2017: The End Of Belshazzar's Feast

Includes: DBB, DBC, GCC, GSG, RJZ, SPY
by: John Overstreet


Expectations of a bull market continuing on until 2025 or 2030 are unrealistic, and perhaps dangerously optimistic.

Using historical analysis, one ought to expect a Depression beginning in 2017 and low returns until 2030.

Bull markets end with severe imbalances in market ecology, these imbalances have been amplified by central banking, and the current imbalances are now at extreme levels.

The next crisis will likely amplify domestic and international political turmoil, making it difficult to address the underlying economic errors.

The most realistic outlook for the period up until 2030 is a scenario that resembles the 1929-1943 years.

What a difference Dow 20,000 makes. As the index first crossed that threshold in late January, Art Laffer, arguably President Trump's most enthusiastic cheerleader on economic issues, promised "an enormous surge in asset values" over the next seven or eight years. "Almost anything [the president] does will be right and will create prosperity", the world's most one-handed economist told FOX's Neil Cavuto. Having said that conditions are even more propitious for Trump than they were for Reagan in 1981, he seems to suggest that the stock market will double in value by 2025. A couple of days later, and somewhat more warily, Barron's claimed that, barring an unlikely trade war, the Dow would hit 30,000 by 2025. "Our projection… is based on our analysis of historical data….This data, which looks at stock market returns… dating back to 1871, suggest stock market gains will fall below the market's typical annual gain of 6% after inflation in the next five years before accelerating above the average in the years after that."

Both of these outlooks are dangerously optimistic. Historical data suggests that we will be lucky to hit Dow 30,000 by 2030, and that once the markets top out later this year, the Dow (NYSEARCA:DIA) will remain closer to 10,000 than 30,000. In other words, the structural imbalances that have been developing in the American economy and around the globe for the last century, which I described in my two previous articles, are likely to cause the greatest economic collapse since the Depression.

But first, let me see if I can build a bullish case for stocks using S&P Composite (NYSEARCA:SPY) data going back to 1871. Look at the chart below, and tell me what you see (this is a rhetorical device not meant to be taken literally).

1885-2016 S&P 500 14-year CAGR

(Source: Robert Shiller data)

Perhaps the first thing one notices is that our stock market expansions are becoming both longer and more intense. Since the end of World War II, we have had two incredible stock market booms, and if that pattern holds now, perhaps we are already well into our third. What is more, there is less downside risk under Pax Americana: over any fourteen-year period, one can be sure to have at least broken even, unlike the years before the war, when you could be down a third or more fourteen years in.

These developments are illustrated below.

(Source: Robert Shiller data)

There is also a cyclical element here. In the last two stock market booms, there were about seventeen years of high growth followed by fourteen years of low growth. Thus, the stock market boomed in 1950-1968 and slowed in 1968-1982. Then, stocks rose in 1982-1999, only to decelerate in 2000-2013. If that pattern were to continue, we may only be four years into the third post-war boom, meaning that the Dow could be at 75,000 by 2030 (using the relatively tame 1954-1968 period of 10% annual gains as a benchmark). Even more optimistically, if the 1986-2000 period were to repeat itself, we could be at Dow 120,000 by 2030!

1886-2016 S&P 500 14yr CAGR with supercycles demarcated (Source: Robert Shiller data)

One reason for pessimism, as some will surely point out, is the rather rich valuations on current stocks. Robert Shiller's Cyclically-Adjusted P/E ratio (CAPE, also called the P/E10) has been approaching 30, a level that has not been exceeded in any significant way since the boom of the 1990s, when it hit 40. With a bit of earnings growth and a multiple expansion, that would be enough to push stocks to the target set by Barron's, and if the historical trend we identified above of more intense booms continues, then perhaps we could reasonably expect the doubling in stock prices intimated by Art Laffer to occur by 2025.

1871-2016 S&P 500 14yr CAGR vs CAPE (Source: Robert Shiller data)

This is highly unlikely, however, and not just because the CAPE ratio is "too damn high". The question one must ask is why stock market booms have become both more durable and more intense over the last century. Except for the persistently high valuations of the last twenty-plus years, it is not clear that there has been an especially strong tendency for multiple expansion to be behind this longer-term phenomenon. There are at least two better (and interrelated) reasons for increasingly huge stock market booms: structurally high earnings growth and inflation.

1871-2030 earnings and consumer price index absolute values (Source: Robert Shiller data)

I assume that these phenomena were in turn caused by central banking. The Fed was established in 1914, and it is from that period that we begin to see surges in earnings and inflation - as well as the structural changes in the economy that are connected with those surges, such as the rise of technocracy, industrial productivity growth, suburbanization (or more properly, urban deindustrialization), social costs, consumerism, state capitalism/social democracy, globalization, the "servicification" of the economy, and perhaps even climate change. I discussed many of these changes in August.

One might assume that as long as we have a central bank that can reliably produce inflation and earnings growth, we will continue to see thrilling stock market returns, just as we have (with a few brief exceptions) ever since the end of World War II.

1871-2030 S&P 500 vs Earnings absolute levels (Source: Robert Shiller data)

So, this all really comes down to a bet, or perhaps a question of faith: Can the Fed continue to generate inflation in the future? Is central banking a sustainable way to achieve growth?

The answer to both questions is No. Setting aside the mysteriousness of the way inflation is generated in the first place and why it is distributed across different sectors and different asset classes at different rates - in fact, one could argue that central banking even causes deflation in some cases (such as durable goods) - historical market behavior suggests that the ecology of global markets has become so unbalanced that we are on the brink of a systemic collapse. Below I will sketch the outlines of that ecology and identify the key historical imbalances.

To put this another way, in the previous two articles, we examined the way in which inflation was expressed in different economic, geographical, and political sectors over the last 150 years, but in this article, we will look at the way that inflation is expressed across asset classes, keeping in mind that no one (as far as I am aware) has yet solved the riddle of inflation.

1. Key relationships: Cycles

So, in periods of stable growth, earnings, stock prices, commodities (DBC, GSG, GCC), bond and stock market yields, and GDP tend to push through market cycles in unison. It is only during stock market booms that this cyclicality becomes disjointed. Again, it is not clear why inflation manifests itself first here and then there; instead, for the moment, we are focusing on the regularities and irregularities of price phenomena.

At the heart of these relationships are three factors - earnings, industrial metals, and the earnings yield. They have always expressed strong tendencies towards cyclicality and correlation with one another. Somewhat paradoxically, periods where those tendencies are strained or even reversed are also periods of stock market booms, but the process of realignment has always been bearish.

At the cyclical level, we can see that cycles in earnings, metals prices (DBB, RJZ) and the earnings yield have always moved in unison under all circumstances. The following chart illustrates the correlations between GDP, earnings, industrial metals prices, and short-term interest rates since the 1960s.

1960-2015 earnings 3m-IR GDP metals prices cycles (Sources: World Bank, St Louis Fed, Robert Shiller)

The next two charts illustrate the ancient (in terms of market history) cyclical connections between earnings, the earnings yield, and commodity prices.

1871-2016 earnings and earnings yield cycles (Source: Shiller)

1871-2016 commodity & earnings cycles (Sources: Shiller, World Bank, St Louis Fed)

And this fourth chart illustrates the way in which the fragmentation of the relationship between stock cycle and the earnings cycle only occurs during stock market booms. As stock market booms have become more persistent and more powerful, the greater the degree to which cycles fragment.

(Source: Shiller)

2. Key relationships: The earnings yield and commodity prices

We also know that the earnings yield and commodity prices have always been highly correlated with one another over longer durations. Below, we present three snapshots of commodity prices to illustrate both the continuity and change within this relationship.

In the first, we present the classical version of this relationship. Prior to the establishment of the Fed, commodity prices were highly correlated with the earnings yield. This is extremely similar to the relationship underlying what Keynes called "Gibson's Paradox", i.e. the strong correlation going back to the early 1700s between the general price level and government bond yields.

1871-2016 commodities & earnings yield absolute values (Sources: Shiller, World Bank, St Louis Fed)

After the Fed was established and began to persistently generate positive inflation, the relationship grew strained, but notice that the period of high stock market returns after World War II line up with the breakdown in the Gibsonian relationship.

We can reestablish that relationship by deflating commodity prices with the consumer price index. Thus, we find that commodity prices still travel with the earnings yield.

1900-2011 earnings yield and real commodity prices

(Sources: Shiller, University of Michigan Library, Stephan Pfaffenzeller)

We can also demonstrate this relationship by comparing the inflation rate in commodity prices with the earnings yield.

1871-2015 commodity inflation 7yr CAG4 vs earnings yield (Sources: Shiller, World Bank, St Louis Fed)

Since the establishment of the Fed, commodity prices have tended to grow at roughly the rate of the earnings yield itself. (As an aside, one of the interesting things about the last century is the degree to which inflation rates and bond yields have tended to gravitate towards the earnings yield).

In short, with respect to the relationship between commodity prices and the earnings yield, there is a remarkable degree of consistency, even despite the collapse of the classical, Gibsonian relationship.

We then come to earnings and commodity prices.

3. Key relationships: Earnings and commodity prices

Earnings growth and commodity price inflation (especially for industrial metals) have been tightly linked over cyclical durations for as far back as we have records. In light of this and the strong relationship between commodity prices and the earnings yield over the long term, what is at least as fascinating (for me, at any rate), are the breakdowns in the long-term relationship between earnings and commodity prices. Of course, this is suggested in the chart above that shows that commodity prices deflated by consumer prices are in sync with the earnings yield.

By extension, we can say that the ratio between commodity prices and earnings is highly correlated with the ratio between real stock prices (the stock index divided by consumer prices).

1871-2011 earnings/commodity ratio vs real S&P 500

(Source: Shiller)

That leads us almost immediately then to the relationship between the earnings yield and earnings themselves.

4. Key relationship: Earnings and the earnings yield

Specifically, if commodity prices (whether looking at the absolute deflated price or at the long-term rate of commodity inflation) are closely linked to the earnings yield, and commodity prices have an enigmatic relationship with earnings, then what might the relationship between earnings and the earnings yield be? Obviously, the ratio between earnings and the earnings yield (i.e. E/EY, or E/(E/P)) simply yields the price of stocks ((E/(E/P)=P)).

More interesting is the relationship between earnings growth and the earnings yield. Now, this too can yield a boringly familiar value: Shiller's CAPE. That is to say, the CAPE is, on one level, stock prices divided by the average of ten years' earnings (i.e. P/E10), but on another level, it can also be expressed in the following equation:

(E/E10)/(E/P), or


Shiller's CAPE is the ratio between earnings growth (E/E10) and the earnings yield (E/P). And, as a simple mathematical fact, the CAPE can reach historically high levels if the earnings yield falls very low and/or earnings growth is very high. CAPE has been used both as a tool for stock market valuations and as proof that stock market returns are dominated by psychological factors, specifically the emotional states of optimism and pessimism.

But, we know from the tight cyclical relationships among commodities, earnings, and yields and from long-term relationships between commodities, earnings, and yields that a) these factors have a tendency to move in sync with one another and b) breakdowns in synchronicity coincide with stock market booms.

As far as the breakdowns are concerned, we have already demonstrated this in the ratio between earnings and commodities, but we can also show this in the relationship between commodity prices and the earnings yield and then with earnings and the earnings yield.

First, we look at absolute commodity prices and their rolling, thirty-year correlation with the earnings yield. Notice that as stock market booms begin, the correlation coefficient weakens and then turns negative. This is not a timing tool, but it illustrates the relevance of the Gibsonian relationship described above.

1871-2016 S&P 500 14yr CAGR vs rolling correlation between commodity index and earnings yield (Sources: Shiller, World Bank, St Louis Fed)

The same thing is illustrated in the correlation between earnings and the earnings yield below.

1871-2016 S&P 500 14yr CAGR vs rolling correlation between commodity prices and earnings yield

(Sources: Shiller, World Bank, St Louis Fed)

But, the relationship that we are interested in with respect to the CAPE ratio is the correlation between earnings growth and the earnings yield. And what we find is that the vicissitudes of the correlation between earnings growth and the earnings yield are at least as accurate as the CAPE itself and more precise in identifying stock market peaks. The following chart shows the correlation between Shiller's E/E10 growth rate and the earnings yield.

1871-2016 S&P 500 7yr CAGR vs correlation between earnings growth and earnings yield (Source: Shiller)

1871-2016 stock market performance vs PE10

(Source: Shiller)

I have found that the seven- and eight-year relationships are best.

1871-2016 S&P 500 7yr CAGR concurrent vs correlation between earnings growth and earnings yield w/ market-top indicators (Source: Shiller)

1871-2016 correlation between earnings growth and earnings yield vs subsequent stock market performance (Source: Shiller)

1871-2016 rolling correlation between earnings growth and earnings yield vs subsequent S&P 500 performance 14yr CAGR (new) (Source: Shiller)

These charts all show that during the late stages of stock market booms, the rate of earnings growth breaks free of the earnings yield (and to a lesser extent commodity inflation), but once there is a significant breakdown in the correlation (that is, once the correlation turns negative), one can expect very low returns over both the subsequent seven- and fourteen-year periods.

Putting this another way: if the CAPE, as noted above, can reach unsustainably high levels through an unusually low earnings yield and/or a high earnings growth rate, we have also found that in most cases, especially since the establishment of the Fed, CAPE has reached those extreme levels by doing both simultaneously (as in the 1920s, 1960s, 1990s, and the 2010s). We have effectively reached that state as of late 2016, meaning that we can expect a bear market to begin in the near term and last until 2023 and 2030.

1871-2016 earnings growth 7yr CAGR vs earnings yield (Source: Shiller)

1871-2016 PE10 vs correlation between earnings yield and earnings growth (Source: Shiller)

Now, even if this forecast is correct, that does not in and of itself indicate a worst-case scenario like a depression, which is what I have been predicting for the aftermath of the 2017 peak. To arrive at that conclusion, we have to combine the Gibsonian relationship popularized by Keynes with market supercycles, or what Joseph Schumpeter called Kondratieff Waves.

Why a depression?

We have had two secular bear markets since the end of World War II, and neither of them coincided with a depression. In fact, in both cases (the 1970s and 2000s), there was an uptick in inflation and a commodity boom. So, what makes this time different?

Quite simply, if we recall the relationship between the earnings yield and commodity prices/inflation, which is the foundation for the Gibsonian equilibrium, it is fairly obvious that since the commodity boom that peaked c.1920, the earnings yield and commodity prices have peaked about every thirty years (e.g. 1950, 1980, and 2010). It has taken about twenty years for the earnings yield and commodities to bottom out after those peaks. So, on that basis, we can assume that the earnings yield and commodity prices will remain suppressed until about 2030.

Prior to 1920, these long waves, as the Soviet economist Nikolai Kondratieff himself termed them, lasted closer to fifty or sixty years. It was only after the Fed was established that we saw the massive structural transformation of the American and then global economies: the alteration in Gibsonian relationships, the shortening and regularization of Kondratieff's long waves, and the rise of cost disease, technocracy, the service sector, consumer culture, the Penn Effect, globalization, and convergence, as well as the demise of the Phillips Curve.

Under the admittedly risky assumption that history repeats itself, we will experience a secular fall in stocks, commodities, and the earnings yield until 2030. Now, if the earnings yield is to remain low or even fall further over the next five, ten, fifteen years, and if it is not going to be pushed down by stocks as it was in every other bull market since World War II, and if stocks will not have the benefit of inflation to keep them afloat as in the 1970s and 2000s, then that means an earnings depression will have to occur, also.

We have not had an earnings depression since the 1930s, and this will constitute the first time since the Great Depression that the economy and the central bank upon which it stands fails to generate inflation anywhere in its system.

For these reasons, I believe that we are on the verge of an extended, severe crisis. In some ways, it has already hit. Stocks have been in a huge bull market since they bottomed in March 2009, but the S&P 500 has only gained about 50% in 17 years. If the analysis made in this article proves correct, in 2030, we might look back and see that the stock market made virtually no gains since 2000. Moreover, the frailty of the global economy since the 2008 crisis and the consequent collapse in faith in the technocratic authority are, it seems to me, indicators, as they were in the 1920s, of a world on the brink of catastrophe.

Ideological risk

Under such circumstances, politics and, more ominously, ideology will play an increasingly important role in market analysis. What kind of economic and monetary reform can we expect under present conditions when the next crisis hits, even if it is only a quarter as bad as I make it out to be here? Will President Trump get the blame for it, or can he manage to pin the blame on domestic obstruction or external threats? He will certainly have to jettison Laffer's creed of presidential infallibility.

The president is an economic nationalist and appears to have no qualms about using state power to goose the economy (not that it will work), but the socialistic left has no such qualms either. Moreover, even if the political rancor can ultimately be traced to economic causes, it is increasingly bound up with battles over identity, ensuring that rational economic reform will get lost in the heat of political battle.

John C. Calhoun, the infamous Southern apologist for slavery, was also, ironically, a gifted theorist on the critical role of minority rights in a constitutional order. Developing Publius's insights into the dangers of popular government, he pointed out that stable republican governments are rare and almost always the product of historical accident; in republics that are reduced to simple majoritarian competition for power and resources, the polity will split into two hardline factions that repeatedly escalate their rhetoric and violence until all that remains is tyranny.

In all honesty, I had hoped that the technocratic elites that got us into the present mess, represented by Hillary Clinton more so than Donald Trump, would win the election simply in order for them to get the blame for the coming crisis, which would then put the US and the rest of the world in a position to reexamine the foundations of economics and possibly put us on a path to disestablishing the central bank. With the success of some rather raw populists in different corners of the world in recent months, it seems just as likely now that the hands of the technocrats will be made stronger by the next crisis or that the right-wing populists will feel emboldened to take desperate measures to advance their interests and/or that we will descend into open conflict between the hard left and hard right.

The reason I am so pessimistic is that in 1929 we were pretty much right where we are now, and it never occurred to the People Who Know that central banking was the cause behind the commodity/inflation/earnings boom of the 1910s, the stock market/earnings boom of the 1920s, the Depression of the 1930s, and the ideological and cultural radicalizations that brought us to World War II, which, if a glorious achievement in turning back the barbarians, also confirmed the victory of American-style state capitalism, or neoliberalism. It probably did not help that the class of people chiefly responsible for thinking about such things, i.e. economists, politicians, and financiers, all stood to benefit in power, prestige, and wealth from such an arrangement. Libertarians, classical liberals, and old-school conservatives were all shunted aside and have been on the fringes ever since; it did not help that their doomsaying never managed to come true, a reliable historical precedent I would also be wise to heed, no doubt.

And yet, something new is going to happen starting this year - my guess is that it will begin in June or July. Either the historical patterns of the last 100, 200, 300 years will repeat themselves, in which case the economic system in place for over a century will descend into an extended (but perhaps not permanent) chaos, or they will not, which would be equally momentous if less disastrous. I could be wrong, but when I look at the historical chart of earnings, I cannot help but notice that the booms of the 1910s-20s had the exact same pattern that they have had in the 2000s.

1871-2030 earnings w/ pre-Depression booms identified (Source: Shiller)

Perhaps I have mistranslated it, but the writing seems clear: Mene, Mene, Tekel, Upharsin. Something is wanting. Globalization, state capitalism, technocracy, the markets - the beneficiaries of the post-war era - are exhausted.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.