I recently came across a chart on Twitter titled "The Most Interesting Chart on Investing You'll Ever See." So naturally I had to hit that clickbait. But you know what, it was pretty interesting. Maybe not the most interesting chart, but still pretty interesting, despite being a little out of date and with some apparent mistakes. So of course I had to recreate it myself — which I did below (click to enlarge):
My version of the chart shows all of the 15%+ streaks in the S&P 500 Index^a (a good proxy for the US stock market) since WWII, based on monthly (the 1st of each month) values of the index, not including dividends. It also adds the cyclically adjusted price-to-earnings ratio (CAPE), a valuation metric I find helpful for forecasting long-term investing returns, at the beginning and the end of each streak in the index. Finally, my chart shows correct figures for the longest and largest streak, the 114 month, 376% run from October 1990 to April 2000, during which the CAPE expanded from 14.8x to 43.5x. The original "most interesting" chart, made by Ric Edelman of Edelman Financial Services, incorrectly shows this particular up streak going up only 155%.
With that out of the way, below are eight interesting takeaways I make from my version of this chart, takeaways that may give clues about what to expect from the US stock market going forward:
1. Markets usually go up
This was Edelman's main takeaway. The US market has usually gone up since WWII because of inflation, population growth, expanding access to foreign markets, productivity growth, and since the early 1980s, declining interest rates. However going forward, it would appear that some of these factors will be much less helpful than they had been over the previous 70 years.
Population growth will be slower due to demographics (smaller percentage of women of childbearing age), declining fertility rates, and in the near term due to the Trump administration's effort to lower both legal, and illegal, immigration.
The Trump administration may also reverse the post-WWII trend of US companies gaining more access to foreign markets and consumers. It pulled the US out of the Trans-Pacific Partnership (TPP), put negotiations with the European Union over the proposed Transatlantic Trade and Investment Partnership (TTIP) on ice, threatened to pull the US out of the North American Free Trade Agreement (NAFTA) and the US-Korea Free Trade Agreement (KORUS FTA) if the US does not get more favorable terms, and just this week applied stiff tariffs on imported solar panels (targeting Chinese companies) and washing machines (Korean companies) while pledging there will be more tariffs on foreign manufactured goods to follow.
Interest rates are likely to go from a tailwind to neutral, or even a headwind, for valuations of US companies. Declining interest rates drive up company valuations as they increase the present value of all future cash flows generated by companies. Declining interest rates also make bonds less attractive relative to equities, thus driving more incremental buying demand for equities. But when rates rise, the opposite occurs. With not much more room to decline, and new upward pressures from a strong economy, tax cuts, and a Federal Reserve that is raising short-term rates, it would appear that the tailwind from declining interest rates is gone.
2. Recent up market streaks have been much longer than down market streaks — but this wasn't always the case
There have been five up streaks that lasted 60 months or longer since WWII, and four of them occurred since 1982. We think that this has a lot to do with the long, steady decline in interest rates since 1981 that were discussed above.
With interest rates unlikely to continue declining from today's low level, we think the US market may go back to the shorter and smaller market streaks that were common from the late 1950s through till the early 1980s.
3. The current bull market is one of the longest, and strongest, since WWII
The current bull market is in its 107th month. Over this time (1MAR09-22JAN18), the value of the S&P 500 increased 270%. Including reinvested dividends, the S&P 500 is up 378% (19.2% annualized). There has only been one bull market since WWII that was longer and stronger — the 1990s run that was culminated by overenthusiastic euphoria over tech, media, and telecom (TMT) stocks.
On the back of the Trump administration's tax cuts and deregulation, a majority of market prognosticators are predicting that the current bull market will extend longer than the 1990s bull market (only seven months to go). And some believe the current bull market will become the strongest — only a further 29% increase is needed.
4. The higher the starting valuation, the larger the eventual decline — generally speaking
This seems logical, but it's not readily visible from my the chart — so I pulled out the 12 major down stretches in the S&P 500 since WWII, and plotted their starting CAPEs against their price declines in the chart below. Although the data isn't particularly tight (R2=0.50), there is a definite relationship between greater CAPE valuations resulting in greater price declines.
A related correlation, but one that is shown by my version of the "most interesting" chart, is the relationship between the magnitude of the price decline and the magnitude of the price increase in the immediately prior up stretch.
The obvious implication from the above mentioned correlations is that when the next major down market commences, it may be a particularly large one given today's elevated valuations (34.3x CAPE) and the large price increase in the current bull market (+270%).
5. Up markets appear larger than down markets — but the chart deceives
The green squiggly lines go higher than the red squiggly lines go down. So of course "this chart clearly shows that when stock prices are rising, they rise a lot and for a long time. When prices fall, they fall a little and for a short period of time" as Edelman states in his piece about the "most interesting" chart. Well, not so fast there Edelman. Comparing up and down percentages, which the chart shows, can be misleading. For example, a 50% decrease is not "equivalent" to 50% increase. A 50% decrease is a lot more damaging and painful than a 50% increase is profitable and pleasureful. Let me explain.
First, there's the simple math. Let's say we start with $10,000, and we invest it in an S&P 500 exchange-traded fund (ETF).b Our timing is horrible and we soon suffer a 50% decline — leaving us with a $5,000 investment. But we're patient and hold on long enough see the investment bounce back 50%. We're back to where we started, right? No. We're up to a $7,500 investment, still another 33% increase shy of getting back to breakeven. Or said another way, if you suffer a 50% decline, you need a subsequent 100% increase just to get back to breakeven. And it works the same if the events are reversed — if you are initially up 100%, and then from there you suffer a 50% decline, you are back to where you started.
Second, there's the psychological damage inflicted by a 50% decline, which is much larger in magnitude than the good feelings caused by a 50% increase. This effect, called loss aversion, is what causes many investors to bail out at the bottom, missing the bounce back entirely — and it's discovery is what helped land Daniel Kahneman a Nobel in Economic Sciences in 2002.
6. Markets always come back after big drops
Yes, but you would be amazed at how many investors, both amateur and professional, forget this in the eye of storm that is a bear market. As a species, we tend to act more like herd animals than lone wolves in times of stress. I saw it firsthand in 2008/9 during the last recession, in 2001/2 after the TMT bubble burst, and in emerging markets during the 1998 Ruble and the 1997 Asian currency crises. The same can be said about human behavior during the late stages of bull markets.
7. Market streaks are driven more by changes in earnings multiples than by changes in earnings themselves
You might have read that earnings drive stock prices. While this is partly true, a much more important driver of stock prices is what investors are willing to pay for a dollar of earnings — the earnings multiple (i.e. price-to-earnings multiple).
Market price increases, and decreases, over a period of time can be calculated by multiplying the growth/(contraction) in earnings with the growth/(contraction) in the multiple that investors are willing to pay for those earnings:
(1 + price change) = (1 + cyclically adjusted earnings change) x (1 + CAPE multiple change)
For example, the current bull market's 270% increase can be broken down into the 157% increase in the cyclically adjusted earnings multiple (34.3x/13.3x) and the 45% increase in cyclically adjusted earnings over the last 107 months:
(1+2.70) = (1+1.57) x (1+0.45) [off slightly due to rounding]
So let's look at these two components in the S&P 500's 26 major price streaks since WWII. In 24 of these 26 streaks, it was the change in earnings multiple that was the larger factor in the index price movement. Only two streaks were predominately driven by changes in earnings.
Fundamentally, the earnings multiple is directly linked to expectations for future long-term earnings growth. The multiple I am willing to pay on a dollar of today's earnings is directly linked to the expected long-term growth of those earnings. The higher the expected long-term earnings growth, the higher the multiple I'm willing to pay. And visa versa. Which leads us to #8.
8. We're paying 2.6x more for a dollar of cyclically adjusted earnings now than we were back in March 2009
One would think that earnings multiples would be fairly stable over time, especially in the context of cyclically adjusted earnings which smooth out the short-term volatility in annual earnings by averaging the last 10 years of earnings (enough time to usually catch most portions of an economic cycle). There's also the stabilizing effect of a diversified index such as the S&P 500 (growth propects may go up for some companies/industries, but they are likely to go down for others). But as the previous chart showed, earnings multiples swing wildly over time.
Below is a chart of the how the CAPE for the S&P 500 has fluctuated since WWII, with periods of upward price streaks in green, and downward price streaks in red.
Nine years ago investors were willing to pay $13.30 for each dollar of cyclically adjusted S&P 500 earnings. Today investors are willing to pay $34.30. Did long-term (decades) earnings growth forecasts for US companies go up 2.6x? Unlikely. The gap between long-term growth forecasts and earnings multiples comes mostly from emotions. Irrational emotions. It is these emotions that we try to take advantage of as active investors.
What strikes you about the "most interesting" chart, or any of the other derivative charts that I presented above? Let me know in the comments section below.