PHILOSOPHICAL ECONOMICS

For most of history, the Shiller Cyclically-Adjusted Price-Earnings ratio (CAPE) oscillated in a pseudo sine wave around a long-term (130 year) average of 15.30.  It spent 55% percent of the time above the average, and 45% of the time below–a reasonable result for a metric that allegedly mean reverts.  Since 1990, however, the metric has only spent 2% of the time below its historical average–98% of the time above.

The metric’s failure to mean-revert over the last 23 years hasn’t been for a lack of reasons. The period covered three recessions, two stock market crashes, and one bonafide financial panic–the likes of which hadn’t been seen since the Great Depression.  Even in the worst parts of the 2008-2009 crash–at levels that we now look back on with nostalgia as the “buying opportunity” of our generation–the metric failed to provide an accurate valuation signal.  In an inexcusable blunder, it basically called the market “slightly below fair value” (see the black circle).

If we’re being honest, there are only two possibilities.  Either the “normal” levels of the metric have shifted significantly upwards over the last few decades, or the metric is broken.  There is no other way to coherently explain why the metric has consistently failed to migrate towards its long-term average, or spend any amount of time below it, as it should do every so often in bear markets.  

Which possibility is it?  In my view, both.  The Shiller CAPE, as constructed by its proponents, utilizes inconsistent data.  In this piece, I’m going to explain the inconsistency in rigorous accounting detail, and then share the results of a modified version of CAPE that eliminates it.  I’m also going to illustrate the distortion that changes in dividend payout ratios create for CAPE.  Finally, I’m going to taunt the bears (slightly facetiously) and argue that valuations have probably reached a “permanently high plateau”, to borrow the famously fatal words of Irving Fisher in October 1929.  

There is no question that the current stock market is more expensive than the averages of certain past eras–the 1910s, 1930s, 1940s, 1970s, 1980s, etc.  Looking forward, long-term equity returns will obviously be lower than they were in those eras.  But the market is not as expensive as the Shiller CAPE suggests.  Moreover, there’s no reason to think that the valuations of those eras–distorted by world wars (1914-1918, 1939-1945, 1950-1953), gross economic mismanagement (1929-1938), and painfully high inflation and interest rates (1970-1982)–were somehow more “appropriate” than current valuations.  The valuations in those eras were “appropriate” to the circumstances of those eras; we live in different circumstances.

The Use of Reported Earnings: Inconsistently Measured Data

(Please note that the points below–related to accounting inconsistencies and dividend payout ratio distortions in the Shiller CAPE–are not new.  Jeremy Siegel, legendary professor at the Wharton School of Business, has been raising them publicly since at least 2008.)

The Shiller CAPE was developed by Nobel Laureate Robert Shiller, the well-known originator of the Case-Shiller house price index.  The metric is calculated by dividing an index’s inflation-adjusted price by the average of its inflation-adjusted annual earnings over the last 10 years.

But how does one define “earnings”?  As far as the metric is concerned, the answer doesn’t matter, as long as the definition is consistent across time.  If the definition is consistent across time, then apples-to-apples comparisons can be made between the metric’s present value and its prior values.  The comparisons will give an accurate indication of how cheap or expensive the index is relative to its history, or to what is “normal” for it.

Unfortunately, the earnings data on Dr. Shiller’s website, which are used to build the Shiller CAPE, are not based on a consistent definition of “earnings” across time.  The data are taken from S&P “reported” earnings, which are formulated in accordance with Generally Accepted Accounting Principles (GAAP).  But the standards of GAAP have changed significantly over the last few decades.

One of the most important changes involves how “goodwill” is accounted.  Conceptually, a company’s earnings can be thought of as the change in its book value before dividends are paid.  When one company buys another, the purchase price is almost always higher than book value–usually multiples higher.  But if the buyer pays a higher price for the company than its book value, then his own book value is going to fall in the acquisition.  He’ll be parting with more cash than he’ll be receiving in net assets from the company that he’s taking in.  To avoid the creation of an illusory loss, he is allowed under GAAP to add the difference between the payment price and the book value to his balance sheet as an intangible asset–called “goodwill.”  This addition keeps his own book value constant, and prevents him from having to report an accounting loss.

In the old days, GAAP required goodwill amounts to be amortized–deducted from earnings as an incremental non-cash expense–over a forty year period.  But in 2001, the standard changed.  FAS 142 was introduced, which eliminated the amortization of goodwill entirely.  Instead of amortizing the goodwill on their balance sheets over a multi-decade period, companies are now required to annually test it for impairment.  In plain english, this means that they have to examine, on an annual basis, any corporate assets that they’ve acquired, and make sure that those assets are still reasonably worth the prices paid.  If they conclude that the assets are not worth the prices paid, then they have to write down their goodwill.  The requirement for annual impairment testing doesn’t just apply to goodwill, it applies to all intangible asssets, and, per FAS 144 (issued a couple months later), all long-lived assets.  

The biggest disadvantage to FAS 142 is its asymmetry.  When a company makes an acquisition that it later realizes was a mistake, as happened with Time Warner in its famous purchase of AOL at the peak of the dot com bubble, the company has to book a loss.  In Time Warner’s case, the loss was a record $54B.  But when a company makes an acquisition that turns out to be a huge success–for example, Google’s brilliant acquisition of Youtube–the company doesn’t get to book a profit.  Youtube is still sitting on Google’s balance sheet today at cost, though it is probably worth 10 times the price paid on a fair value basis.

Now, FAS 142 may be a more accurate accounting standard than its predecessor, but that isn’t the issue for the Shiller CAPE.  The issue for the Shiller CAPE is that the accounting standard is not being applied consistently across time.  None of the “reported” earnings numbers used in the Shiller CAPE for years before 2001 were held to the harsh standard of FAS 142.  But all of the “reported” earnings numbers used in the metric for years after 2001 were held to that standard.  Consequently, any comparison between the present value of the metric and pre-2001 values is a comparison between inconsistently measured data points.  The present values end up looking more expensive relative to the past than they actually are.   

You might think that these accounting changes aren’t a big deal.  But they’re a huge deal, especially in the present environment, where the prior 10 year period includes the aftermath of the Tech bubble and the earnings chaos of the Housing Downturn and Great Recession.  These periods were littered with painful writedowns.  In the fourth quarter of 2008 alone, writedowns were large enough to wipe out almost all of the earnings in the index.  If the same standards had been applied during the painful recessions of the mid 1970s and early 1980s, or the M&A binge that followed, reported earnings would have been significantly lower.

We can visually observe the significance of the changes by comparing S&P reported earnings to Pro-Forma (non-GAAP) earnings measures.  Bloomberg offers a time series of trailing twelve month earnings for the S&P 500 (T12_EPS_AGGTE) that dates back to 1954.  The following chart shows the trajectory of this series alongside reported earnings.

For most of history, the two series closely tracked each other.  But since the beginning of the last decade, they’ve significantly deviated, especially in periods around recessions. The biggest reason for the deviation is the introduction of FAS 142, first implemented in 2001, near the break in the chart.  

For a better picture of the sharpness of the deviation, the following chart shows the difference between reported earnings and Pro-Forma earnings divided by Pro-Forma earnings.  

Some have proposed that we fix the problem by excluding recessionary periods from the metric.  This would certainly help remove the penalty that the market faced from 2009 to 2012, when, purely by chance, its prior 10 year period included two big earnings recessions.  But it won’t fix the problem of the inconsistent earnings measurement.  The deviation that has emerged in reported earnings extends to post-recessionary periods as well.

Now, let me say one more time, I’m not arguing that Pro-Forma earnings are more accurate than reported earnings, or that FAS 142 (or 144) is inaccurate as an accounting standard, or that any other accounting change instituted since the late 1990s is inaccurate. The question is not a question of accuracy–it’s a question of consistency.  The reported numbers are not consistent across time, therefore they cannot be used to build a reliable valuation metric.

We can confirm the inconsistency with a second source: profits in the National Income and Product Accounts (NIPA), aggregated by the Bureau of Economic Analysis (BEA) for all U.S. corporations (NIPA Table 1.12, Line 15).  For most of history, there has been a reasonable correlation between NIPA profits and reported earnings for the S&P 500.  As expected, the correlation has broken down over the last decade.  The following table shows the correlations between NIPA Profits, Reported Earnings (GAAP), and Pro-Forma earnings for each completed decade back to 1954.

Notice that the correlation between NIPA and Pro-Forma seen in the decade 2000-2010 is roughly in line with the average correlation seen over the prior 5 decades–72.0% versus an average of 84.3%.  In stark contrast, the correlation between NIPA and GAAP seen in the decade 2000-2010 is more than 3 standard deviations off of the average correlation seen over the prior 5 decades–45.0% versus an average of 85.0%.  The substantial, isolated deviation between NIPA and GAAP in the decade 2000-2010 is conclusive proof that GAAP standards changed materially during the period.

Fixing the Shiller CAPE

The Shiller CAPE is one of the few metrics that provides a non-cyclical indication of market valuation.  Such an indication is particularly useful during recessions, when the market’s true value is hidden by temporary corporate weakness.

There are other non-cyclical metrics that we can use to assess valuation–two examples are price to book (or Q-ratio) and price to sales (or market capitalization divided by GDP).  But these metrics carry the disadvantage of being static with respect to profit margins.  The Shiller CAPE, in contrast, is dynamic.  To illustrate the difference with a relevant example, suppose that an economy sees structural reductions in its corporate tax rates and interest rates that lead to structurally higher profit margins and (deservedly) higher stock prices. Valuation metrics based on price to book and price to sales ignore the bottom line, and therefore cannot detect this change.  In the presence of the higher stock prices, these metrics will forever show the market as “overvalued.”  The Shiller CAPE, however, takes an average of earnings over time, and therefore will detect the change, especially if it is gradual.   

Fortunately, we can fix the inconsistency in the Shiller CAPE by building the metric with the Pro-Forma earnings data set rather than the GAAP earnings data set. The following chart shows the Pro-Forma CAPE alongside the GAAP CAPE, going back to 1954 (to calculate the average earnings in the 10 year period prior to 1954, we used GAAP for both metrics, since the two were very close in that general era):

With the S&P at 1775, the GAAP CAPE is currently at 24.51, a frightening 60% above its historical (130 year) average of 15.30.  The Pro-Forma CAPE, in contrast, is currently at 20.63, a more modest 19% above its historical (59 year) average of 17.35.  To be fair, the Pro-Forma data doesn’t extend past 1954, and so the Pro-Forma CAPE average doesn’t include the depressed valuations of WW1, the Great Depression, and WW2, as the GAAP CAPE average does.  But in terms of accuracy, that’s a good thing.  As we’ll see later, the market dynamics of those eras are of little relevance to the present era.

The following chart shows Pro-Forma CAPE relative to its average since 1954.

To get to fair value on Pro-Forma CAPE, we would need a garden-variety correction of 15%, down to around 1500 on the S&P.   But to get to fair value on GAAP CAPE, we would need a plunge of 40%, down to around 1100 on the S&P.  This difference has crucial implications for tactical investors.  A bearishly-inclined investor who pulls out of the present market in the hope of seeing the S&P revert to its “fair value” of 1500 is at least being realistic (whether you agree with the tactical call is obviously a different question).  But a bearishly-inclined investor who pulls out of the present market on the expectation that the S&P will revert to its “fair value” of 1100 is being ridiculous.  It’s fine to make calls for those kinds of levels–we all know that shit happens in markets.  But let’s not make them on the basis of considerations that are as fickle and unreliable as “valuation.”

Comparing the Predictive Power of GAAP and Pro-Forma Constructions 

Proponents of the Shiller CAPE will argue that the true measure of a valuation metric is not whether it makes intuitive or logical sense, but how well it correlates with future returns. As shown in the table below, the Pro-Forma CAPE and GAAP CAPE exhibit essentially the same correlation with future returns.  The reason is that they are very similar for most of the historical data set.  They only begin to appreciably deviate from each other in 2001.

Notice that for the period after 2001, the Pro-Forma CAPE correlation with future returns is slightly stronger. We don’t have data on 10 year total returns past 2003, and so this fact doesn’t tell us very much–the size of the data sample is only two years.  However, it’s clear that as additional 10 year total return data come rolling in over the next few years, the Pro-Forma CAPE is going to continue to outperform the GAAP CAPE.  Since 2001, the GAAP CAPE has been notably unreliable as a predictor of future returns.  The Pro-Forma CAPE, in contrast, has been quite reliable.  

At the bear market low of spring 2003, when the market was a reasonably attractive buy, the GAAP CAPE called it 40% overvalued.  The Pro-Forma CAPE called it 12% overvalued, in line with the respectable–but not spectacular–returns that it ended up producing.  

At the panic low of March 2009, when the the market was a screaming buy, the GAAP CAPE called it a mediocre 15% undervalued.  The Pro-Forma CAPE called it 40% undervalued–much closer to what we would expect from a historical bear market low, and consistent with the fantastic returns that the market subsequently produced (north of 20% annualized).

Notably, in March 2009, on a pro-forma CAPE basis, the market traded down to early 1980s valuation levels. That is something we would expect, given that the early 1980s recession and the 2008-2009 recession were of similar intensity (with the latter arguably more intense).  But on a GAAP CAPE basis, the market didn’t even come close to reaching early 1980s valuation levels–to the contrary, it barely broke below the valuation highs of the 1980s.  

Changes in Dividend Payout Ratios

When the corporate sector earns money, it can pay the money out to its shareholders as dividends, or it can deploy the money internally to generate growth in future earnings (via investment, acquisitions, and share buybacks).  All else equal, the more the corporate sector favors dividends, the lower the Shiller CAPE will be.  The more it favors deploying earnings into investment, acquisitions, and buybacks, the higher the Shiller CAPE will be. That’s just the way the math of the Shiller CAPE works out–to the extent that earnings growth is reflected in present valuation, it is penalized.

To illustrate, suppose that we live in a zero growth, zero inflation world in which all stocks trade at 17 times earnings.  Consider a hypothetical company in this world that pays out 75% of its earnings as a dividend, and uses the other 25% to buyback shares on the open market.  As shown in the table below (using dummy data), the company’s Shiller CAPE for the period will end up being 18.16.

Now, consider the ten year history of a second company that is identical to the first company except for one feature: it pays out 25% of its earnings as a dividend, and devotes the other 75% to buybacks (the ratios are switched from before).  As shown in the table below, the second company’s Shiller CAPE for the ten year period will end up being 20.65.

But these companies are identical in all respects–they have the exact same businesses and trade at the exact same P/E multiples–17.  Why, then, does the Shiller CAPE for one of them end up being 14% higher than the other?  The answer is that because the second company chooses internal reinvestment–buybacks–over dividends, it ends up with higher EPS growth over the period, and therefore a higher final share price (because they both trade at 17 times trailing earnings).  Thus the second company ends up with a higher Shiller CAPE, even though its true valuation is the same.

The structure of the Shiller CAPE unfairly penalizes the corporate sector for reinvesting profit into EPS growth instead of paying dividends.  But that is exactly what modern corporations do in comparison to corporations of the past: they provide a return to their shareholders by reinvesting profit rather than by distributing it.  From 1954 to 1995, the S&P 500 dividend payout ratio averaged 52%, while the real EPS growth rate averaged 1.72%.  From 1995 to 2013, the S&P 500 dividend payout ratio averaged 34%, while the real EPS growth rate averaged 4.9%.

To make comparisons between present and past values of the the Shiller CAPE, we need to normalize for differences in payout ratios.  A crude way to do this is to note that at the current trailing twelve month P/E ratio–around 17–the difference between a 52% payout ratio (the average of 1954-1995) and a 34% payout ratio (the average since 1995) corresponds to around 1 point worth of Shiller CAPE.  Taking 1 point off the current value of the Pro-Forma Shiller CAPE, we get an adjusted value of 19.63.  On this basis, the market would need to correct to approximately 1550 to get back to its average valuation of the last 60 years.  Unpleasant for longs, but hardly catastrophic.  We were there just a few months ago.

A Permanently High Plateau

There is no external, divinely-imposed valuation level that the stock market has to take on.  Rather, the stock market takes on whatever valuation level achieves the required equilibrium between those that want to get in it, and those that want to get out of it.  At all times, every investor that wants to get in the market needs to connect with an investor that wants to get out of it.  If there are too many that want to get in, and not enough that want to get out, the price will rise until the imbalance is relieved.  If there are too many that went to get out, and not enough that want to get in, the price will fall until the same.  The process is reflexive–investors want to get in or out based on where the price is and what it is doing, but they also make the price be where it is and do what it is doing, through their efforts.

For this reason, context–the set of environmental variables that shape investor outlook and risk appetite, and that influence the preference to be in or out, given the price–is crucial to normative claims about valuation.  A valuation level that is “appropriate” in one context–adequate to achieve the required equilibrium–may not be “appropriate” in another.

Consider the following chart of the GAAP CAPE back to 1881:

Notice that the periods of below-average valuation, circled in black, generally involved three different types of environments: (1) war (destructive violence between countries or within a country), (2) high inflation (with tight monetary policy and high interest rates), and (3) financial crisis (with debt deflation and deep recession).  These environments, which we will call The Big Three, represent classic, recurring challenges for the stock market.  They create fear, pessimism and malaise on the part of investors, and serve as catalysts for deeply depressed valuations.  We can see their damaging effects not only in U.S. market history, but in the history of stock markets all over the world.

The major bull markets of this century and the last century were each preceded by at least one of The Big Three.  The bull market of the 1920s, for example, began after the victory in World War 1, as the economy moved out of the severely deflationary downturn of 1920 and 1921.  The 1930s bull market began at the end of the Great Depression, after FDR took office and ended the gold standard, allowing the Federal Reserve (Fed) to inject desperately needed liquidity into the financial system.  The 1950s bull market began in the years after World War 2, as the country worked through inflation challenges and monetary policy disputes and successfully transitioned into a peacetime economy.  The bull market stumbled during combat in the Korean War, and then moved into full speed after the armistice of 1953-1954.  The 1980s bull market began in 1982, when Paul Volcker finally conquered inflation (or so the narrative goes), making it possible for the Fed to shift to a looser monetary policy.  The 2009 bull market began at the resolution of the acute phase of the Great Recession, as the Fed and the Federal Government aggressively collaborated to stabilize the banking system and restart the economy.  

Notably, each of the ensuing bull markets involved a sustained period in which the economy saw the opposite of The Big Three: (1) peace, (2) low inflation with low or falling interest rates, particularly at the short end of the curve, and (3) stable, expansionary growth in an environment of financial stability.  It’s hard to think of any time in history when these three conditions were met and where the economy was not in a rising bull market, with a “natural” bias for higher valuations.

Granted, the market saw corrections (1962, 1987, 2011) and there will surely be corrections in this market going forward.  But there were never bear markets of the magnitude that would be necessary to bring the current Shiller CAPE back down to its long-term historical average.  The only real exception to this point was the 2001-2003 downturn.  But in that period, the market was unwinding a legitimate investment mania. Valuations did not “mean revert”–rather, they fell from egregious, unconscionable levels to levels that were just “expensive.”  Recall that the theme of war was a relevant catalyst for the move: there were the 9/11 terrorist attacks, the operations in Afghanistan, and the invasion of Iraq, the runup to which helped push the market to its ultimate low for the period.

Investors that are patiently waiting for the Shiller CAPE to “mean revert” from the elevated level that it has hovered at over the past few decades, towards its long-term average, are implicitly calling for at least one of The Big Three to recur (either that, or something new that markets have never seen).  The dominant presence of The Big Three in market history is the very reason that the long-term historical average of the Shiller CAPE has the value that it has–without them, its value would be higher, in line with or above the levels of the current era.

But why do The Big Three have to recur?  If they do recur, why do they have to recur with the same frequency and intensity?  If they don’t recur, why does some other bad thing have to occur in their place?  Why can’t human beings make progress?

Think optimistically for a moment.  What if large scale war is a thing of the past?  The suggestion might sound naive, but there is significant evidence to support it.  The human species has become dramatically less violent and war-prone as it has advanced intellectually, technologically, and economically.  In the modern era of globalization, the idea of two advanced countries–the U.S. and China, for example–fighting each other in a real, no-kidding war is almost inconceivable.  

What if low inflation and low interest rates are a permanent fixture of the modern economy, rather than something temporary?  Trend inflation has been falling for over 30 years.  Population growth is slowing, society is getting older, and therefore there’s less of a need to build and invest for the future.  The process of building and investing for the future–a process that puts pressure on the present supply of labor–is arguably the main driver of inflation in a normal economy.

The current lack of inflation is made worse by advances in technology that have reduced the marginal value of labor, and by a process of globalization that has created an excess supply of cheap labor internationally.  Moreover, a decline in union influence has significantly reduced the bargaining power of workers.  If workers don’t have bargaining power, wage-price spirals can’t occur, and neither can meaningful inflation.

The Fed has absolute control over short-term interest rates, and significant control over long-term interest rates.  It sets the former, and influences the latter, based on the level of inflationary pressure that it sees in the economy.  If structural changes in the economy mean that there isn’t going to be meaningful inflationary pressure going forward, then interest rates can conceivably stay low forever.    

What if deflationary financial crises are once-in-a-generation events?  What if each time we have such crises, policymakers and the community of economists learn valuable lessons about how the system works, so that they can avert future crises, or at least address them more effectively?

Once again, the historical evidence provides strong support for such a view.  The 2008 crisis had all of the makings of a new Great Depression.  In terms of the amount of unstable private sector debt that existed, the starting point was actually worse than the Great Depression.  But unlike in 1930, policymakers in 2008 quickly arrested the downward spiral–fiscally, monetarily, and especially through their controversial actions to stabilize the banking system.  Now, here we are, only a few years later, in an economy that is growing healthily, with output well above the prior peak, arguably on the verge of the strongest expansion the country has seen since the mid 1990s.  That’s a remarkable achievement, a reason for admitting that progress in economic policymaking is actually possible.

The reason that policymakers were able to bring about a different outcome in 2008 than in 1930 is that the field of economics advanced dramatically in the interim period. Policymakers learned the lessons of the Great Depression, and emerged with a much better understanding of how the system works.  The good news is that in the present crisis–to include what is happening in Europe and Japan–new lessons are being learned. These lessons will help to avert future crises (or at least lessen their severity).

If you asked me what the biggest long-term threat to the U.S. economy is, my answer would be demographics.  Over the next few decades, as our economy ages demographically, it runs the risk of going too far on the low inflation front–and entering deflation, which would eventually lead to financial crisis.  But, again, it’s important to anticipate progress.  Japan is experimentally fighting the problem of demographic deflation as we speak.  It is going to learn critical lessons that we will be able to draw from should we ever face a similar situation.

Deflation is an unacceptable condition for an economy.  If it occurs, policymakers will attack it with increasingly extreme monetary policies: “whatever it takes.”  These policies tend to provoke higher than normal valuations.  And so, ironically, even if the U.S. does eventually enter a Japanese-style deflation, Shiller bears are unlikely to be vindicated for very long.

It sounds overly ebullient to propose, as Irving Fisher famously did, that stocks have reached a new era of elevated valuation.  But the point needs to be taken seriously, as there are strong reasons to believe it.  The historical record leaves no other option but to admit that something about valuation has changed.  The market has spent more than 20 years at a historically elevated Shiller CAPE.  A 20+ year period is way too long to dismiss as an “outlier.”

Instead of assessing the valuation of the current market by making casual comparisons to the markets of the 1910s, 1930s, 1940s, and 1970s, we need to make comparisons to markets that shared relevant similarities with our own.  Examples include the market of the late 1950s and 1960s, the market of the 1990s (excluding the mania that emerged after 1997), and the post-bubble market of the 2000s (excluding the financial crisis).  If the current market were “attracted” to any kind of valuation, it would be to the kind of valuation seen in those periods, which were marked by peace, low-inflation growth, easy monetary policy, and financial stability (as opposed to wars, inflation spirals, punitively high interest rates, and financial panics.)

The following table shows the average and peak Pro-Forma Shiller CAPEs for each of the periods:

With the S&P at 1775, the current value of the Pro-Forma Shiller CAPE is 20.63.  To compare the value with the 1955-1969 period, we need to adjust it for differences in the dividend payout ratio.  From 1955-1969, the dividend payout ratio averaged 55%.  At a P/E of 17, the difference between 55% and the current value of 34% amounts to roughly 1 point worth of Shiller CAPE.  Subtracting, we get an adjusted Shiller CAPE of 19.63.  That number almost perfectly matches the average for the 1955-1969 period, a period in which market valuations were generally fair and reasonable.

To reach the valuation high of the 1955-1969 period–23.88, which was registered in January 1966–the current S&P would have to rise to 2125.  To reach the high of the 2003-2007 period, which was registered in January 2004, the S&P would have to rise by a slightly higher amount, to 2130 (note that there is no need for a dividend payout ratio adjustment in this case).  To reach the high of the 1990-1997 period, which was registered in December 1997, the S&P would have to rise to 2900.

Admittedly, a call for Shiller CAPE to go to 33.82 and for the market to go to 2900 is pushing it–not a smart bet.  But there’s no reason why the market shouldn’t at some point go back and touch the valuation peaks that it reached in the other comparable periods.  It’s showing every sign of wanting to do so.

It’s also going to fall appreciably at some point, as all markets do, but it’s unlikely to fall in a way that would sustainably restore “historical averages” and vindicate Shiller bears.  In their boycott, Shiller bears are making a blind bet on the mean reversion of a poorly constructed metric, without paying attention to context–the set of variables that drive the preferences of market participants to be in or out, and that determine the valuation that the market naturally gravitates towards.  The outcome they are calling for requires the market’s context to return to the low points–war, tight money inflation, financial crisis–of prior eras.  The reality of human progress reduces both the likelihood that such a return will occur, and, in the specific case of financial crisis, the intensity and duration of the pain that it would bring.

In this piece, I’m going to derive and explain the Kalecki-Levy Corporate Profit equation, and then demonstrate its truth empirically using data collected by the BEA.  I’m then going to present tables that show the values of the components of the equation, as a percentage of GNP, from 1929-2012 annually, with recessionary periods shaded in.  I’m going to conclude with a brief comment on mean reversion.

Profit Equation Derived and Explained

Let’s define some terms.

• Saving means increasing one’s wealth. (And dissaving means depleting it).

• Hoarding means accumulating already-existing wealth.

• Investment means creating new wealth that didn’t previously exist.

• Wealth can mean whatever you want it to mean, as long as the definition is applied consistently, and on a net basis. Usually, when we speak of wealth, we mean things that are of economic value to human beings, to include both real and financial assets.  One’s wealth is constituted by the real and financial assets that one owns, net of one’s liabilities.

Individuals within an economy can save by hoarding, or by investing.  Suppose that I earned $100 last year and put all of it into a piggy bank.  Over the period, I’ve increased my wealth–saved–by hoarding, by accumulating money that already existed.  Now, suppose that instead of putting the $100 into a piggy bank, I use it to pay for the construction of a new home for myself.  In this case, I’ve increased my wealth–saved–by investing.  I’ve created a new source of wealth–shelter–that didn’t previously exist.

Hoarding is a zero sum game.  To  illustrate, suppose that you and I are the only people in the economy.  You have $100, and I have $100.  You take your $100 and give it to me as payment in exchange for cooking you a meal to eat.  If I save the $100 by hoarding–by putting it into a piggy bank, I will be taking it out of circulation.  It won’t be coming back to you.  Thus, my saving of the money will necessarily come at the cost of your dissaving.  My wealth will increase from $100 to $200, but only because your wealth decreased from $100 to $0 (what you ended up with after sending money to me and never getting it back).

Investment, in contrast, is a positive sum game.  If I take the $100 that you paid me, and I pay it back to you in exchange for building me a new home, you will get the money back. Because I saved by investing, my saving will not require you to dissave.  I will end up with increased wealth–my original $100 plus the home that was built–and you will up end where you started, with your original $100.

Because hoarding is a zero sum game, the overall economy–considered in aggregate–can only save by investing.  What follows is a fundamental macroeconomic identity:

(1) Saving = Investment

This identity applies only to the overall economy.  It doesn’t necessarily apply to individuals or entities inside the economy.

Now, let’s arbitrarily divide the economy into four aggregate entities: All U.S. Households, All U.S. Corporations, the U.S. Government at all levels (Federal, State, Local), and the Rest of the World (ROW).

Trivially, the saving of the overall economy is the saving of each of these entities.  Thus,

(2) Saving = Household Saving + Corporate Saving + Government Saving + ROW                Saving

Now, Corporate Saving is simply Corporate Profit (wealth accrued by corporations over a period) minus Dividends (wealth paid out to shareholders in that period).  Thus,

(3) Corporate Saving = Corporate Profit – Dividends

Substituting (3) into (2), and (2) into (1), and rearranging, we get,

(4) Corporate Profit = Investment + Dividends – Household Saving –                                      Government Saving – ROW Saving

This is the Kalecki-Levy profit equation, a macroeconomic accounting identity derived by Jerome Levy in 1908.  What it says is that if dividends are held constant, then any increase in savings that the non-corporate sector conducts that is not offset by an increase in investment (from either the non-corporate or the corporate sector) will require the corporate sector to save less–i.e., increase it’s wealth less, have less profit.  Note that governments save and dissave by running government surpluses and deficits, and that the ROW saves and dissaves by running trade surpluses and deficits, net of other transfers.

A few points of clarification:

• The Kalecki-Levy profit equation describes a required result.  It doesn’t describe how the economy gets to that result.  The variables that determine how much companies earn in profit are the same as they’ve always has been–revenue, expenses (depreciation, wages, interest, etc.), and taxes.  These variables are the true drivers of changes in profits.  However, when they change, terms in the equation change.  So, for example, if the government eliminates the corporate tax, profits will jump.  But the government will be losing revenue.  To keeps its spending constant, it will have to dissave–increase its deficit.  Now, if the government reduces its spending by the same amount as the lost taxes, so as to keep its deficit unchanged, the corporate sector will lose that amount in revenue.  And so the entire process will end up being a wash for profit–what the corporate sector gains in lower taxes, it will lose in lower revenue. 

• Borrowing doesn’t imply dissaving.  You can borrow to invest–for example, to build a home.  In that case, your wealth stays constant.  You take on a debt, but you receive an asset in exchange for it.  The kind of borrowing that involves dissaving is borrowing to consume.  When you borrow to consume, you take on a debt, but end up with nothing to show for it–and thus you deplete your wealth.

• Investing doesn’t mean trading already-existing securities such as stocks and bonds.  It means the creation of new wealth that didn’t previously exist.  Define “wealth” however you wish–it doesn’t matter, as long as the definition is applied consistently.  But to count as investment, the wealth must be new.

• Operationally, the Government is in a different place from the other sectors because it (normally) can create money directly.  It therefore has an infinite capacity to borrow and dissave.  More importantly, it doesn’t exist for the sake of its own wealth–it exists for the sake of the wealth of households.  For this reason, in periods where the desire to save is high, and the desire to invest is low, it has an obligation to the other sectors to dissave by running deficits.  Its dissaving prevents the economy from seizing up as everyone tries to hoard the same assets in a zero sum game.  It can be helpful for the government to run smalls deficits even during prosperous times, to ensure that there is enough cushion in the system to allow other sectors to save.  Other considerations aside, we all want to increase our wealth–and our prosperity–over time.  The government exists, in part, to help support us in that effort.

Profit Equation Empirically Confirmed

The profit equation can be confirmed empirically using data from the National Income and Product Accounts (NIPA).  It doesn’t need to be confirmed, because it’s true by definition. But confirming it can provide us with a second check, to make sure it’s right.  The fact that it can be confirmed is a testament to the prowess of the Bureau of Economic Analysis (BEA)–that it would be able to gather data for a 300 million person economy with trillions of dollars in annual transactions, and that the results would fit together so closely in the final analysis.

We’re going to take each saving and investment term net of depreciation (the wealth that is lost naturally over time as assets depreciate).  We’re also going to use corporate profits after taxes.  The following table gives the NIPA references for each of the terms, available on the BEA website:

To normalize, we divide each term by Gross National Product (GNP) (NIPA Table 1.7.5 Line 4).  The reason that we use GNP rather than Gross Domestic Product (GDP) is to be consistent.  GNP refers to the total production supplied by the labor and property of U.S. resident entities, whether the production takes place inside U.S. domestic borders, or abroad.  GDP, in contrast, refers to the total production that occurs inside U.S. domestic borders, whether the production is supplied by the labor and property of U.S. resident entities, or of foreigners.  The corporate profit term that we used was corporate profit for all U.S. corporations, regardless of where the profit was earned; that term feeds directly into GNP, not GDP.   

The following chart shows Corporate Profits as a percentage of GNP from 1929-2012, calculated using the Kalecki-Levy Equation and gathered directly by the BEA in NIPA.

We can construct the same chart using terms available in FRED (click here for link).  

NIPA Table 5.1, Savings and Investment, has a discrepancy term to account for errors in data gathering and statistical estimation.  Unsurprisingly, when the discrepancy term is subtracted out of profit, the numbers match exactly.

A History of Corporate Profits, 1929 – 2013

The following table show the components of the corporate profit equation annually from 1929 to 2012, neglecting statistical discrepancy.

NYSE Margin Debt just reached another record high, and an increasing number of market skeptics are expressing concerns.  They reason as follows.  A willingness to buy stocks on margin suggests confidence and optimism.  But markets don’t perform well when investors are already confident and optimistic.  Moreover, people invested on margin are less able to tolerate price fluctuations.  As their presence in a market grows, so does the risk that otherwise healthy challenges to the market’s advance will provoke rapid, self-fulfilling unwinds of positions.

To ground these concerns, the skeptics appeal to history.  They point out that extreme margin borrowing was the primary cause of the 1929 stock market crash, and that the market crashes of 2000 and 2007 both involved blow-off peaks in margin debt.

If you think the chart above looks frothy, I have some charts to show you that are even frothier.  To start, take a look at the chart below, which shows margin debt alongside a market index from an unspecified period of history.  The x-axis is in years.

It’s clear that investors in this chart are not respecting downside risk.  They’ve pushed their margin debt levels well above the $70B-$80B level that presaged nasty market crashes at prior peaks (years 6 and 11). To make matters worse, the index itself is showing signs of weakness, failing near the prior peaks and now slouching downward.  If everyone is already maximally invested in this market, as evidence by the large jump in margin debt, who is going to buy it when it finally tips over, and the forced selling begins?

If you were an investor in this market, how would you position yourself?  Take note of the answer, and we’ll fast forward through time to see how things turn out.  The following chart shows the market’s evolution over the next five years.

Wait, what is going on here?  The margin debt was supposed to fuel another ugly correction, but that hasn’t happened.  It pulled back briefly, but now it’s right back up at a record high, 25% higher than when we initially became concerned.  Is the herd being smart here?  Is this going to end well?   

Now, to the moment of truth.  Let’s suppose that you are a participant in this market holding cash.  What would you choose to do?  Would you jump in alongside the rest of the margin-heavy buyers, or would you step aside in favor of caution, discipline, patience, restraint–the venerable traits that “wise” investors demonstrate?

If you would choose patience and restraint, I’ve got news for you: you would be missing out on one of the greatest bull markets in history.  Far from some bubble about to go bust, what we’ve been charting and discussing is the S&P 500 from January 1963 to March of 1983.  To conceal the chart’s identity, I’ve multiplied the index and margin debt level by a factor of 10.  The following chart shows the true numbers of each.

The following chart offers a longer-term perspective:

An investor can sound wise and insightful by highlighting the risks of record margin borrowing, suggesting that individuals are repeating the mistakes of 1929, but the argument is no more valid now than it was when it was made in August of 1983:

Or when it was made in March of 1979 (what turned out to be a magnificent long-term entry point for investors):

Or two months later, in May of 1979:

To be clear, I’m not suggesting that the October 2013 market opportunity is analogous to the market opportunities of August 1983 or March 1979.  Not even close.  The valuations, earnings growth prospects, and demographics that underpin the current market are substantially different from what they were then, and the future returns are almost certain to be lower.  But the example illustrates a deeper property of margin debt that those who worry about it often forget.  Margin debt rises with the market’s price level and total capitalization.  It always has.  The reason that margin debt is at a record high right now is that the market is at a record high.  That is also the reason it was at a record high in 1983, and continued to make new highs for years thereafter.  

The following chart shows the annual price change in the S&P 500 alongside the change in NYSE margin debt from January 1960 to present.  The correlation is clear:

When we scrutinize the chart closely, we see that there have been a number of times in history when margin debt increased significantly faster than the S&P 500.  This happened, for example, in August 1972, in March 2000 and in July 2007.  It goes without saying that those were not ideal times to be taking on market exposure.  One might therefore argue that if margin debt is quickly rising without a similar rise in the index level, then investors should proceed with caution.  But even if the argument is true, it doesn’t matter right now, because we’re not in that situation.  The trends we’re seeing with respect to the growth of margin debt are perfectly normal for a bull market.  Continual new highs in margin debt are happening alongside continual new highs in the market itself, and the pace of growth of the two are not far off.  Certainly, the new highs we’ve been making, and the speed at which we’ve been making them, could be increasing the risk of future downside.  But the driver of the increased risk is the price advance itself, not the increase in margin debt.

Now, why does total margin debt rise with the level and total capitalization of the market? There is a simple, intuitive answer.  In any environment, a certain percentage of investors borrow against their portfolios.  They borrow for a number of reasons.  Examples include: (1) They may be engaged in investment strategies that hedge and pare risk by applying leverage to uncorrelated assets. (2) Margin debt might be the cheapest type of debt they have access to, and therefore they may use it to pay off other more expensive debts. (3) They might be waiting for money to arrive at a broker, and need temporary liquidity to fund purchases. Or (4), they might just be really bullish and really reckless.

We should expect the amount of margin debt that these investors take on to vary with the size of the portfolios they are borrowing against.  Thus, we should expect the total quantity of margin debt in existence to vary with the total capitalization of the stock market (the sum of the value of all equity portfolios).  That is roughly what we see.

The following chart shows NYSE margin debt as a percentage of the total market capitalization of all equities (Z.1: market value of financial plus non-financial corporate equity liabilities).

This chart is the only margin debt chart that matters.  It puts worries about margin debt into proper perspective.  When we talk about margin debt, we’re talking about a tiny portion of the aggregate investor portfolio.  That portion has hardly changed since the 2009 low–and in total portfolio terms, it has hardly changed since the late 1970s.  

Suppose that the S&P is at 1350, and that you own 100 shares of it.  You’re really bullish, so you decide to go 200% long in your portfolio, adding another 100 shares on margin. Shortly after you add the shares, a 20% correction occurs.  The S&P falls from 1350 to 1080. The odds are very high that you’re going to be forced to sell–if not by your broker, then by your amygdala.  Your situation in that respect is no different from the situation of a bullish investor who owns 100 shares of the S&P at 1700 and adds another 100 on margin before getting caught in a 20% correction to 1360. Nominally, he has more debt than you, but he also has a larger net worth.  In portfolio-adjusted terms, the level of debt is the same, and so is the loss.  Therefore, the felt pressure to unwind is the same.   

Now, one can argue that a 20% correction is more likely from 1700 than from 1350 because 1700 is a richer valuation.  The point conflicts with the data, of course, given that we actually experienced a 20% correction from 1350 in 2011, and haven’t come close to experiencing one since, despite much loftier valuations.  But the point makes sense: more important considerations aside, valuation matters, and 1350 on the S&P is more attractive than 1700.

However, if the criticism of the current market is its valuation (a potentially legitimate criticism that Bulls should take seriously, even if they disagree), then there is no reason to bring up margin debt.  Margin debt is not what distinguishes current market conditions from those of ’09, or of ’11, or of ’84, or of ’86.  As a percentage of assets, margin debt has not substantially changed relative to those years.  Aside from the dramatically reduced interest cost, the debt looks and feels exactly the same to present market participants as it looked and felt to market participants back then–representing, in aggregate, between 1.4% and 1.6% of total equity assets.

The reader will notice that margin debt as a percentage of total stock market capitalization has gradually increased since the early 1990s.  Market skeptics will want to attribute this increase to “froth” and “lack of discipline.” But there is a better explanation. Hedge funds have grown dramatically since the early 1990s.  The strategies they employ to smooth out and optimize their risk-adjusted returns involve more leverage than the rest of the market has traditionally used.  Additionally, the cost of borrowing against a portfolio has fallen significantly since the early 1990s.  Borrowing on margin is cheaper now than it has ever been, not only because interest rates are at record lows, but also because brokerage competition has produced an outcome where clients are offered much better terms.  Finally, with the development and mass expansion of online trading, portfolios are easier to monitor and quickly adjust.  This reduces the stress of being on margin.  

To summarize, it’s usually a waste of time for investors to worry about margin debt levels.  For the most part, margin debt just follows the level of the market itself.  If it is at record highs, that is probably because the market is at record highs.  If  your strategy is to never buy a market with a record amount of margin debt in it, then, effectively, your strategy is to never participate in secular bull markets.  Not a very good long-term strategy.

There may be a relationship between how quickly margin debt increases relative to the market’s advance and the probability of subsequent corrections, but there are too many confounding variables and too many historical counter examples in the data for us to be confident about it.  More importantly, we as investors are unlikely to be able to convert the relationship into an actionable investment strategy.  The proof is in the data–margin debt and the change in margin debt show essentially zero correlation to future returns.  One cannot build a simple, intuitive, non-data-snooped market timing strategy that outperforms using margin debt as the timing criterion.  For every correction one avoids, one will miss even greater market advances.  Regardless, the question doesn’t matter to present market conditions, since we’re not in a situation where margin debt is flagging anything unusual.  It’s current trajectory is perfectly in-line with historical norms for a bull market.  

With all of that said, we have to remember that the market is ultimately a Keynesian Beauty Contest.  What matters is not what is actually the case, but what market participants believe (or will eventually believe) is the case, and how they choose (or will eventually choose) to act on that thinking.  In the United States, the experience of 1929 has ingrained the dangers of excessive margin borrowing into the market’s moral ethos.  Consequently, “too much margin” has become a story that investors instinctively tell to explain why the market falls (usually after the fact).  The story was told in the 2000s, in the 1990s, in the 1980s, in the 1970s, in the 1960s, as far back as you look.  For the most part, it’s not true.  But it doesn’t necessarily need to be true to matter; sometimes, it just needs to be believed by enough people.  And so if the chart of margin debt starts to do funky things that genuinely scare the investment community (we’re not there yet), we will need to at least pay attention, because that narrative, even if wrong fundamentally, can affect behavior, and can therefore become true reflexively.

At each quarterly meeting, FOMC members are asked to state what they think would be an appropriate forward path for the Fed Funds Rate, given their present expectations about future inflation and unemployment.  The following table and chart show how the averages of these estimates have evolved over the last three meetings:

Suppose that FOMC projections turn out to be accurate.  Rates rise through 2015, and then stabilize at long run levels.  For each of the different future path estimates, we can ask, how much would $100 of cash earn over the next 10 years under that estimate? If we assume that the interest on the outstanding cash balance is paid on the first of the year, at the prior year’s interest rate, and that the transition from ~1% to the long run interest rate occurs over two years, consistent with the 2003 – 2007 cycle, we get the following result: 

Cash under the June path estimate would earn slightly more than cash under the March path estimate, and cash under the September path estimate, released today, would earn less than both.

Now, to get an idea of what sort of effect on long-term bond yields the differences between the path estimates might be able to justify, we can ask, what would a brand new $100 face value 10 year risk-free bond–a symmetric alternative to holding $100 of cash for 10 years–need to yield to match the cash return under each rate path?  If we assume that the coupon is held as cash and reinvested at the Fed Funds rate, we get the following result.

So the yield on a 10 year risk free bond that would be necessary to match the yield on cash is about 20 bps higher on the June path estimate than the March path estimate.  Did the market respond to the March-June Fed shift by increasing the 10 year bond yield by 20 bps?  Hardly.  On May 1st, a little less than a month and a half after the March meeting, the 10 year treasury yield was trading at a ridiculous 1.64%.  On August 1st, a little less than a month and a half after the June meeting (which all but cemented the Fed’s “taper” plan), it was trading at 2.72%.  The market adjusted to the news of the Fed’s plan by increasing the 10 year yield 108 bps, when only 20 bps was necessary (assuming anyone had taken the path guidance seriously).

This result helps to illustrate the mechanism through which QE lowers long-term interest rates.  Surely, there are stock and flow effects.  If you reduce the supply of bonds that investors can invest in, and increase the supply of zero-interest money that they must hold, you are going to push bond yields lower.  Similarly, if you enter a market as a dominant player and promise to buy a large number of securities, you are going to push the prices of those securities higher.  But far more powerful than either of these effects is the behavioral impact of QE, the way it influences investors’ perceptions of the Fed’s intentions and their expectations about likely future Fed policy.  Bond yields rose in May and June because of that effect.  

The Fed’s perceived insistence on “tapering” in the presence of a still-weak economy created doubts in the investment community about its commitment to a dovish, pro-cyclical policy stance.  The selling in the bond market and consequent rises in yields–that the Fed seemed to be OK with–made the shift feel acutely real to investors.  A reflexive feedback loop then took hold.  More and more investors began to question whether bonds were a smart place to be invested, which led to more doubts, more selling, lower prices, more discomfort, more questioning, more doubts, more selling and so on.  

Some have proposed that the Fed should counteract this process by telling the market that it intends to keep rates lower for a longer period of time.  But talking is easy to do, and can be taken back in the future on a moment’s notice.  QE, in contrast, is an overt action.  It cannot be feigned or faked, nor can it be easily taken back.  For all practical purposes, once the Fed’s balance sheet is increased, it has to stay increased until the securities roll off (to try to sell them would create chaos).  Thus, when the Fed aggressively expands its balance sheet, when it steps into what the market views as “uncharted territory” and exposes itself to controversy and criticism, it demonstrates its commitment to an easy policy, its willingness to stay true to such a policy despite the risks.  It puts its money where its mouth is–quite literally.

In 2003 and 2004, a debate emerged in academic and financial circles over the appropriateness of the Fed’s “easy money” response to the crash of the tech bubble and the ensuing recession.  Hawks argued that the Fed’s low rate policy was inevitably going to create another asset bubble, and ensure a repeat recession in the future when it bursts. Doves argued that if the Fed imposed tight money on the economy in the lingering aftermath of a recession, simply to eliminate the risk of a future bubble, the recession would turn into a depression. Both were right.

Unsurprisingly, a similar debate is taking place now, in the aftermath of the crash of the housing bubble. The difference, of course, is that the dovish case for monetary stimulus is stronger now than it was then.  The US economy is organically weaker and more inclined towards recession than in 2003 and 2004, and it faces stronger non-monetary headwinds. In terms of bubble risks, conditions in the US economy are not conducive to the formation of another asset bubble.  The two primary candidates for asset bubbles–stocks and homes–are tarnished by their recent track records.  Investors have vivid memories of what happens when you buy stocks and homes at extreme prices.  The reflex they experience in the face of rising prices is more on the side of caution now, rather than excitement.  The situation is further stabilized by the society’s aging demographics.  Older people are more risk-averse than younger people.  It’s harder to get them to play in bubbles.

But let’s go back to 2003-2004 for a moment. What would the correct Fed response have been? Dean Baker says that if he had been the Fed, he would have publically threatened to raise interest rates until homes prices returned to orbit.  But this would have caused a recession–then what?

The truth is that the Fed in 2003-2004 was in a genuine bind. The choice was depression or bubble. Pick your poison. The 2003-2004 Fed chose a bubble; the 1930 Fed chose a depression.  The only reason that the 2013 Fed can escape from both is that it’s been bailed out by weak demographics and the scars of ugly memories. It probably couldn’t create a bubble in stocks or homes if it tried.

The Fed was right to choose a bubble over a depression (or, more generously, to choose the risk of one over the risk of the other).  A depression surely would have caused more suffering. But this is not to say that bubbles don’t matter.  They matter a ton.  In the U.S., they’ve had a significant negative effect on the well-being of millions of low and middle-income citizens.  If there are stimulatory policies that can avoid provoking them, those should be a goal.

What are the costs of a bubble?  The costs are the damaging effects of the unwind–on balance sheets, on the banking system, and on confidence, the foundation of economic growth.  If a housing bubble emerges in response to overly easy money, then when inflation picks up in the cycle, and tighter money becomes necessary, a lot of people are going to lose a lot of money–not money they own, but money they’ve borrowed, money that ultimately belongs to other people. The negative effects on their behavior and on the behaviors of those who lend to them will spread throughout the system, culminating in a recession.  If the Fed can prevent, or at least avoid contributing to, such an outcome, shouldn’t it want to? Shouldn’t it want to help foster an economic environment in which asset prices, like consumer prices, are stable and predictable over time, rather than violently ballooning and then crashing?

Now, it would be a terrible mistake to try to pursue asset price stability at the cost of perpetually-induced recessions. And so using the blunt instrument of monetary policy alone to fight the excesses of financial markets, without consideration for the negative effect on the economy, is obviously the wrong approach. But what if we gave the Fed more tools? Specific to the purpose of this essay, what if we put the Fed in control of a certain amount of fiscal policy, so that it could prudently mix the two types of stimuli in ways that create the optimal economic outcome–a healthily growing economy, without an asset bubble?

The current recession–especially the last 12 months of it–has taught us an important lesson. Fiscal policy matters–a lot. Here we have a Fed that is engaged in arguably the most aggressive balance sheet expansion in the history of the post-war developed world. But even a relatively small amount of fiscal tightening has been enough to entirely neutralize the economic boost–and leave us with nominal growth rates that are the lowest they’ve been at any time in the last 50 years, outside of recession.

The reason that we leave monetary policy to a group of “enlightened dictators” is precisely because it’s too economically important to be subjected to the chaos and corruption of the political process. Noisy politicians who do not understand it are not in position to intelligently vote on it. And if they were allowed to vote on it, they would vote purely on the basis of their own competitive electoral interests, rather than the interests of the general public. It would devolve into a mess of lies and brinksmanship–similar to what congressional discussions on fiscal policy have presently devolved into.

It’s quite ironic then, that we would be comfortable with an approach that relegates fiscal policy–the most powerful type of economic policy there is, at least in the current environment–to the whims of uninformed democracy.  Like monetary policy, fiscal policy belongs in the hands of those who understand it and who can be trusted to employ it prudently for the general benefit of all, not in the hands of those who know nothing about it and who are bound to try to exploit it for personal political gain.

The solution I propose, then, is this. We formally desegregate retirement programs from their trust funds, and fund them instead from general revenues. We then give the Fed full control over the rate on the payroll tax. We allow the Fed to set that rate anywhere from 0% to 10% (or higher). Because the payroll tax is a tax that primarily hits low and middle income paychecks, it has a significant effect on aggregate demand, and can therefore be used to both stimulate employment and fight inflation–the two responsibilities that we entrust to the Fed.

This proposal is filled with hidden benefits:

First, the proposal gives the Fed the ability to safely stave off asset bubbles in early-stage economic recoveries. If the Fed finds itself in a 2003-2004 type situation where the economy still needs stimulus, but where bubble risk is growing quickly, it can proceed with the Dean Baker approach–raise rates punitively to send a message to speculators–while lowering the payroll tax to bolster aggregate demand. The fiscal offset would concentrate the destimulatory effect of the rate increases precisely where the Fed wants it to hit–in speculative asset markets–and not everywhere else.

Second, the proposal allows the Fed to effectively deal with the opposite scenario–a stagflation similar to the late 1970s, where it needs to put downward pressure on incomes without disincentivizing needed capacity-expansive investment.  In such a scenario, the Fed could maintain lower interest rates to avoid excessively discouraging corporate investment in new capacity, while maintaining a higher payroll tax to reduce disposable income and keep inflation in check.

Third, the proposal makes it possible for the Fed to move quickly and apolitically on the fiscal front in response to recessions and crises. Under the current paradigm, a significant part of the policy equation–fiscal policy–is entangled in the congressional morass, and can therefore only move at a snails pace. In late 2008, when the economy needed every bit of fiscal stimulus it could get, it had to wait six long months, as congress dithered.  On the monetary side, in contrast, the Fed was able to cut interest rates and expand its balance sheet immediately after Lehman.  Had it been in control of the payroll tax, it could have done the same on the fiscal side–cut the tax to zero, putting money directly into the pockets of income-starved households.

Fourth, the proposal would dramatically increase the Fed’s policy credibility, particularly at the zero lower bound. Right now, in a zero interest rate environment, it’s very difficult for the Fed to credibly stimulate investment. Sure, the Fed can buy treasuries and mortgages, and make stock and bond markets rise on the excitement, but this is very different from causing corporate managers to invest.  To invest, they need to believe that they are about to see more customers with more money coming through their doors.  The experience of the last 4 years confirms that simply “swapping” assets with the private sector–treasuries and mortgages for cash–does very little to stimulate that belief.

But what if the Fed had the ability to boost everyone’s income 10% in one fell swoop? Then, the game would change. Empty jawboning would turn into talk that you better listen to. The Fed would be able to target a nominal income with significant reflexive suasion, because it would have direct, unmediated control over nominal income (unlike now, where all it can really control is the rate paid by low-risk borrowers, if they want to borrow–a huge if).

Fifth, the proposal would allow the Fed to influence the private-public mix of debt so as to optimize the long-term stability of growth. All growth involves the expansion of debt in some form. Private sector debt tends to be unstable–each individual must carry it on his own back, and stomach its associated psychological pressures, pressures that, when they get heavy, tend to cause everyone to want to deleverage at the same time, to catastrophic effect.  Public sector debt, in contrast, is stable–it is collectively borne, something that no individual person ever has to lose sleep over, or carry on her own back.  For this reason, it doesn’t tend to provoke the dangerous waves of deleveraging that private debt tends to provoke (though it sometimes provokes lawmaker stupidity).  In the worst case scenario, the society as a whole defaults on it through inflation–or, in the case of debt denominated in a foreign currency, through explicit default.  Importantly, in explicit default, no individual citizen’s life is ever destroyed, as would be the case of a private default.    

The payroll tax cut side of the Fed’s stimulus option would be a way of fueling growth through increased government debt. The interest rate side of the Fed’s stimulus would be a way of fueling growth through increased private debt. The Fed could alter the mix of the two types so as to maximize long-run economic stability. It could also tweek policy to address any shortages that emerge in the supply of “safe assets” used by financial intermediaries as collateral for funding, a shortage of which can create squeezes in entangled financial markets where each lender is rehypothecating the collateral it receives.

But what of the national debt?  It’s certainly true that the policy would give the Fed a certain amount of indirect control over the national debt.  But that’s a good thing, not a bad thing.  The people that are best equipped to focus on long-term issues of debt sustainability are leaders at the Fed who manage and understand the economic system, not uninformed politicians and voters. Conveniently, because those people are appointed to their positions, they are also the most likely to be capable of stepping back from the political process and doing the right thing.

An important question that few bother to ask: why is high government debt a problem? For one reason and one reason alone: because it constrains monetary policy.  When the stock of government debt gets really large, the Fed is put in a situation where it can’t raise short-term interest rates without destabilizing the economy. If the Fed can’t raise short-term interest rates, then it can’t control inflation in accordance with its mandate.

To make the point clear, suppose that a country has a debt to GDP of a 1000%. Inflation, which had otherwise been tame for demographic reasons, starts to pick up, as the age distribution of the population shifts.  The central bank is supposed to be able to respond by raising short-term interest rates. Suppose that it needs to raise the short rate from 0% to 5% to keep inflation subdued. If the debt is financed short-term (as most government debt tends to be) the rate increase will increase the government’s interest expense from 0% of GDP to 50%. Absent a dramatic change in fiscal policy, the government will have to borrow 50% of GDP.  But borrowing 50% of GDP and using it to pay interest (lender’s income) to a stock of money and debt that is three times the size of the economy is going to be wildly stimulative–at a time when inflation is already a problem.  Thus, the 5% short rate that the Fed needs to control inflation is going to turn into 25%, then 50%, then 100%, and so on–with each hike, a ballooning deficit, more borrowing, more interest paid, more interest income to households, and therefore more inflation.

Absent fiscal austerity, the only way such a spiral can resolve itself is if rates are held low, despite the rising inflation.  Over time, the real value of the debt will then burn off, which will allow for a more stable equilibrium to be reached.  But this is unfair to the savers who end up holding the bag–enduring large inflation in an environment where their money earns no interest.  The beauty of putting the payroll tax under Fed control is that it gives the Fed the ability to protect their interests by preventing the sovereign from choosing this option.  The Fed would be able to impose the needed medicine –fiscal austerity, the only option that can tame the inflation without ballooning the sovereign’s interest expense.

Some might object that the proposal would be unconstitutional, because, constitutionally, only congress has the power to levy an income tax.  But congress is free to pass laws that delegate that power to other branches of government–in this case, it would be delegating the power to the Fed.  This is exactly what it does with with monetary policy–it delegates its constitutional monopoly on creating currency to the Fed.

Let me close by explicitly outlining the proposal.  We fund retirement programs from the general fund, and turn the payroll tax into a fiscal mechanism.  We let the FOMC adjust the rate on the payroll tax in accordance with its primary mandates: to maximize employment, and maintain low inflation.  But we add a secondary mandate, explicitly stated to be less important than the primary mandates: to employ where possible, a fiscal-monetary policy mix that prevents asset prices from reaching levels that are dangerous to the health of the economy.

If we really wanted to empower the Fed, we could give them the ability to adjust the capital gains tax rate on specific types of assets, so as to discourage excesses from growing.  But that might be taking things too far.

The following chart shows nominal total equity returns in local currency terms for select developed market countries from the beginning of January 1970 through the end of August 2013.

The UK is way out ahead of the pack, almost doubling the returns of the US and France, and more than quadrupling and octupling the returns of Germany and Japan.  But most of the UK market’s nominal outperformance has been due to higher domestic inflation.  The following chart shows the consumer price index for each economy:

The following table separates annualized returns for each country into nominal, inflation, and real:

Note that the sum of the annualized real returns and the annualized inflation rates are almost equal to, but not exactly equal to, the annualized nominal returns.  This is just an artifact of the math.

The following chart shows the real total return for equities in each nation:

We can draw two conclusions from these charts.  First, future nominal US and European equity returns are likely to be low relative to the past, and second, German equities are probably cheap relative to those of the US.

Future Nominal Returns Are Likely to Be Historically Low

As far as equity performance is concerned, inflation matters–a lot.  Over the last 50 years, it has accounted for a little less than half of the total nominal returns of developed market equities.

Looking out into the future, developed economies face a number of anti-inflationary headwinds:

• Aging and shrinking population demographics, a secular trend that has emerged as a result of economic and medical progress and a general elevation of women’s status in society.
• Stagnation in opportunities for genuine, economic-value-adding innovation, and therefore in opportunities for real investment, which stimulate inflation by putting pressure on the supply and price of labor.
• Reduced bargaining power on the part of labor due to labor-displacing technological advancement and globalization.

Consequently, we can conclude that nominal total returns for developed market equities are likely to be meaningfully lower than they’ve been in the past.

German Equties are Probably Cheaper than US Equities

Justifications aside, the Germany equity market is probably more attractively valued right now than the US equity market.  Let’s explore why this is likely to be true.

The “equity” (or total business value, understood not only in potentially inaccurate accounting terms, but also in real terms) contained in a share of a stock market trades at a certain price, and therefore at a certain valuation.  The total return from any time T1 to any time T2 is the total profit generated by that share of equity (which is either reinvested internally, thus increasing per-share equity, or distributed to shareholders via dividends), plus any return that results from a change in valuation.

Now, ignoring differences in corporate taxes, if we assume that equity markets in two different countries start at time T1 at the same valuation, and proceed to generate the same return on equity from time T1 to time T2, then any difference in their total returns over the time period must be a consequence of changes in their relative valuations.  The equity market that produced the lower total return must have become cheaper than the equity market that produced the higher total return, with its falling valuation being the very cause of its lower total return.

From 1970 to 2013, the total real return of German equities has been roughly 25% less than that of US equities.  If we make the reasonable assumption that German equities were not already significantly more expensive than US equities in 1970, and that the return on equity of the corporate sector in each country has not been significantly different, then it would have to be true that German equities are presently cheaper than US equities.

It’s tempting to draw the same conclusion about Japan, given the dramatic underperformance of its equity market.  I hesitate to draw this conclusion because Japanese corporations have been notorious destroyers of value over the last 25 years. Their relative underperformance can be attributed to relative differences in their return on equity.  The same is not true, however, of German corporations, which tend to be just as well run and shareholder friendly as their US counterparts.

The conclusion that German equities are cheaper than US equities fits with the signals generated by other valuation measures.  According to MSCI, Germany’s trailing twelve month price-earnings multiple is roughly 12, around 30% less than the US multiple of 17. Germany’s Shiller CAPE and price to book value are similarly depressed relative to that of the US.

Why are German Equities So Much Cheaper? 

We can say at the outset that the reason has nothing to do with any opportunity cost relative to risk-free bonds.  Long-term German bund interest rates are significantly lower than long-term US treasury interest rates, more than 100 bps for the 10 year, which reflects the uncontroversial fact that the future paths of short-term central bank rates in each country are likely to be meaningfully different.

An obvious reason for the difference in valuation is the effect of the Eurozone crisis. Given the anti-growth fiscal, banking, and monetary straightjacket imposed by the currency union, as well as the likely possibility of eventual exits, risk appetite is significantly subdued in Germany, and fear of tail events significantly elevated, relative to the US.

But there’s another interesting difference: demographics. The point isn’t often discussed, but Germany’s population, and Europe’s in general, has an age distribution that is older than that of the US.  Aging populations are bearish for equity valuations because older people tend to allocate their wealth with a preference for low-volatility, low-risk assets (such as savings accounts and short-term bonds) over high-volatility, high-risk assets (such as equities).

The following charts show the 2015 population pyramids for the USA, Germany, and Japan:

 As we see from the charts, Germany’s demographics, both in terms of population growth and age distribution, look much more like Japan than the US.  Of course, equities are traded internationally, and so non-German investors could relieve downward demographic pressure on German markets by buying up “cheap” German equities.  But investors tend to display home bias when they invest.  Absent a compelling reason to shop elsewhere, they tend to want to stay within their borders.  Therefore, we should expect differences in country age distributions to maintain an effect on country valuations, even though there is international access to capital markets.   

I think that the differences in the economic, financial and business conditions in Germany and the US–and, crucially, their carryover effects on investor sentiment and appetite for risk–have had a significantly more powerful effect on valuations than simple demographic differences, the impacts of which, in my opinion, are exaggerated.  Granted, it is true that if given a choice as to how to meet their income needs, older investors will lean towards cash and short-term bonds, which have minimal credit and interest-rate risk, rather than the more volatile options of equities and long-term bonds.  But older investors in the US and Europe aren’t currently being given such a choice, and won’t be given one for a very long time.  If they want any income at all in the present environment, they have to take on either market risk, credit risk, or interest rate risk (note that in 2013, this latter risk has been the most costly of the three, and could continue to be the most costly over the next few years).  

In the current world of zero interest rates, older investors have proven that they are willing to own equities, provided that they feel comfortable with the health of the economy and the market, and do not see painful losses on the horizon.  Ultimately, fears of losses are what keep all of us away from equities, when we stay away.  If the possibility of losses, especially deep losses–crashes–could be entirely removed, we would all invest 100% in equities, since, when purchased at normal valuations, they offer much stronger long-term returns than their alternatives. 

The difference between older investors and middle-aged investors is that older investors are more sensitive to the risk of near-term loss, given their liquidity needs and time horizons (getting your money back in the eventual recovery is worthless if you won’t live long enough to see it).  But this is not to say that middle-aged investors are willing to jump into equity markets that they think are dangerous, or that older investors can’t become perfectly comfortable investing in equity markets that they think are healthy and strong.  With any investor, of any age, the key question that determines the investor’s willingness to take on equity exposure is: how do you think equities will perform going forward?  Every other consideration is minor compared to that one.

In the course of the cycle, as economies and markets recover from whatever their last negative event was, and as the memories of losses associated with that event abate, older investors, like all investors, become more willing to take on equity exposure.  As they accumulate positive experiences in the markets, they grow more comfortable, more confident that their wealth is being invested in the right place.  There is no reason why they can’t end up in a situation where they are willing–indeed eager–to be invested in equities.  

We have to remember that the current generation of wealthy retirees in the US and Europe (who hold the vast majority of each regions’ investment firepower) made their fortunes over the last 30 years precisely by being invested in the stock market.  If you take out the shock of 2008, which is now more than recovered, they are likely to feel a certain amount of goodwill towards the asset class–at least enough to give it a chance, when the only alternative is to sit on the sidelines without any retirement income and watch everyone else make money.

Earningless bull markets are bull markets in which stock prices rise substantially despite falling earnings.  Consider two examples from U.S. market history (price and EPS are for the S&P 500):

Both of these earningless bull markets started from low valuations.  We can therefore attribute a significant portion of their moves to mean reversion.  For whatever reasons, sentiment towards equities improved, and better valuations followed, despite falling earnings and rising interest rates.

But why did the moves go as far as they did?  The P/E ratio in August 1961 was 22.  The P/E ratio in June 1987 was 21.  The 1987 advance was particularly puzzling, given how high interest rates were.  Investors were holding equity at an earnings yield (inverse P/E) of 4.75% when the 10 year treasury offered almost twice that yield, risk-free.  What were they thinking?

It’s often assumed that June 1987 equity buyers were wrong to be buying, that they were victims of irrational exuberance.  But if we neglect the unlucky crash they walked into, a crash that no one could have foreseen or predicted, they turned out to have been right. The 10 yr total return they ended up earning in equities was a whopping 14.7% annualized, whereas those who bought and held the 10 year treasury (zero coupon) only earned 8.4%. The equity risk premium (difference in return between stocks and bonds) ended up being 6.3%, significantly higher than the historical average of 4%.

Now, some might say that 1987 buyers got lucky–that they road the coattails of the multiple expansion of the 1990s.  But actually, on a 10 year time horizon, they received no multiple expansion at all.  The multiple in June 1997 was 21, the same as in June 1987.  The S&P’s stellar total return was driven entirely by the growth of earnings, which tripled during the period.  The success of June 1987 equity buyers shows the danger of equating earnings yield (which is not actually a yield, just an internal ratio) with total return, in neglect of other important variables such as profit margins and the trajectory of earnings (profit margins are inversely correlated to earnings growth, and were at a historical low in 1987).

But even if we try to explain the earningless bull market of the 1980s by pointing to low profit margins and high anticipated earnings growth, we still aren’t telling the whole story. We’re ignoring the role of asset supply.

In the previous post, we laid out the mechanics of asset supply in rigorous detail.  Looking at the three types of financial assets–money (cash), credit (bonds), and equity (stocks)–we found two rules to be fundamental:

(1) For every financial asset that exists, someone must willingly hold it at its current market price.  If no one can be found that wants to hold it at that price, the price will fall until someone is found.

(2) The “supply” of a financial asset is its total market value–the total dollar amount of it in existence at market prices, which is the amount that investors can allocate their wealth into.  Equity assets have a flexible supply, meaning that if demand to allocate wealth to them exceeds supply, the price per share will rise, therefore the total market value will rise, therefore the supply will rise, fixing the problem.

Ultimately, the only sector of the economy that genuinely owns any financial assets are households. There are different ways that households own financial assets–directly on their own, or with the help of a manager, or through mutual funds, or through hedge funds, or through pension funds, etc.  But the final ownership always comes back to them.  They are the people that make up the economy.

When we use the Flow of Funds report (Z.1) to draw a picture of the universe of financial assets–how much cash, how much credit, and how much equity there is in the overall economy—what we are actually looking at is the aggregate household investment portfolio: how households, in aggregate, are allocating their wealth.  From rule (1) above, someone must hold every financial asset in existence, so if there is a large supply of cash and credit floating around, and a small supply of equity, then households (or whoever represents them financially) have to want their portfolios to be allocated in that way–with big positions in cash and credit, and a small position in equity.  If that’s not what they want, if they want a bigger portion of their wealth invested in equity, then the per-share price of equity–and therefore the market value, and therefore the supply–is going to get pushed up under the pressure, until aggregate allocation preferences are achieved.

Every investor has a preferred asset allocation, informed by both structural (age, risk tolerance) and cyclical (outlook, sentiment) factors.  Allocation is the most important decision that investors make when they build and manage portfolios.  Where should I put the money: in stocks, in bonds, in cash?  How much to each?  Which sectors of the stock market should I invest in?  Which sectors of the bond market?

Consider the classic allocation that efficient market gurus recommend: 60% to stocks, 40% to bonds.  What rule says that there has to be enough equity, at “fair” valuations, for everyone to be able to allocate their portfolios in that way?  There is no rule.  If everyone wants a 60/40 allocation, but equities are not 60% of total financial assets, they will quickly become 60%, as they get lifted to higher prices by all the bidding.  There is nothing to guarantee that the prices they end up at are going to be “fair.”

Now, valuation concerns can certainly push back, and motivate investors to accept a lower allocation to equities (lest they have to buy really expensive stocks, expensive because of the short supply).  But, outside of extremes, “valuation” doesn’t tend to be a very strong counter force, especially given that (1) it’s hard to reliably determine–it involves making tenuous predictions about the future, (2) it’s ambiguous–there’s a wide range of what can be considered a “reasonable” valuation, and investors tend to be quite willing to rearticulate that range as needed to rationalize their deeper market biases, (3) it’s self-fulfilling–investors base their assessments of valuation on future estimates that are as much a reflection of a given sentiment towards equities as they are a cause of it, and (4) equity investors invest for price appreciation, rather than yield (which is what valuation actually hits, concretely).  Because price is the bottom line, the ultimate source of return, investors tend to get very uncomfortable sitting out of rising markets on the grounds that the valuation is too high (a call they can’t even be that confident in).  They miss out on real, tangible profit.

These insights can help us explain one of the reasons why earningless bull markets happen. Whenever an economy is growing, money and credit are being created through the lending process. The rising supply of money and credit end up in investor portfolios–they have nowhere else to go. To the extent that investors’ desired allocation to equities remains stable in the presence of the rising money and credit supply, the supply of equity has to rise at a similar pace. Otherwise, its share of the total–which literally is the aggregate household investment allocation to equities–will shrink.

As an aside, it’s worth noting that if you look at the available historical data, which goes back to the early 1950s, you will see that money and credit have both grown at about 8% per year over the period.  The market value of all equities in existence has also grown at about 8% per year.  These numbers are not the same by coincidence.  If the market value of all equities in existence had grown at some smaller clip, say 4%, while money and credit were growing at 8%, the equity share of overall investment portfolios would have collapsed to near nothing.  Investors would have had to have wanted that.  Very unlikely.

Now, there are two ways that equity supply can rise.  The number of shares outstanding can rise (corporations can issue new shares), or the price per share can rise.  Historically, corporations have been very hesitant to increase their number of shares outstanding. With the exception of the 2008 crisis, annual share issuance over the last 60 years has never exceeded 2%.  For most of the last 35 years, corporations have been operating in share consumption mode–reducing their shares outstanding through buybacks and M&A.  The situation now is not as extreme now as it was in prior decades, but net consumption continues to take place.  The chart below shows total equity issuance as a percentage of total outstanding.  Note the extreme level of share count reduction that took place in the 1980s.  This phenomenon is not unrelated to the earningless bull market that the 1980s produced:

As stated earlier, money and credit grow at a rate much faster than 2% per year–over the last 60 years, at an average of 8% per year.  Unlike share count, money and credit rarely, if ever, contract. The following chart gives an approximation of the growth of money (red) and credit (blue) since 1959:

For these reasons, if investors maintain a constant desired allocation to equity, that is, if the percentage of their portfolios that they want to have invested in stocks remains stable, then stock prices have to rise.  There is no other way.  Rising stock prices are the only mechanism through which the supply of equity can keep up with the supply of everything else, so that equity allocations can stay at the level that investors want them to be at.

If, as happened in the 1970s, the stock market doesn’t rise commensurately with the rising supply of credit and money, then equities will steadily become a smaller and smaller percentage of the aggregate pool of total financial assets.  To the extent that the demand for allocation to equities remains stable, a shortage will develop.  This shortage will put upward pressure on prices–regardless of what is happening on the valuation front.

Normally, there isn’t a problem.  Equity prices need to rise to keep equity supply growing at a pace consistent with money and credit growth, but they are already rising for fundamental reasons–specifically, because earnings are growing.  And if corporations choose to buy back shares or conduct M&A, no problem is created, because the buyback leads to a de facto earnings per share increase–therefore prices can rise to keep overall supply constant without affecting valuations.  

But what happens when you have a long period of time during which money and credit grow normally, but earnings fail to grow, or fall?  Unless the aggregate investment community’s preferred allocation to equities as a percentage of the overall portfolio falls, prices are going to have to go up–regardless of what “valuation” says they should be doing.

To better understand what I’m saying here, let’s do an experiment.  Let’s calculate how the U.S. investor allocation to equities would have had to have changed if the economy had gone through ten year “lost decade” periods where credit and money grew at their historical rates, but where stock prices stayed constant.  The purple line shows what the portfolio allocation to equity would have ended up being from each point forward (into the lost equity decade), in comparison with what it actually was (the orange line).   

As we see from the falling purple lines, for market prices to stand still over ten year periods in which credit and money grow normally, investor equity preferences have to fall precipitously.  There is little reason to expect this to happen inside a healthy economic expansion, where investors are optimistic, and want to be invested for the future.   

Notice from the chart that the aggregate allocation to equities rose during every healthy economic expansion (late 1970s inflationary mess excluded), and fell during every recession.  This trend is consistent with what a behavioral view would predict.  The key assessment that determines the eagerness of investors to invest in equities as opposed to other asset classes is the assessment of risk, specifically tail risk–how much might I lose? Will there be a selloff, a crash?  Large losses really hurt; they are the feature of equities that keep investors away.  As investors become more confident that a selloff or crash is not coming, they become more comfortable with equities, and more eager to allocate wealth to them.  Thus, as the economy moves farther into periods of healthy expansion, where social mood is high, and where memories of the traumas and tail events of the last bear market have faded, investors display a rising preference to be invested in equities.  However, for market prices to remain constant over long periods, so as to catch up with earnings and valuation, we need the opposite–we need investor equity preference to fall.  In healthy economic expansions, this just isn’t likely to happen, and therefore earningless bull markets are sometimes unavoidable.      

Let’s do a similar experiment with the two earningless bull market examples that we studied earlier. Let’s assume that investors respected “valuation”, and kept the P/E ratio constant from 1956 to 1961, and from 1979 to 1987.  The market would not have risen, because earnings didn’t rise.  In fact, the market would have have fallen, because earnings fell.  Therefore, the supply of equity would have fallen, even as the supply of money and credit expanded rapidly.  The aggregate allocation to equity would have shrunk–and investors would have had to have accepted this result in their portfolios, accepted a situation where less and less of their wealth is allocated to the stock market.  In a vibrant expansion, where sentiment towards equities is improving, why would that happen?  The point is that it wouldn’t happen–and it didn’t.  

The orange line below shows what the actual investor allocation to equities was from 1951 to 2013.  The dark blue line shows what the allocation would have had to have been reduced to if “valuation” had been respected, and if the market were to have fallen (or stood still) consistent with the falling (or stagnating) earnings.

Alarmingly, from 1979 to 1987, investors would have had to have willingly allowed their allocation to equities to fall from an already generationally low, post-1970s-bear-mkt-allocation of 25%, to an even lower 13% of assets–this, in the middle of a vibrant economic renaissance where consumer and business sentiment, plagued by the inflation, regulatory, and tax nightmares of the 1970s, were finally improving (or at least that was the narrative), where investors were gaining confidence that policymakers were on their side, and where the economy was reaching a demographic sweet spot in which a large number of baby boomers were entering the 30-something age range where equity allocation is naturally favored.  It just wasn’t going to happen.  That’s part of the reason why the market was pushed up to such a high P/E multiple, despite a sky-high interest rate.  The market was not wrong in making that move–it was following the forces that were driving it.

Now, you might ask, shouldn’t rational investors have been content with such a low equity allocation, given that bonds and cash were offering such an attractive yield relative to equity?  But that isn’t how markets work.  We’ve seen time and again, in the U.S. and abroad, that though valuation unquestionably drives long-term returns, it is easily overshadowed by sentiment and demographics in determining how investors choose to allocate in the here and now.     

Much of the time, as in 1987, investors don’t even know what the market’s proper valuation is, because they don’t know how future earnings are going to evolve.  To determine value, they have to rely on forward estimates, which means they have to rely on their own mood-influenced visions of the future, rather than concrete, simple-to-measure facts.  The 1987 market looked really expensive from the vantage point of earnings and interest rates, but in hindsight it turned out to have been very attractively priced.  Investors that irrationally chased it before the crash were doing exactly the right thing.  Much of their being right probably had to do with luck, but they were right nonetheless, and were compensated handsomely for it in the final analysis.

This piece looks at issues related to the “Great Rotation”, the view that U.S. investors are in the early stages of a reallocation of their wealth out of bonds and into stocks.  The interesting charts are at the end, so if you’re pressed for time, just read the summary, and fast forward. 

Summary  

• The relevant universe of financial assets is composed of equity (stocks), credit (bonds), and money (cash).  The extent to which individual sectors of the economy (Households, Pension Funds, Life Insurers, etc.) and the economy in aggregate have allocated wealth to these categories can be determined from the Flow of Funds report.   

• For every asset in existence, someone must willingly hold it.  Prices will adjust to ensure a holder is found.

• The “supply” of an asset is the total market value of it in existence.  Changing market prices are the means through which asset supply expands or contracts to meet demand.

• “Asset Shortages” are actually income or yield shortages.

• “The Great Rotation” entails two portfolio processes: (1) a reduction in the preferred allocation to bonds, relative to cash and stocks, and (2) an increase in the preferred allocation to stocks, relative to cash and bonds.  These processes do not represent a “re-allocation” of wealth from bonds into stocks.  They instead represent a separate destruction of wealth in bonds, and a separate creation of wealth in stocks.

• Contrary to market narrative, the aggregate U.S. investor is not “underweight” U.S. equities, but is instead “overweight” (but not extremely so), relative to the average of the last 60 years. Sectors notably overweight: Pension Funds, Life Insurers, and Households.  Sectors notably underweight: Property-Casualty Insurers and Foreigners.

• If a “Great Rotation” is coming, it may be a rotation into cash (an increase in the amount of a portfolio that the average investor wishes to allocate to cash).  Of all asset classes, cash is clearly the most hated–and with some justification, given the zero interest environment.  If the rate environment normalizes (and there is nothing to say that it must), cash is likely to be the prime beneficiary, at least relative to bonds.

The Universe of Financial Assets

We can divide the universe of financial assets into three categories: (1) money, (2) credit, and (3) equity.  Money (or “cash”) is whatever the government designates as tender for the payment of taxes and fines.  Under this definition, dollar currency and dollars deposited at a bank are money.  Treasury bills, money market mutual fund shares, and open market paper are not.  Credit is a claim to be paid money by some entity.  Equity is ownership of an entity, including its monetary income.  The Flow of Funds report (Z.1) provides detailed information on the money, credit and equity assets of each sector of the economy.

Let’s take a closer look at credit assets.  The following chart uses Flow of Funds data to decompose the universe of dollar-denominated credit assets by type and share from 1951-2013:

The reader will note that there is a significant amount of double-counting in this decomposition.  For example, we see the category “agency securities”, and we see the category “mortgages.”  Both of these credit assets are associated with the same fundamental act of lending.

If a government-sponsored enterprise (GSE) sells a debt security, and uses the proceeds to purchase a mortgage loan from a bank, the GSE will own a credit asset–the mortgage loan.  But the person who purchased the debt security from the GSE will also own a credit asset–the debt security.  In this way, the same fundamental act of lending gets multiplied into two securities: one covering the obligation of the homeowner to pay money to the GSE, and the other covering the obligation of the GSE to pay money to the investor holding GSE debt.  This process is called “financial intermediation”; it is what financial institutions do.

In truth, anyone can engage in financial intermediation–the task is not limited to financial institutions. Anytime a person purchases an asset with money that she does not own, and that she has to get from somewhere else by borrowing, she is engaging in financial intermediaton.

When financial intermediators purchase and hold assets, they do not change the quantity of assets that the rest of the economy must hold.  They only change the mix.  Some examples:

• Bank of America ($BAC) buys a treasury bond.  If it funds the purchase by creating and selling its own debt security, then it takes the treasury bond out of the market, and puts it’s own debt security in (a $BAC bond that didn’t previously exist).  If it funds the purchase by creating money (in the various ways that banks create money), then it takes the treasury bond out of the market, and puts cash in (which didn’t previously exist).  Notice that this activity is very similar to QE, which is an activity that any depository institution can engage in–not just the Federal Reserve. If it funds the purchase by selling equity, then it takes the treasury bond out of the market, and puts an equity security in (a share of $BAC which didn’t previously exist).

• A Hedge Fund manager borrows a share of stock from person A, and sells it to person B–he goes “short.”  Through the sale, the Hedge Fund manager is taking cash out of the market, and putting a new share of stock in.  It’s a new share of stock because the person who originally lent it to the hedge fund manager still owns his shares, but the person who bought it from the hedge fund manager, who took the other side of the short, also owns shares.  The supply of shares has been expanded through the stock lending process, just as the supply of money gets expanded through the cash lending process.

The following chart shows the share of credit assets held by financial institutions (blue) and the share held by the rest of the economy (green) from 1951-2013 (Z.1 Flow of Funds):

Here, financial institutions include not only the traditional commercial banking system, but also the entire shadow banking system: money market mutual funds, REITs, finance companies, funding corporations, etc.

Neglecting the role of equity in the financial sector (which is small relative to assets) we can conclude from the chart that somewhere around 1/3 of all credit assets cover genuine acts of lending from owners of money who want to lend it to users of money who want to borrow it.  The other 2/3 cover the borrowing and lending of intermediaries whose job it is to get the money across.  When we try to determine how much wealth each sector of the economy, and the economy as a whole, has allocated to the three asset classes, credit assets associated with intermediation can be ignored, since the actual owners of the economy’s wealth–the non-financial sector–do not have to hold them.

A Model For Thinking About The Great Rotation

Financial assets follow two fundamental rules:

(1) For every share of every asset in existence, someone must willingly hold that share at all times.  If no one can be found who wants to hold a share, its market price will fall until someone is found.

(2) The total “amount” or “supply” of a financial asset is the total market value of it in existence: the number of shares outstanding times the market price.  Asset “amount” or “supply” is therefore flexible for all assets except cash, whose market price is always unity.  If there is more financial wealth that wants to be allocated into an asset than exists of that asset, the market price of each share of the asset will rise, which will expand the supply of the asset so that the demand can be satisfied.

From the perspective of someone seeking to buy, the “market price” is defined as the lowest “ask” price being offered.  From the perspective of someone seeking to sell, the “market price” is defined as the highest bid price being offered.  In a market with free exchange, this price adjusts to ensure that both rules are satisfied.

Investors tend to describe the Great Rotation as a movement of money and people: money “comes out of” bonds, and “goes into” stocks… investors “leave” the bond market and “enter” the stock market. We sort-of know what investors are trying to say when they say this, but the description is wrong.

Money is not something that can go into or come out of assets; rather, it itself is an asset that is traded for other assets.  The offered rate of exchange is the price.  Changes in the price can create the perception that money is moving, but, in reality, nothing needs to be moving at all.  Any movement that does occur is incidental to the underlying process.

Likewise, investors cannot leave or enter any asset class.  All they can do is fight with each other over who will hold each asset class, offering to exchange money at various rates in exchange for the privilege of holding something else.  The consequence of shifting preferences and exchange rates may be a destruction or creation of wealth in various places, but it is never a “movement” of wealth.

An intuitive example will help clarify.  Suppose that the financial universe is composed of 6 persons, each with a portfolio of assets.  Each person starts out with 10 shares of stock, with each share worth $1 at market prices (therefore, $10 worth of stock in each portfolio, $60 aggregate supply of “stock” in the total universe), 10 bond certificates, each worth $1 at market prices (therefore, $10 worth of credit in each portfolio, $60 aggregate supply of “bond” in the total universe), and 10 dollar bills ($60 worth of “cash” in the total universe).  The allocation is therefore 33.3% to stocks, 33.3% to bonds, and 33.3% to cash.

Each investor is value-insensitive, and makes buy and sell decisions strictly based on what is necessary to achieve desired allocations.  Now, consistent with a “Great Rotation”, suppose that each investors’ desired allocation shifts.  Each investor decides that he wants to allocate a larger portion of his assets to stocks, and a smaller portion of his assets to bonds.  He wants to keep his cash allocation constant.  We’ll assume he wants to allocate as follows: 33.3% to cash, 6.6% to bonds (credit), and 60% to stocks (equity).

Naturally, investors are going to take their cash and try to use it to buy shares of stock. To avoid lowering their cash allocations, which they want to keep constant, they’re going to try to sell bonds in the equivalent amount.  But there’s a critical problem.  When they try to buy and sell shares, they are trying to buy and sell shares from and to each other, through the exchange.  Right now, there isn’t enough stock on sale to meet the desires of all the buyers–indeed, there isn’t any, because everyone is trying to buy.  Likewise, there are too many bonds on sale for the amount of interested buyers–of which there are none.  What will happen?

The answer is simple.  Since the investors are value-insensitive, they each are going to increase their bids on stocks, and decrease their asks on bonds, until they get what they want–a trade.  Thus the market price (highest bid) on shares of stock will rise, and the market price (lowest ask) on bonds will fall.  As the market price of each asset rises and falls (respectively), the supply is going to increase and decrease (respectively), making it possible for each investor to meet his desired allocation, and for the aggregate group to meet its desired allocation.

The shifting bidding and asking process will stop when prices have shifted such that the bid on stocks is $1.8 and the ask on bonds is 20 cents.  Each investor will then have 6.6% of his portfolio in bonds at market prices, 60% in stocks at market prices, and 33.3% in cash, as desired.  If the prices shift any farther than that, the investors will need to go in the other direction to preserve the desired allocation–they will need to sell stocks and buy bonds. This is what will create an ask for stocks and a bid for bonds above and below $1.8 and 20 cents respectively.

Notice that all that is theoretically required to change the allocations is a bid or ask–no actual exchange of shares needs to take place.  The exchanges are incidental to the process. The “paper” wealth that is created and destroyed in markets is not created or destroyed through physical transfers, but through simple bids and asks–alterations in individual preference, nothing more. 

The moment someone bids $1.8 for each share of stock, the total supply of stock–the total market cap–will have magically increased 1.8-fold.  Assuming no trades, each person will magically have $18 worth of stock in his portfolio, at prevailing market prices.  Likewise, the moment someone bids 20 cents for each bond certificate, the total supply of bonds will have magically decreased by 80%.  Assuming no trades, each person will have $2 worth of credit in his portfolio, at prevailing market prices.  The final universe will look like this: 

Voila! 33.3% cash, 60% equity, 6.6% credit–and not a single trade was even necessary. Now, trades probably will take place, and there is no guarantee that we will ever reach an equilibrium.  The fight over who will  hold what asset may never end–especially considering that they are not likely to be value-insensitive, and that the shifting prices will have reflexive effects on their perceptions and preferences.  In the real world, markets never reach equilibrium.  But if they were to reach an equilibrium in this case, we know for certain that the price of each stock share would be $1.8, and the price of each bond share would be 20 cents.

What happened in this “Great Rotation”?  Did wealth move from bonds into stocks? No. Two distinct and separate processes took place.  Wealth was destroyed in the bond market due to a falling allocation preference to bonds, and created in the stock market due to a rising allocation preference to stocks.  The stock prices could have gone up without the bond prices falling, and the bond prices could have gone down without the stock prices rising.

Asset Shortages

The above example illustrates that there can never be a true shortage of a non-cash investment asset.  If the demand for the asset exceeds the supply, the market price of the asset will rise, magically creating new supply, and vice-versa.  What there can be, however, is a shortage of yield.  The supply of equity and credit at market value might increase in response to increased demand, but this won’t increase the quantity of dividends, coupon payments, and interest payments that investors are really after, or should be after.

Ultimately, a shortage of yield is the mirror image of an excess of cash, a condition in which there is more cash in the economy than any investor wants to hold idly in a portfolio at a zero interest rate (because the investor needs income, or because the banking system is suspect and the investor wishes to avoid deposit risk), but also more cash than anyone wants to put to use in new productive investment, which would create new income that could be used to sustainably pay interest to lenders.  The frustrated investors who are stuck holding the cash therefore try to exchange it for existing yielding assets, which pushes up prices, sometimes to unreasonable levels–but again, this doesn’t solve the problem, because no new yield, which is what investors are really facing a shortage of, is created.

Some have worried that an increasing propensity on the part of corporations to use their earnings to buy back shares could create an equity shortage in the market.  This fear confuses share count with equity supply.  The supply of equity is the total market value of shares outstanding, not the total number of shares.  If a corporation buys back its shares, the market value outstanding need not fall.  Indeed, to the extent that the market is functioning efficiently, the price of each share will rise to reflect the rising earnings per share, which will keep the market value exactly constant, though each shareholder that holds on as others are bought out will end up with a greater share of that market value.

The chief risk of an asset shortage is that it cause assets to trade at unsustainable valuations.  This is a risk in the credit market because credit has a price above which it cannot coherently trade–the price corresponding to a 0% yield to maturity (YTM).  Absent negative deposit rates or perceived credit risk in deposits, there is no reason for credit to ever trade at a higher price than 0% YTM.  For this reason, the supply of credit does have a practical bound.

Asset shortages can also emerge in the equity market if the market value of equity, properly valued based on earnings or book, does not grow at the same pace as the supply of credit and money.  If investors choose to maintain a constant allocation to equity, credit, and cash respectively in the presence of such a dynamic, the price of equity shares will necessarily have to rise beyond what fundamental valuation can justify.  This very interesting phenomenon has happened at various times in market history (though it is not happening at present).  It will be thoroughly investigated in the next piece.

Using the Flow of Funds to Determine The Aggregate U.S. Financial Portfolio

Asset allocation is the single most important decision that portfolio managers have to make.  How much do I allocate to equity, how much to credit, and how much to cash? Where in the equity space should I focus? Where in the credit space?  Even short-term traders have to make this decision–indeed, they have to make it on a daily basis, when they decide what market trends to ride.

Because asset allocation is so important, there is significant value in knowing how the market is allocated at a given time.  An extreme allocation can be a sign of a coming reversion, which will affect market prices.  Fortunately, information about allocation–of individual sectors and of the economy as a whole–can be extracted from the Flow of Funds report.  The challenge, of course, is to avoid double counting in the credit space.  Two-thirds of all credit assets in the economy are held by the financial sector.  But the financial sector doesn’t actually “own” the underlying loans from which these intermediary assets spawn out–it simply “transports” the loans from root lender to root borrower.

Person A might borrow from Person E.  But with intermediation, what actually happens is: Person A borrows from Financial Entity B which borrows from Financial Entity C which borrows from Financial Entity D which borrows from Person E.  The total quantity of loans is just one–the loan from Person A to Person E.  But with intermediation, we have four credit assets that spawn out–the loan from E to D, D to C, C to B, and B to A. With respect to asset allocation, we’re concerned with what the owners themselves are holding, how their wealth is being allocated, so we need to ignore this intermediation.

The most reliable way to eliminate double counting in assessing the equity, credit, and cash allocation of the overall economy is to look only at two variables: total quantity (market value) of equity, and total quantity of non-financial liabilities, that is, liabilities of Households, Non-Financial Corporations, State and Local Governments, the Federal Government and the Rest of the World.  Ultimately, total non-financial liabilities are what money and credit are made out of.  When a non-financial liability is held by a regular entity–such as you or I–it takes the form of a credit asset–a bill or bond.  But when it is held by a depository institution–a private bank, or the Federal Reserve banks–it takes the form of actual money.

We know that when the Federal Reserve buys treasury bonds, it turns a bond into money.  This action is not limited to the Federal Reserve.  A private bank can do it too.  When a private bank borrows an idle deposit from another bank, or from a depositor, or from the Fed (which creates deposits from thin air), and uses the money to purchase a bond, it puts new money into the economy that did not previously exist, and that investors have to hold, and takes the bond out.  This is hardly different from the “quantitative easing” that the Federal Reserve conducts when it does the same thing.

The following chart shows non-financial credit liabilities (the seeds of money and credit) and equity assets as a % of total financial assets in the Overall U.S. Economy from 4Q 1951 to 1Q 2013, with money and credit in the negative (to delineate on the chart), and recessions shaded in gray.  Note that the total quantity of non-intermediated financial assets in the U.S. economy is around $60T:

The first thing to notice is that contrary to the market’s current narrative, investors are not underallocated to equities.  At 39.23% (extrapolated to an S&P level of 1650, August 2013), the allocation is higher than the historical average of 34.63%, though not by an extreme amount.

To double check our work, we can calculate the same value differently, by taking the financial universe to be composed of M2 (a valid proxy for total cash in the economy), equities, and credit assets not held by non-mutual-fund financial institutions.  The following chart reveals a similar allocation to the one above, but splits money and credit apart:

As the chart shows, the asset class to which investors are the most underallocated is cash.  Per the chart, cash represents 18% of total assets, versus a 54 year average of 29%.  Bonds, in contrast, represent 38% of total assets, versus a 54-year average of 30%.  If a great rotation is coming, it could well be a rotation out of bonds and into cash (a shift in the relative preference for each, provoked by rising interest rates).

The following chart shows equities, credit, and cash as a % of Total Household Financial Assets, with mutual fund holdings included, but not pension reserves (all subsequent allocation charts will include mutual fund holdings):

Note that total household financial assets amount to approximately $35T, not including pension fund assets.  Again, the chart makes it clear that households are not underallocated to equities. If anything, they are overallocated to credit and equity, and underallocated to cash. As the economy improves and interest rates rise, the “Great Rotation” everyone is anticipating may well end up being a rotation back into cash–unquestionably the most hated of all asset classes.

The following chart shows the composition of Total Pension Fund Financial Assets (~$6T in total).  Pension Funds are clearly overallocated to equities, and have been trending towards a higher allocation since the 1950s.

The following chart shows the composition of the total financial assets (around $15T) of the Rest of the World.  The Rest of the World is underallocated to U.S. equities relative to cash and bonds: 

The following chart shows the composition of total financial assets of Life Insurers (~$5T). Life Insurers have historically been trending towards a higher equity allocation, which is near record highs: 

The following chart shows the composition of total financial assets of Property-Casualty Insurers (~$1.2T). Property-Casualty Insurers are underallocated to equities relative to the past:

In comparison with the past, the overall market is clearly tilted towards a higher equity allocation.  What this ultimately reflects is that equities have a higher market value relative to other assets in existence than they used to have–there is a greater supply of them to be held than in the past.  As we will show in subsequent pieces, the relative supply of equities (or, alternatively, the aggregate allocation to them) is tightly anti-correlated (~95%) with future returns.  

The fact that the supply of equity is elevated right now suggests that equity investors face the prospect of lower future returns than they have received in the past (something that no one should find surprising, given where profit margins are).  But the likely returns are not unacceptably low–and still significantly more attractive than what any other asset class can offer.  To give a brief preview of what will be discussed in a later post, using the current allocation to equity, I estimate around 5%-7% per year total return (including dividends) over the next 10 years, based on an August 2013 SPX price of 1650: 

• The Performance of the Fed Model
• A Charitable Interpretation
• The Fed Model’s Mistake
• Legitimate Ways to Use the Equity Risk Premium
• Buying the Nikkei at 9,600 in September 2010
• Buying the S&P at 1400 in December 2012

In the last post, we examined how liquidity–the ability to trade an asset, rather than hold it to maturity–radically changes the dynamics of investment.  When there is no liquidity, the return of an investment can only come from one place: the underlying cash flow.  That cash flow is all that matters.  There is no rising market price to use as confirmation of a successful investment, “Great job buying that dip. You nailed it!”, just as there is no falling market price to fear and fret over, “Geez, how low do you think it will go?”  There is just the investment.  The investor has already psychologically parted with his money, said his goodbyes, and is now patiently collecting the income that the investment is producing and delivering, with a focus on the very long-term.

When liquidity is introduced, the dynamics change completely.  The investor no longer approaches the investment as a genuine parting with his money–a long-term goodbye.  He views his money as still there, at a fingertip’s reach, contained within the market price.  For this reason, he takes the market price extremely seriously.  When he buys a stock that plunges in price, he conceptualizes the plunge as a real loss–and feels real regret and frustration.  When he buys a stock that rises in market price, he conceptualizes the rise as a a real profit–and rejoices internally.  In addition to serving as the arbiter of his performance, the market price impacts his anchoring, his assessment of the fundamentals, and his future expectations–these in turn affect the prices that he is willing to pay, which in turn affect future market prices.  The process is recursive and reflexive–with a tendency to exhibit momentum and to generate price equilibria in highly path-dependent manners.

The Performance of the Fed Model

A number of investment approaches wrongly attempt to evaluate liquid investments in the stock market in the same way that an illiquid investment in real business would be evaluated–based strictly on the underlying cash flow.  A classic example is the popular “Fed Model”, coined by analyst Ed Yardeni, and recently touted by Hedge Fund Manager David Tepper and researchers from the New York Fed.  For those that aren’t familiar, the “Fed Model” argues that the attractiveness of the stock market as an investment should be measured by comparing its earnings yield (trailing twelve month earnings divided by price) to the yield on long-term bonds.  On this model, the stock market at 20 times earnings (5% earnings yield) is “expensive” if the 10 year treasury bond is yielding 6%, and “cheap” if the 10 year treasury bond is yielding 1.5%.

What do the terms “cheap” and “expensive” mean, precisely?  Granted, if we define the terms to mean “has a higher yield than long-term bonds” and “has a lower yield than long-term bonds” respectively, then, tautologically, the Fed Model is the arbiter of cheapness and expensiveness.  But if that is all that “cheap” and “expensive” mean, then there is no immediate reason to care about whether the stock market is cheap or expensive.  The terms “cheap” and “expensive” are only worth caring about if they can be linked to future returns, on some time scale.  What we need, then, is a market-based definition.  Is the stock market, given its current price, likely to produce high future returns, or low future returns?  If high future returns, then it is cheap. If low future returns, then it is expensive. With the terms appropriately defined in this way, the Fed Model ceases to be the arbiter of anything.

With respect to future returns, the Fed Model does not appear to have any independent predictive power.  The following chart plots future 2 year annualized total returns (y-axis) versus the “equity risk premium” or ERP (the difference between the market’s earnings yield and the 10 year treasury yield, x-axis) from April 1933, the month that FDR ended the gold standard, to present:

The coefficient of determination is roughly 0.  You could splatter paint on a page and get the same chart.  There are a number of instances where low future equity returns followed high ERPs, and a number of instances where high future equity returns followed low (or negative) ERPs.  Let’s extend the horizon out to 10 years.  

The coefficient of determination increases to 0.10.  Still not attractive.  Granted, when the ERP approaches double-digit extremes (and it is nowhere near such extremes right now), it seems to correctly signal that high future equity returns are coming.  But at those extremes, the earnings yield itself (inverse of the P/E ratio) produces the same signal, without any reference at all to the bond yield.  Here is the simple earnings yield versus subsequent 10 year returns back to April 1933:

The coefficient of determination increases to 0.48, far more attractive.  Demonstrably, then, all that the comparison to bond yields is doing is taking an otherwise meaningful signal–the earnings yield–and distorting it, screwing it up.  The comparison is making markets that were generationally cheap–for example, the market of the early 1980s–look historically expensive (indeed, as expensive as the market of the late 1990s), and markets that were expensive–for example, the market of late 1936 and early 1937–look cheap.  

It makes perfect intuitive sense that the bond yield would represent a distraction in the analysis of long-term future equity returns.  To the extent that long-term future equity returns are driven strictly by the underlying cash flows of the equities themselves, in the approximation of an illiquid investment, we should expect the equities themselves–the cash flows that they produce–to be the drivers of the return, regardless of how those cash flows stack up to the cash flows produced by other investments: bonds, real estate, collectibles, whatever.

The fundamental problem with the Fed Model is this.  It is simply incorrect to assess the cheapness or expensiveness of one asset class by blindly comparing it to another.  Such an approach dismisses the very real possibility that both asset classes are cheap or expensive at the same time–that they will both produce strong or weak future returns.  History clearly demonstrates that such an outcome is possible.  In the spring of 1937, stocks and bonds were both expensive, they both produced unattractive future returns.  In the summer of 1982, stocks and bonds were both cheap, they both produced excellent future returns.

A Charitable Interpretation

But maybe the Fed Model isn’t saying that stocks should be considered cheap or expensive based on how their yields compare with bond yields.  Maybe the model is simply telling us which asset class offers a higher return, and therefore which asset class an investor seeking to maximize return should choose, if she is forced to choose.  In 1937, an investor should choose stocks.  In 1982, an investor should choose bonds.  

But if this is the model’s contribution, then it isn’t of much value.  With the exception of the ends of recessions and bubbles, stock earnings yields are almost always higher than bond yields.  You don’t want a model that will have you out of the market at the end of recessions–those are the most attractive times to invest.  And though the Fed Model can spot bubbles, so can a plain vanilla P/E approach–where the bond yield is thrown out of the calculation.  What value, then, does the model add?   

Surprisingly enough, if an investor reliving the 1933-present period demands any kind of premium at all between earning yields and bond yields as a condition for investing in equities, he will underperform “buy and hold” on an absolute basis.  The following chart illustrates the total return performance of various Fed Model approaches back to April 1933.  Each line shows a Fed Model that uses a different cutoff risk premium (where you own the stock market if the difference between its earnings yield and the 10 year bond exceeds that premium, and you own 10 year treasury bonds otherwise).  The decision whether to switch is made at the end of each month.  To simplify the calculations, and also to capture the full upside of the 1981-present bond bull market, it is assumed that when 10 year bonds are owned, they are rolled over each month into new issues at no tax or transaction cost, with the capital gain or loss pocketed:

To make the points of relative performance more clear, let me introduce a different type of chart.  This chart plots the ratio of the performance of each strategy (numerator) to the performance of buy and hold (denominator), over time.  When a line is rising, the associated strategy is accumulating outperformance relative to buy and hold.  When a line is falling, the associated strategy is accumulating underperformance relative to buy and hold.  When a line is straight, the associated strategy is tracking buy and hold (usually because it is invested in stocks):

All premia except a 0% premium generate a lower return than buy and hold.  The 0% premium generates a higher return by about 17% (total, not annualized).  This excess return is good, but in the real world, would not be enough to make up for transaction costs, switching (slippage) frictions, and tax hits, which we have not modeled.  The strategy spends essentially all of the time before the early 1970s invested in stocks, and then proceeds into bonds at various points during the 1970s inflation, where it generates the bulk of its excess return.  The strategy avoids the downside of the tech bubble, but there is no net gain, because it also avoids the upside.

Note that during the 1970s, cash invested at the Fed Funds rate outperformed both stocks and bonds.  From Jan 1972 to Jan 1982, $1 in cash became $2.44, versus $2.31 for the Fed Model, $1.64 for Buy and Hold, and $1.28 for the rolled-over 10 year.  Surprisingly, then, the only period in which the Fed Model reliably outperforms the return of Buy and Hold is a period in which cash is king–certainly not the kind of period that most analysts have in mind when they tout the model.

The Fed Model’s Mistake

The Fed Model makes an appeal for a certain a type of investment consistency.  If investors are willing to pay a lot for the cash flow that one type of asset–e.g., the aggregate bond market–offers, they should be willing to pay a lot for the cash flow that another type of asset–e.g., the aggregate stock market–offers.

As discussed in the previous post, if we lived in a world without liquidity or trading, where all investments had to be true investments, held to maturity, with all returns generated directly from the underlying cash flows–the earnings, coupons, and interest payments, rather than the appreciation of price–then this appeal might make sense.  If you are willing to pay 33 times earnings for a 30 year treasury bond, why wouldn’t you also be willing to pay 33 times earnings for a well-diversified index of blue-chip stocks?  With the stocks, you would accrue the same initial yield, plus decades of eventual growth in yield.  The only real uncertainty, over the 30 year period, would be the upside: how much would the earnings grow?  Over 30 years, probably a lot.  Maybe we would have a recession in a few years, and the earnings would fall by 20%.  Big deal.  Even if we assume no earnings growth between now and then, the yield would fall from 3% to 2.4%, before coming back to 3%.  The difference is 0.6%–the equivalent of a paltry one-time transaction fee, certainly not a reason to forego 30 years of growth in yield.

But the truth is this.  In a world without trading, where each investment had to stand on its own merits, investors wouldn’t want to be involved with either option!  They wouldn’t want to buy the aggregate stock market at 33 P/E, nor would they want to buy a 30 year treasury bond at 33 times coupon.  They would hold cash instead, because the financial and psychological value of decades worth of liquidity for them would far outweigh the paltry 3% annual return that they might receive.  Maybe 8% or 10% would be worth it, but certainly not 3%.

Bonds and stocks are fundamentally different types of investments.  Bonds have a lower duration, and a defined maturity date, when the principal must be returned.  Stocks have no such date–to earn a decent profit over the time horizon of an adult human life, you must find someone to sell them to.  Bonds pay out all of their cash flows to the investor.  Stocks only pay a fraction–with the amount ultimately at the mercy of management, a group of strangers that can do whatever they want with the money.  Bonds are a guaranteed cash flow, a repayment contract that can be recovered in court and that moves investors to the front of the line in bankruptcy.  Stocks, in contrast, are risky ventures that have no repayment contract to back them, and that almost always go to zero in the event of business failure.  Finally, in the case of treasury and mortgage bonds, these bonds can be legally bought–suppressed in yield and supported in price–by the Federal Reserve as a stimulus measure, a measure that the current Federal Reserve is very much inclined to use.

Within a secondary market, these differences have caused each type of  investment to behave differently, with different financial and economic correlations, different trading properties and conventions, and a different audience.  In an environment where the market determines the outcome, the differences make all the difference in the world.  4% on a 10 year treasury feels reasonable to most bond investors, they will comfortably buy at that yield–at least in the current environment, where they know that short-term rates will be kept low for a long-time.  But to stock investors, the same 4% yield, which implies a 25 P/E–an S&P 500 price of 2,500–categorically does not feel reasonable, regardless of where short-term or long-term rates are.  Now, one can argue that it is irrational for investors to fear owning a market at 25 times P/E, if the bond yield is low enough to compensate.  But “rational” doesn’t matter.  All that ever matters is what investors feel, because what they feel determines what they do, and what they do determines the market’s outcome.  If you poll investors, they will make it very clear: outside of a recession, where profits are artificially depressed, 25 times earnings for the S&P 500 does not feel like a reasonable price.

Legitimate Ways to Use the Equity Risk Premium

If the equity risk premium has value as a market signal–and there may be cases in which it does–the reason is not that “consistency” dictates that an investor should choose higher yielding assets over lower yielding assets.  In a secondary equity market, on realistic time horizons, returns don’t come from yields, they come from capital appreciation.  So the point is a non-starter.

Maybe a high equity risk premium reflects excessive fear in the market–fear that is depressing equity prices, and that will create significant price gains when it normalizes.  Maybe a high equity risk premium reflects an aggressively expansionary Fed that, through its pro-growth stance, will eventually rekindle risk appetite, and push investors back into the stock market.  Or maybe, in a reflexive sense, “high equity risk premium” plus “stocks are the only place left to get a return” plus “yield chase” are memes that will implant themselves into investors’ minds and create the expectation of rising prices, which will cause investors to buy, which will push prices higher, which will validate the prior expectation of rising prices, which will increase confidence in this way of thinking, which will lead to even more buying, and so on, in a positive feedback loop.  (But be careful, because when the Fed hints that they might “taper”, everyone will freak the f— out and rush to the other end of the ship).

In each of these uses, the equity risk premium is being linked to price–and therefore the appeal has validity.  But a simple appeal to investor consistency in the purchase of yield does not.  The yield alone, especially the small part that investors actually collect–the dividend–will take years or decades to accrue in meaningful amounts, and can be lost in a single day of trading.  Whatever they may say, that yield is not what equity investors–or even many bond investors–are ultimately after.

Note that when a large equity risk premium is appealed to in this way, as a condition that sets the stage for higher future prices, the details associated with the premium become important.  Why is it large?  Is it large because the earnings yield is high?  Or is it large because the bond yield is low?  And if the bond yield is low, why is it low?  The fact that a 10% earnings yield and a 6% bond yield are correlated with fantastic future equity returns is not, in and of itself, a reason for investors to expect similarly fantastic equity returns from a 6% earnings yield and a 2% bond yield, especially if both the 6% earnings yield and the 2% bond yield were engineered through a policy intervention that will eventually end.

Put differently, any noteworthy return that an investor generates from buying a market with a 4% equity risk premium will not come from the premium itself, the “spread”, but from the future willingness of others to pay higher prices than they are willing to pay now.  For this reason, the underlying factors that are driving the premium, and the way that those factors are “setting the market up” for bullish future changes in investor behavior, are critically important to the calculation.    

Buying the Nikkei at 9,600 in September 2010

In September of 2010, the Nikkei 225 traded at roughly 9600, a P/E of 16.4.  The S&P 500 traded at roughly 1125, a P/E of 16.0.  The P/Es for all countries at the time are shown below, courtesy of FT:

The 10 year JGB yield was around 1%, and the 10 year US treasury yield was around 4%.  So the Japanese stock market had an equity risk premium of 5%, and the US stock market had an equity risk premium of 2%.

What was the investment outcome two years later, in September 2012?  The S&P was up 31%, and the Nikkei was down 6%.  All while Japanese investors were (supposedly) collecting an (invisible) 5% spread, versus an (invisible) 2% spread for US investors.  Did the (invisible) spread matter?  Not in the slightest.  Yields of 2%, 5%, or 6% collected over a year don’t matter when they aren’t actually collected, and when the price can change by twice that amount in a period of a few days or weeks.

Now, fast forward to May 2013.  Voila, the Nikkei is up 51%.  Have the Fed Model and the Equity Risk Premium been vindicated?  Hardly. What drove the abrupt increase in returns was not Japan’s high ERP (due mostly to a low bond yield), but “Abenomics”, which turned horrid investor sentiment into budding optimism, a contagious desire to be invested for the coming “recovery.”  That’s all it took.  On the time scale that most investors care about, changes in sentiment, outlook, expectation, and the effect on price, are what drive returns–not small yields or spreads that take years to collect in meaningful amounts.

Buying the S&P at 1400 in December 2012

If you joined David Tepper last December and bought the S&P 500 because it was “cheap” relative to its earnings, or relative to what bonds can offer, you’ve made about 17% on your investment so far, including dividends.  Congratulations.  But you haven’t made that 17% directly from the earnings themselves, or from their spread relative to bond yields.  All that the earnings themselves amount to–assuming they were paid directly to you, which of course they weren’t–is about $40-$50, 3%.  The added 14%–almost 5 times as much–came from the changing sentiment of others, who will now pay you 1630 for each share, instead of 1400.

What has driven this changing sentiment?  The answer: steadily improving U.S. economic conditions, especially in the housing market, a sense of a return to normalcy after the resolution of the “fiscal cliff” and the stabilization of the European debt crisis, both of which suggest an end to the “era of crisis” and a return to a more normal era of steady progress, an investor-friendly Fed that continues to promise to support the economy and the market with zero interest rates for an extended period of time, corporate earnings that are managing to hold up despite feared headwinds, and finally, the market’s best friend: QE.  The Fed is perceived to be endlessly “flooding” the market with liquidity–newly “printed” money that, allegedly, has to go somewhere–thereby lifting the prices of every asset that is considered to be worth owning.  It doesn’t matter whether this dramatized description reflects the actual mechanics of the low-yield asset swap that takes place in QE–all that matters is whether investors think it does.  And many do.

These trends and conditions have not lifted the markets directly.  They’ve lifted the markets indirectly–by functioning as “inputs” into the minds of investors.  They’ve caused investors to focus more on the long-term gain of equities, and less on the short-term risk, which seems to be dwindling.  As a general rule, whenever investors are thinking optimistically about the long-term–personally, financially, or economically–they tend to want to be invested in equities, the asset class that has historically offered the highest long-term returns.  They exhibit this tendency regardless of where P/E ratios, or ERPs, or whatever other metric you prefer, happen to be.  If they see the future as sufficiently promising and exciting, as in the fall of 1999, they will buy equity markets all the way up to 40 times earnings to be “in”, even with bond yields at 6%.  And if they see the future as sufficiently scary, as in the spring of 2009, they will sell equity markets all the way down to below 10 times earnings to be “out”, even with bond yields at 2%.  Compare this 400% swing in price to the 7.5% difference in earnings yield that it represents–a yield that would take a full year, 365 days, to collect–and you will see what really drives returns.

Rest assured, if the improving trends and conditions that have materialized in the US and global economy had not emerged, you would not have your 17% return right now, even though you bought at a “cheap” price.  And if the trends and conditions had emerged at a more expensive price–if, in December, the market had been treading water for months at 1630, rather than 1400–the bottom line would have likely been the same.  Instead of moving up to 1630 from 1400, the market would have moved up to 1860 from 1630, or to some other attractive number.  Investors in the 1630 to 1860 move would be just as happy as you are now–completely unaware of the fact that they bought the market at a P/E a couple points higher, or an ERP a point lower, than you did–and those left out, just as frustrated to not have bought the market, regardless of what some manufactured metric says.  Moves in a secondary market are constrained not by absolute levels, but by the levels that participants are psychologically anchored to.  The outcome is path dependent; where you start matters.

Arguably the biggest “input” of all is the steadily rising price itself.  It creates a positive feedback loop in the minds of investors that confirms the “rightness” of the decision to invest and strengthens the optimistic expectation of what future investments will produce.  It validates the narratives that investors have been using to rationalize why they should be in the market, why it is heading higher–in this way, it increases their confidence and willingness to take future risks.  On the other end of the spectrum, the price action creates a negative feedback look in the minds of those that are not invested.  It confirms the “wrongness” of their decision to be cautious, to wait, and creates fears of perpetually missed future gains–a train that will permanently leave the station–if they stay on the sidelines.  Eventually, the pain of being the only person on the block that is not participating in the collective prosperity becomes too much to bear, and they jump back in.

When things are good, and the market is in an uptrend, most people–even those near or in retirement–want to be involved.  They want to be invested in the high-return asset class, participating in the gains that “everyone else” is enjoying.  But per the “Hold Rule”, not everyone can be invested in the high-return asset class–someone has to hold the other stuff.  It certainly doesn’t help, in the present situation, that the other stuff is yielding zero, and will continue to yield zero for many years.  But even when the other stuff is not yielding zero, these types of uptrends still happen–sometimes, more aggressively than they are happening now.

We can distinguish between two types of investments: liquid investments and illiquid investments.  An example of a liquid investment would be the online purchase of a stock.  If you change your mind two seconds after making the purchase, you can sell the stock at virtually no cost.  An example of an illiquid investment would be the construction of a new production line in a factory.  If you–the CEO or business owner–change your outlook after the new line has been built, you can’t easily turn around and sell it to someone else.  The engineering and labor costs are unrecoverable, and the scrap prices that you will get for whatever materials you used will be a fraction of what you paid.

A mistake that fundamentally-oriented stock market investors sometimes make is to assume that liquid and illiquid investment decisions are governed by the same considerations.  They are not.  Liquidity in an investment is an enormously powerful feature.  In addition to having significant economic (option) and psychological (comfort)  value in itself, its presence radically changes the way in which investors evaluate their investments, as well as the mode in which they receive their returns.

In an illiquid investment, all of the return comes from the payout of the underlying cash flows, over decades of time.  And so evaluating the attractiveness of the investment is straightforward.  Project out the cash flows, and discount them–for risk and uncertainty, the cost of money, and the cost of illiquidity, which is not small.

In a liquid investment, the cost of illiquidity is extremely small (only the bid-ask spread and transaction costs).  The majority of the return comes not from payout of the underlying cash flows, but from the ability to sell the investment to others.  For this reason, the piece of information that really matters, that effectively decides the outcome, is the market price: what others offer to pay for the investment in the future.

Some investors will claim that their time horizon is forever, and that they don’t care about the market prices of their investments.  If they make this claim, ignore them; they are bullshitting.  They definitely care.  When the prices of securities they own rise rather than fall, you will not hear them talking about “infinite time horizons.”  You will hear them touting their investment process, celebrating their “return”, which they consider to be very real.

In the paragraphs and posts that follow, I am going to explore the different considerations that pertain to liquid and illiquid investments, and present a general framework for how we should think about each type–with a particular emphasis on how to think about liquid investments in a stock market.

A Simplified, Easy-To-Understand Model

As always, we begin with a simplified, easy-to-understand model.  We will use this model to help clarify the true, illiquid value of assets, as well as the importance of liquidity preference as a determinant of the price that will be paid to invest in them.  Try, if you can, to think about the model as if you were really in it, right now, making the decisions that each person has to make.  You will see first-hand how impactful considerations about liquidity are.    

Suppose that you, John and Laura are investors in a closed market.  This market contains three types of assets: stocks of different companies securitized into an index ($SPY), government bonds similarly securitized ($TSY), and cash.  

There are 90 oustanding shares of $SPY, 90 outstanding shares of $TSY, and $9,000 of cash.  That is the universe of existing assets.  At present, each share of $SPY earns $6 a year in profit from the underlying companies, with the earnings growing over the long term at roughly the rate of inflation, which we will assume is around 2%.  The earnings grow only at that rate because there is no internal reinvestment; 100% of the earnings are paid out as a dividend each year.  Each share of $TSY pays out a constant $4 of cash per year in interest, and will mature in 10 years, at which point each share will be exchanged for $100 in cash.  The $9,000 of cash is just money: dollars held electronically in a bank account.  Currently, for each dollar that is not used to make purchases, the bank pays out 1 penny of interest per year, in compensation for the ability to lend it to others.  This interest rate is set by the government, and changes based on the government’s management of economic conditions.

Because you, John and Laura are the only individuals in this market, one of you must hold each outstanding unit of each asset class at all times: each share of $SPY, each share of $TSY, and each dollar bill (or byte).  We call this rule the “Hold Rule.”  There are no exceptions to it.  If no one wants to hold a give unit of an asset, the “price” of that asset, expressed in terms of other assets, will fall, until someone emerges that does want to hold it.  

Suppose that you, John and Laura come into existence with the assets distributed equally, so that each of one of you owns 30 shares of $SPY, 30 shares of $TSY, and $3,000.  This distribution satisfies the “Hold Rule”, because the sum of what each of you is holding equals the total outstanding quantity of each asset.  But this randomly chosen distribution is unlikely to satisfy your various preferences.  So I am going to “open the market.”  That is, I am going to give you the opportunity to trade the assets freely among yourselves, at whatever exchange rates you choose.

The million-dollar question is, once the market is opened, what will the exchange rates between the assets be?  How many dollars for each share of $SPY?  How many dollars for each share of $TSY?  And if shares of $SPY and $TSY can be swapped directly, without going through the medium of cash, what will the ratio between them be?

Illiquid Investment: No Uncertainty, No Trading

To get a better picture of the true, fundamental value of each asset, we need to eliminate trading, wherein an investor can generate large returns by correctly anticipating the changing preferences of others, rather than by collecting the cash flows of the asset itself.  So let’s assume that the market will remain open for only one day.  On that day, you, John and Laura will be able to freely conduct exchanges in accordance with your preferences.  Once the day is over, the market will close forever, and the three of you will have to stick with whatever you have.  Whoever is holding shares of $SPY will have to stick with those shares, forever.  Whoever is holding shares of $TSY will have to stick with those shares for the next 10 years, at which point each share will be redeemed for $100 of cash that will have to be held forever.  Whoever is holding cash will not be able to hold any of the other assets, but will have the ability to use the cash to purchase goods and services at any time.  In theory, the person holding cash would have the ability to use it to create a new asset, but to be fair to the other asset classes, we will consider this “real investment” to be a type of trading, and assume it is not a viable option.

If you, John and Laura are perfectly rational agents, then 3 questions will determine the relative rates at which you will offer to exchange the assets.  These 3 questions are:

(1) What will the future $SPY payouts be?     

(2) What will the future interest rate on cash be?

(3) What is your liquidity preference?  How important or valuable to you–both economically and psychologically–is the ability to have and spend your money now, versus later?  What is the cost to you–again, both economically and psychologically–of having to separate with your money, not be able to touch it, for long periods of time?  As an investor, what is your time horizon?

Unlike question (3), questions (1) and (2) are outside of your control.  In the present scenario, we are going to eliminate them.  Suppose, then, that I tell you exactly what the future payouts of $SPY will be, and exactly what the interest rate on cash will be, in each future year up to eternity.  Because we’ve eliminated the ability to make money by trading, you will know everything there is to know about the returns offered by each asset class.  Those returns are shown in the table below:

As we see from the table, the advantage of $SPY relative to cash is that it will pay out more over time. $TSY also has this advantage, but the advantage is lessened by the fact that the payouts do not grow with inflation, and also by the fact that the fund matures into cash in 10 years.  

The disadvantage of $SPY relative to cash is that you will have to wait to get the money.  If you exchange $100 of cash for $SPY, it will take you 14 years to get that amount of cash back (by then, you will have $100, plus your shares of $SPY).  In the interim, you will only be able to spend what you have accrued.  $TSY carries this same disadvantage relative to cash, but the disadvantage is reduced by the fact that there is a maturity date, a time at which a lump sum of cash will be paid to close out the fund.

Think about the scenario in real terms, as if you actually had to make the investment choice right now.  From the initial distribution, you have $3,000 of cash, 30 shares of $SPY, and 30 shares of $TSY.  You can trade any of these assets for any other asset, at whatever ratio you choose.  The only catch is that John and Laura–the individuals with whom you will be making the exchange–have to agree to the ratio for a trade to be carried out.  

Suppose that John makes the first offer.  He has a very long investment time horizon, and doesn’t need the cash.  To get the highest long-term return, he wants to acquire shares of $SPY.  So he offers to exchange $100 of cash for each $SPY share.  He says “Look guys, I’m offering 15 times earnings, that’s a very fair price.”  Would you take his offer, and sell your shares?  If not $100, then what would your minimum price be?  $150?  $200?  $300?

Let’s assume that, like John, you and Laura both have very long investment time horizons.  Whatever wealth you have, you plan to put it aside and not spend it for decades.  If that’s the case, then the eventual price in cash that each of you will end up offering to pay for each other’s shares of $SPY will be much higher than $100–maybe $150, or $200, or $300.

These prices would represent price-earnings (P/E) ratios of 25, 33.3 and 50 times earnings respectively–quite expensive relative to what most of us are used to.  But it doesn’t matter.  There is no rule written into markets that says that the P/E ratio must equal some number.  The only rule is the “Hold Rule”–the rule that each unit of each asset in existence must be willingly held by someone at all times.  If each of you wants to hold $SPY, and none of you wants to hold cash, then the exchange rate between $SPY and cash will rise until one of you changes your mind.  Period.  The same is true of $TSY.  The only difference is that there is a limit to the price that an investment in $TSY can rationally command.  In our example, if, to hold $TSY, you pay more in cash than $145.98–that is, $100 of principal plus $40 of collected coupon payments compounded for the relevant period of time at the cash interest rate–then you will have essentially given away your money to someone else for free.

Ultimately, $SPY and $TSY are a type of cash–with the access delayed over time, in accordance with the intervals specified in the table.  And so if you think about the dynamics of the decision in front of you, you will see that everything comes down to that same question: what is the difference in value, for you, between “cash now” and “cash later”?  What is your liquidity preference?  If the three of you perceive there to be no difference whatsoever between the value of cash now and the value of cash 10 years from now–if having the actual money in your hand, being able to spend it whenever you want over the next 10 years, is worth zero to you–then the maximum amount of cash that you will be willing to exchange to hold $TSY will approach $145.98, an excess return of zero.  And if the three of you perceive there to be no difference whatsoever between the value of cash now and the value of cash in a million years, then, assuming low and stable cash interest rates in the interim, the amount of cash (or $TSY, if you exchange the security directly) that you will be willing to exchange for $SPY will approach some absurdly high number.

Being willing to pay enormous amounts for $SPY might sound crazy, but it makes perfect sense, provided that you have a long enough time horizon.  Even if you pay $1,000 a share, there will come a time when your investment will have produced more than that amount, and, assuming low and stable cash interest rates, more than cash or $TSY will have produced.  The only question is whether it’s worth it to you to wait that long.  For most of us, it isn’t.

Illiquid Investment: Add Uncertainty 

It is common for economists to speak of a “risk premium”–a premium, in added return, that the holder of an asset with uncertain cash flows (such as equity in a company) demands in exchange for holding it, as opposed to holding a guaranteed asset.  In the first scenario, no risk premia were necessary, because we disclosed the future cash flows of each asset out to eternity.  In this scenario, we will reintroduce uncertainty to see how the problem changes.  What we will see is that because we are utilizing indexing, the problem doesn’t change much at all.

Before we begin, let’s ask the question, why does uncertainty even require a risk premium?  For each investment, there is an expected (or mean) return.  The uncertainty around that return applies in both directions–to the upside and downside–therefore it doesn’t change the mean.  So why should we consider the uncertainty to be a net negative that requires compensation?

The answer lies in the disparate way that the human mind evaluates profits and losses of the same magnitude.  They are not the same, and they do not cancel each other out.  To illustrate, suppose that there is a company that has a 50/50 chance of generating a $150 total profit, or a $50 total profit, tomorrow, after which it will dissolve.  Mathematically, the expected (or mean) return of an equity investment in the company (+$150, +$50) is $100.  Even though the expected return is $100, investors are not going to pay $100 in exchange for the equity.  The reason is that the prospect of a $50 gain is not commensurate with the prospect of a $50 loss.  Investors are, on average, risk-averse.  The cost of a loss is perceived to be greater than the benefit of an equally-sized gain, and therefore investors demand to receive compensation over and above the expected return on the investment.  That compensation is the risk premium–the compensation for taking the risk, which could be avoided altogether.  For a given expected return, the greater magnitude and probability of the potential loss, the greater the risk premium needs to be.

To test this out in your mind, consider the following proposition.  I’m going to flip a coin. If it comes out heads, I will pay you X dollars.  If it comes out tails, you will pay me X dollars.  Would you accept the offer?  The expected return of the offer is 0, which is exactly what I am you charging to take it.  So will you take it?

Granted, if X is really small, like a few dimes or pennies, you might take it.  Gamble a bit, for fun.  On the scale of extremely small potential losses, human risk-aversion approaches zero, and the rewards of excitement, humor and leisure can outweigh it.

However, as the amount of money that can be lost grows, the risk-aversion grows–in non-linear fashion.  To illustrate, suppose that X equals your entire net worth.  If heads, you double your net worth, if tails, you lose your net worth.  Would you accept the offer?  Of course not.

To get you to accept the possibility of losing everything that you own, I would have to compensate you by dramatically skewing the expected return–with the extremity of the skew determined by how much the loss of your net worth would hurt right now.  Depending on your personality traits and life situation, I might have to offer to pay you as much as, say, 10 times your net worth if you win the flip.  But even 10 to 1–an expected return equal to 9 times what you are putting at risk–might not be enough.  When your net worth is at stake, we could very well be at a singularity, where no expected return, no matter how large, is worth a 50% chance of losing everything that you own.

Obviously, with respect to an individual company, there is significant uncertainty around the outcome.  The cash flows that the company ends up producing could fall dramatically below the expectation.  Worse yet, the company could go bankrupt, disappear forever, creating a permanent loss of capital.  Subjecting one’s wealth to that uncertainty demands compensation.

But in our scenario, we are not contemplating an investment in an individual company.  Rather, we are contemplating an investment in a collection of thousands of companies pooled together.  When  pooled together, the winners among the companies cancel out the losers.  The result is a much tighter distribution around the expected outcome, and therefore a much smaller required risk premium.

In the case of $SPY, we can be reasonably confident, based on history, that next year the asset will produce something close to what it produced this year–$6.  We can also be reasonably confident that the $6 will grow over time, at about the nominal growth rate of the economy.  There will be expansions and recessions in which earnings will rise and fall around the trend, and so a risk premium is still necessary, but there is no reason why it needs to be particularly large.  As a consideration in the problem, it is going to be dwarfed by the far more important consideration of liquidity preference.

To capture the point intuitively, put yourself back in the scenario, except without the table that tells you what $SPY’s returns will be.  You know that $SPY is currently paying out $6, but you don’t know for sure what it will pay out in the future.  John offers to sell you $SPY for $150.  Based on your estimates, the payback will be roughly 20 years.  But you can’t be sure of that estimate.  Depending on how the economy performs, the payout could be 15 years, or it could be 25 years.  How much does this uncertainty dissuade you from the investment?  Probably not much at all.  The key consideration for you, far more important than the uncertainty around the trend in $SPY’s payout, is your liquidity preference, the difference in value for you, economically and psychologically, between having cash now and having cash in the future.  Ultimately, it is that preference that will make the difference in determining the maximum price that you will offer to pay.      

Liquid Investment: Enter the World of the Stock Market

What we have in our first two scenarios, where you, John, and Laura must decide who will hold a set of outstanding equity, fixed income, and cash assets over the long-term, is a nice, neat problem that we can solve rationally with only two pieces of information: (1) the total future cash flow that each asset will deliver, and (2) each investor’s liquidity preference, i.e., the cost to each investor–both economically and psychologically–of parting with money, not having it or being able to use it, for extended periods of time.

Now, let’s remove the artificial constraint that the market will close forever tomorrow, and that each individual will have to stick with whatever she chooses to hold.  Assume that assets can be freely exchanged indefinitely, and that cash, in addition to being spent on goods and services, can be used to create new assets.  

Unfortunately, this new formulation radically changes the dynamic of the problem.  Because we’ve introduced the concept of trading in a secondary market, the problem is now recursive.  With respect to (1) above, the total future cash flow that the asset will deliver to the owner is no longer just a function of what the asset will earn or generate in its own operation.  It is now a function of what the asset can be sold to others for.  With respect to (2), the consideration that encapsulates the true cost of the investment in terms of lost liquidity is no longer the maturity or payback period of the asset, but rather, the future willingness of others to purchase the investment from the owner, should the owner want to sell it.  Each investment becomes fully liquid as if it were cash, except that its value fluctuates every day based on how eager others are to own it.  If you choose to get out of the investment, you will have to expose yourself to that fluctuation, which is a net negative to the proposition, given risk-aversion.    

It is extremely difficult, if not impossible, to try to logically model how these considerations should interact to determine the price of an asset.  If I don’t plan on holding the asset until maturity, and generating a return from its cash flows directly, I can’t know, with precision, what the highest price I should be willing to pay for the asset is, unless I know what price others will be willing to pay, at various points in the future.  But I can’t know what price others will be willing to pay, at various points in the future, unless I know what price those others think yet others–to include me!–will be willing to pay, at various points farther out into the future.  

The logical intractability is made worse by recursion inside the individual: what George Soros calls reflexivity.  The prices that the investments are trading at are displayed on a “tape”, for all to see.  The investors’ views about the investments determine those prices, but those prices also determine the investors’ views about the investments.

Each investor is a human being with insecurities.  Whatever he might proudly say, his views are influenced by the aggregate views of others–which is what the “tape” expresses to him, the collective wisdom of his peers.  The tape can make him scared, cautious, greedy, confident, impulsive, excited, elated, bored–emotions that have the power to alter his investment time horizon, his tolerance for risk and uncertainty, his assessment of the underlying merits of his positions, and his sense of how well those positions will perform.  The result is a pricing mechanism that exhibits both momentum and path dependence.  Past prices can influence future prices, and securities with the same fundamentals can easily arrive at disparate prices, depending on the paths they take to get there.

Because a liquid market functions in this way, the insights that matter are not insights about what is worth what, or what is “cheap”, or what is “expensive”, or what my “discounted cash flow model” says, based on information already given or information about to arrive.  The insights that matter are insights about what other people are going to choose to do in the presence of that information.  What other people will do in the face of whatever path reality takes is what decides returns, and is therefore what investors in liquid markets should be trying to understand and model.  Anything that cannot in some way be related to that question is a distraction, an attempt to treat liquid investing as if it followed the radically different rules of the illiquid.

In the next piece, we will use insights gained here to refute the popular “Fed Model” for equity investing, which evaluates the attractiveness of the stock market as an investment based on how its earnings yield compares with long-term bond yields.