Editor's note: Seeking Alpha is proud to welcome David Rimkus as a new contributor. It's easy to become a Seeking Alpha contributor and earn money for your best investment ideas. Active contributors also get free access to the SA PRO archive. Click here to find out more »
The ten-year outlook for stocks looks bleak. Investors continue to pour money into U.S. equities, driving the Average Equity Share (average investor percentage allocation to stocks vs. bonds and cash) up to levels that have proven to be dangerous in the past. There is a clear inverse relationship between Average Equity Share and forward-looking U.S. stock market nominal and real returns. At this moment, the risk in U.S. equities appears to vastly outweigh the potential rewards to be gained during whatever time we have remaining in the second longest bull run in history.
Recently, I stumbled upon one of the most interesting articles I have read regarding market behavior and market indicators. The “Average Equity Share to U.S. Market Return” ratio described in this piece is the most meaningful market indicator I have encountered. The regression in the analysis totes an R-squared of 0.913. In other words, the variation in the x variable explains 91.3% of the variation in the y variable, which is incredibly significant. The x variable is Average Equity Share, or the percentage share of the average investor’s portfolio (all investors) that is allocated to stocks. The y variable is the elusive expected market return figure that keeps much of Wall Street in business. More specifically, it is the ten-year forward average return of the S&P 500. Because of the extraordinarily high R-squared value, there were questions lingering in my mind from the analysis:
- Why does the analysis begin in 1952?
- Where is the Average Equity Share level today, and how has it tracked since 2013?
- Does the analysis still hold when adjusting for inflation?
- How does this regression look when using the CAGR of the S&P, rather than average annual return?
The dataset covered 60+ years of market data and numerous market cycles, but I was concerned whether data mining played a factor in determining the start date. When first compiling my data, I noticed that the quarterly FRED (Federal Reserve Economic Data) used in the original analysis only went back to October 1951. After doing some research, I was unable to find this data freely available anywhere else online, which derailed the data mining suspicion.
I attempted to mimic the analysis using data that I gathered, including five additional years of data that have since surfaced (through April of 2018). Upon plotting the data, I discovered that the R-squared of my regression of 0.815 was still very much significant, although lower than 0.913.
Upon using virtually the same time frame (January 1952-October 2013), my R-squared was closer to 0.831. Perhaps the difference between the 0.913 and 0.831 lies in the market return metric used. I chose to use the ten-year compound annual growth rate (CAGR), which is more precise than using the average return in calculating expected performance. (Quick exercise: A $1 investment increases by 50% one year then decreases by 50% over the next. Your original investment is now worth $0.75. Your CAGR is -13.4%, and your average return is 0%. Which metric would you rather use?) While the 0.831 statistic is, again, lower than the 0.913, this reading is still statistically significant. When expanding the analysis to include the most continuous data available (October 1951-April 2018), my R-squared was closer to 0.815. The variability in Average Equity Share has strong explanatory power regarding the variability in the nominal ten-year forward CAGR of the S&P 500 (81.5%); this is an inverse relationship.
The above chart plots the quarterly data points of Average Equity Share on the x-axis and the nominal (non-inflation adjusted) CAGR of the S&P 500 in the subsequent ten years. The black dotted line is where the most recent estimate of Average Equity Share registered (April 2018). Since 1951, the highest ten-year forward nominal CAGR with an Average Equity Share above 44.5% was 3.7%. Keep in mind, this is before factoring in inflation. The linear relationship of the two variables can be clearly illustrated below.
By adjusting for inflation, you can get a better idea of the spending power gained over the subsequent ten years.
The R-squared of the new regression drops to 0.558 after adjusting historical S&P pricing, using the CPI data from FRED (Consumer Price Index for All Urban Consumers: All Items, Index 1982-1984=100, Monthly, Seasonally Adjusted). The plot of data clearly becomes looser, indicative of inflation’s effect on real market returns. Inflation only moves return in one direction, but the extent of the downward shift has been more drastic at different points in time (more on that later).
The second y-axis on the line graph below was adjusted 3% downward (more negative) to imitate a similar relationship to the nominal graph above (both were roughly eye-balled for the best fit).
As shown, there is more of a tendency to underperform relative to Average Equity Share than overperform, particularly in the 1970s. The ten-year forward CAGR becomes more volatile due to differing economic policies causing vastly different inflationary periods in the past. However, this is to be expected. The bottom line is this: inflation will not do any favors for your expected return. Using the regression shown at the beginning of the article (not adjusted for inflation) with the April 2018 Average Equity Share of 44.5%, the predicted forward ten-year CAGR of the S&P is 0.86%. Factoring in inflation, the expected real return would be -2.67%.
As a counterpoint, perhaps there has been a long-term or permanent shift in the Average Equity Share upwards. In other words, perhaps a higher Average Equity Share is less indicative of a lower ten-year CAGR than it used to be. After all, trading is more accessible than it has ever been. Markets are much more efficient, and trades are executed within fractions of seconds. You, quite literally, can trade at any point of the day with the stroke of a finger, so long as the markets are open. Even during the worst financial meltdown in recent memory, the Average Equity Share did not stoop to the previous lows of the 70s and 80s. It also recovered with incredible speed. If there is merit to this explanation, how much of the variance between these two lines would it explain? If the span in the first chart from ’03 to ’07 is an indicator, it would suggest two things: a) that this shift (higher investor preference toward stocks) occurred after ’03; and b) that the shift has only accounted for a difference in the expected return of about 2%. However, I would tend to think the data will revert to the mean.
The biggest flaw in this indicator is its ability to predict returns over the next five years, or even over the next two years. It is important to use your own judgment and to evaluate whether the risks are worth the reward in the current market environment. Even if this indicator was 100% accurate, it does not spell doom for each of the next ten years. The market tends to correct itself in brief, abrupt maneuvers. It is worth noting that the last couple of times that the Average Equity Share reached the levels of today, it immediately preceded recessions. Further, the Average Equity Share topped at 42.0% shortly before the Great Recession in 2008, which is 2.5% lower than the most recent FRED data point.
While there is still value to be found in U.S. equities, it becomes increasingly difficult to come across as the equity market becomes more saturated. If history is any indicator of what is to come and this relationship continues to hold, there are likely far more enticing entry points lying ahead. While many investors continue to try and squeeze as much return out of this latest bull run as possible, I predict that the true winners will be those who took caution when the warning signs were flashing red.
For more recent regression results:
Since January 1987: Nominal R2 - 0.894, Apr’18 expected nominal 10 yr. fwd. CAGR: 0.06%
Since January 1987: Real R2 - 0.877, Apr’18 expected real 10 yr. fwd. CAGR: -4.89%
Thank you for reading. Your thoughts and/or criticism are welcome below.
Disclosure: I am/we are short SPY. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: Currently buying continual out of the money puts on SPY as hedging instruments for my overall portfolio.