- This market has all the characteristics of a strong bull: weak commodities, a wide yield curve spread, disinflation, P/E expansion, and counter-cyclical stock behavior.
- This market and the crisis that preceded it is often compared to the Depression, but a far better comparison is the Roaring Twenties and the 1921 crisis.
- That leaves us with three difficult questions to answer: when will this bull market end, what will the aftermath look like (Depression? Stagflation?), and do we have any policy alternatives?
- The market will probably grow for a few more years, and the aftermath will probably be pretty grim, as our understanding of market forces has barely progressed since 1929.
How strange for the stock market (SPY, DIA, QQQ) to perform so well despite such persistently dire conditions. But, how unusual is this, really? I have argued in previous articles that a bull market in stocks typically follows a recession that occurs during a commodity (GSG, DJP, RJA) boom, but in this article, I would like to point out that this market not only resembles the mid-1970s and the 1980-1999 market, but perhaps as much as any other, the early 1920s. One can argue, therefore, that the comparisons between our current economic troubles and the Great Depression are somewhat premature.
More specifically, in the 1920s, as in this market after 2007, a sharp, short-term P/E expansion caused by a severe recession after a commodity boom ultimately morphed into a huge stock market run that coincided with deflation and global economic and political turmoil. Although this Great Recession is often compared to the Great Depression, from the perspective of the behavior of financial assets, as well as inflation, the more apt comparison is with the 1920s.
Anatomy of a Bull vs. the Anatomy of a Bear Rally
In the modern market, during commodity booms, stock markets tend to be highly cyclical, rising and falling with commodities, earnings, and interest rates (UST, IEF). As we came out of the commodity boom of the previous decade, a lot of commentators were thrown off the scent of the market because they were sticking with patterns that only apply during such booms, and they failed to note the transition to a secular bull market in stocks.
During secular bull stock markets, commodities tend to remain moribund, although individually they can certainly experience quite impressive runs from time to time, such as the strong metals market in the mid- to late 1980s that coincided with the even more impressive performance of Asian stock markets.
Inflation also tends to fall dramatically. Inflation fell from nearly 15% in 1980 to under 2.5% in 1983 and 1% in 1986. In 2007, inflation hit 5.6%, went to -2.1% in 2009, and has hovered between 1 and 2% since. In 1974, inflation was over 12% but fell to 4.9% by 1976 during the brief P/E expansion (and gold collapse) of the mid-1970s.
During secular bull markets, stocks tend to be somewhat more countercyclical, accelerating when interest rates and profits decline, and decelerating or falling when interest rates and profits rise rather than being tied slavishly to the cyclical movements of commodities, interest rates, and profits.
In other words, the difference between a secular bull market and a bear market rally is not merely one of duration but of kind. The difference is night and day.
One other way of characterizing a secular bull market is to observe that, at least since the establishment of the Fed a century ago, it is nearly synonymous with an expansion of the P/E multiple. And, that's where a problem arises.
Anomaly of the 2009 Bull
During the financial crisis, the P/E ratio rocketed from 22 in early 2008 to 124 in 2009 while the S&P was going from 1379 to 902 (using monthly data points). Of course, if you use CAPE, that sort of noise is tuned out, but considering the intimate relationship between the cyclical behavior of earnings, interest rates, and stocks and how that then plugs into the secular disposition of the market, I suspect that that bizarre leap in 2009 is telling us something.
So, for all of the similarities between the markets of the 2010s and the 1980s, there are three notable differences:
1. CAPE and P/E were in virtual lock step thirty years ago, but not in 2008-2009.
2. There was a cleaner end to the commodity bull market back in the '80s. In this market, commodities peaked in 2007-2008, but then many of them peaked again (especially agricultural commodities and precious metals) in 2011.
3. In 1980, P/E clearly bottomed, whereas in this instance, P/E appeared to bottom in 2011, well after the bull market in stocks had begun, i.e. in 2009, when P/E was in the triple-digit neighborhood.
The problem, from my point of view, is that we can make nice, clean demarcations of the initiation and conclusion of secular bull and bear markets in stocks and commodities (and, less so, bonds) using the P/E ratio and its long-term negative correlation with commodity prices, but the current bull market in stocks seems to have begun a good two years before the commodity bull truly died, and it began when P/E hit an all-time high. It is only since the eurozone crisis of 2011 that things have gone back to the post-World War II "normal," where stocks have tended to rally into falling interest rates and earnings, and commodity prices have corrected sharply.
In other words, we seemed to transition from a bear market rally (2009-2011) to a bull market P/E expansion (2011-present) without going through the pain of recession. (It should be said that pain has been felt elsewhere around the world, to be sure, and briefly in the stock market in the summer of 2011; the recession promised by the NBER never materialized, however).
I think that this is where the 1920s analogy fits in and helps put this market in some perspective. The sequencing is not identical, but that is not the point. It is the overall combination that is significant, but in the text below, I have tried to lay them out sequentially while presenting the combinations in the graphs.
The Roaring Twenties and the 2010s
In December 1916, earnings peaked after having tripled in two years (or ten times over two decades). Stocks, which had risen more slowly, also peaked.
In 2007, earnings peaked after their longest run of double-digit gains in a century.
In 1917, P/E bottomed at 5.3, and real commodity prices peaked. P/E then rose, and real commodities declined, although post-war inflation remained high, or rather real commodities declined because inflation remained high. (Nominal commodity prices remained at roughly their 1917 highs, but real prices were diminished by tremendous consumer price inflation up until 1920. Metals prices collapsed after 1916; oil prices seem to have peaked with inflation in 1920, according to annualized data). Stocks fell 30% by year end.
In 2006-2007, P/E bottomed out just under 17, and most commodity prices peaked in 2008, although metals peaked in 2006. Inflation peaked in 2008. Stocks peaked in 2007, and were down 40% over the course of 2008.
(Sources: Shiller and Pfaffenzeller)
(Sources: Shiller and World Bank)
In 1920, inflation hit 20% again; a severe recession soon followed.
Inflation peaked in 2008 at 5.6%; the Great Recession had already begun the year before, however.
In 1921, P/E went from 8.5 to 25.2 as earnings collapsed 64% (in the Depression, it took about three years for earnings to fall 75%). Although P/E was spiking, stock prices were bottoming. Commodity prices (in both real and nominal terms) bottomed. The rate of inflation fell to as low as -15%. Unemployment skyrocketed.
In 2008-2009, P/E went from 25 to 123, as earnings collapsed 85% year-on-year. Stock prices were bottoming, however. Commodity prices (notably, crude oil) also bottomed. Inflation went to as low as -2% in 2009. Unemployment skyrocketed.
(Sources: Shiller and Pfaffenzeller)
(Sources: Shiller and World Bank)
By early 1923, stocks were up 45% from their 1921 lows, but earnings had recovered much faster, driving P/E back down to 12; it bottomed at 8.6 later in the year. Inflation bounced back, peaking at 3.6%. Real commodity prices peaked and plateaued into 1925. Stocks, however, did take a tumble in the middle of the year, bottoming in October, while Europe was experiencing a grave political and economic crisis (recall Germany's hyperinflation and Hitler's Munich putsch). Earnings boomed throughout, although they began to decelerate in late 1923.
By early 2011, stocks had roughly doubled from their 2009 lows, but earnings had recovered much faster, driving P/E back down to 16, bottoming at 13.5 later in the year. Inflation bounced back, peaking at 3.9%. Commodity prices peaked again and have declined since. Stocks, however, briefly tumbled in the summer of 2011, as the Mediterranean (and Black Sea region?) began a severe economic and political crisis. Earnings growth began to decelerate.
(It was no picnic in other parts of the world, either. In a strange coincidence, Japan experienced a devastating earthquake and tsunami in 1923, and again in 2011).
(Source: Shiller and World Bank)
In late 1924, after a robust earnings recovery from 1921, earnings decelerated somewhat; stock prices accelerated in response.
In 2012, earnings declined about 0.5% over the course of the year. Late in the year, the stock market accelerated in response.
In late 1925, inflation peaked again, as did real commodity prices; from that point on, deflation would set in and not pause except briefly in late 1929/early 1930 (hitting a whopping 1.2%) and not stop until 1933. Earnings would decline about 10% in 1927, only for stocks to (again) accelerate in response and not look back until late 1929--and only, it must be emphasized -- after earnings growth was back to 20%.
Since 2011, P/E has gone from 13.5 to 18. "Lowflation" seems to have set in. P/E expansion in the 1920s didn't really get underway until 1926, doubling by September 1929.
One thing I left out of this comparison was interest rates. In 1921, long- and short-term rates would peak. Long-term rates would then slide until 1941, with a brief bump in 1928. Short-term rates (using annual data) bottomed in 1925, and then jolted upwards rather sharply in 1929, only to resume their slide into 1941. In a recent issue of The Economist, it said that in 1929, the Fed was stuck between the hard place of a runaway stock market and the rock of deflation and "made a catastrophic error" when it attempted to pop the stock market bubble. That seems to be the scenario that is unfolding in the present decade.
With the global recovery still remaining fragile, inflation low, and unemployment less than optimal, there does not seem to be much reason to expect that the Fed will raise rates anytime soon. But, what if the Fed were to face another 1929 scenario in 2019? Let's say, the S&P 500 hit 5000, roughly where we will be if earnings rise 50% and P/E doubles? Imagine, in the period between now and then, a foreign crisis that dampens earnings growth, sends investors scuttling back to Treasuries, pushes unemployment up just a little, leaves inflation at one-point-whatever, and GDP growth less than optimal. Stocks might get a nice 20% scare at some point (as in 1998 and in 2011, ever so briefly), but never experience a year-on-year decline in any month. In 2020, as the crisis conditions pass, things start to heat up on the "real" side (commodities, inflation, earnings, employment, growth), the Fed raises rates, and poof! There it all goes. I am not saying that is what is going to happen; I am saying that that is how these things happen, and that is the direction we are moving in.
The 1920s analogy suggests that, over the long term, the importance of Fed policy is overrated; it is only over the short-term that it is decisive, but long-term transitions tend to coincide with cyclical crises. Disinflation, falling interest rates, and falling commodity prices were already a fact well before the Depression. Economic and political chaos was already boiling in the Old World before it landed on the shores of the New in the 1930s.
But, really, what does this analogy tell us about what investors should do in the immediate future?
On the one hand, it says nothing at all. Just because things may have unfolded in a similar way ninety years ago does not mean that they will follow even a remotely similar course as the current decade unfolds. That is not really the point of the comparison, though.
The point is rather that we can look back at history to observe how variables historically relate to one another to help us as we weigh the arguments that are made today in favor of a given course of action. As I have argued before, we can wave off arguments based on bearish predictions based on cyclical declines in earnings, declines in profit margins, commodity weakness, and economic slack. Historically, these are bullish scenarios for stocks.
In a bull market like the one we are experiencing now, one would have to expect an inversion of the yield curve, an oil shock (OIL, USO, CRUD), a spike in gold (NYSEARCA:GLD) prices, a spike in inflation, or a surge in earnings or interest rates, and preferably some combination of those, in order to turn bearish in the current environment. Until at least some of those things begin to happen, we are likely to be surprised to the upside well into 2015 and, it seems, for a few years to come.
But, how sanguine can we be about this market if its most applicable precedent was the bubble that popped and caused the Great Depression? In order to answer that question, we need a better understanding of the relationship between commodities, inflation, and stocks. It is possible that the Depression was not caused by the popping of the stock market bubble in 1929 but rather the popping of the commodity bubble after World War I.