"It's tough to make predictions, especially about the future" - Yogi Berra
It's often tempting to look backwards as investors. People are lured in by the investment that's up 200% over the past three years, and want to keep away from the one that's down 50%, even if the latter is the better investment.
Perhaps this "backward-looking" tendency is part of human biology. We are hard-wired to segregate between what's "good" and what's "bad" based on our prior experiences. If we had a bad experience after buying a particular car, we want to avoid that brand in the future. Conversely, if we had a great experience with a product or service, we're more likely to trust that brand and go back for more. This is generally a good instinct to have; it's just that it doesn't always work that well with stock investing.
Buying a stock isn't quite like buying a washing machine or a refrigerator. It's true that we might care about the quality of management in the same way we'd care about the quality of a consumer appliance, but there are other factors with stocks. When we buy appliances, cars, or other major consumer goods, there's generally a competitive market, and all companies are fighting to keep prices down, and gain business. There's no equivalent with equity investment. To understand stocks, we must evaluate earnings, margins, cash flows, assets, and growth, within the context of a constantly-shifting price.
It's this constantly-shifting price that can be confusing to many people. Its true prices change with consumer goods, as well, but not in the same way. You won't find a TV that costs $500 today, $750 six months now, $900 in twelve months, and $350 in two years. With stocks and investment in general, however, prices can change quite rapidly.
What's more, to effectively value a stock, it requires a lot of research and knowledge. Many investors (even professional ones) either lack this knowledge or lack the will to do the research. It can be even more difficult to understand the "valuation" of an asset class such as housing or US equities (in general), given the vast multitude of factors involved. Given this, it should be no surprise that bubbles do form.
In this article, I want to examine the US stock market and try to predict the future returns for US equities over the next 5-10 years. We'll use the S&P 500 (NYSEARCA:SPY) as our proxy for this. My conclusion is that US equity returns are likely to be below-average over the next decade.
Past Returns and Future Performance
The investment cliché is that "past returns are not necessarily indicative of future performance." No period better exemplifies this than the mid- to late- 1990s.
An investor who passively bought into the S&P 500 on the first day of 1995 and sold out on the last day of 1998 would have achieved a spectacular 29.7% annualized return. If one had invested $100,000 at the beginning of this timeframe, he or she would have had $282,831 at the end of it. Not a bad four years, eh?
After those four years of otherworldly returns, more investors poured into the market. We were in the midst of a tech revolution, after all, and the prior four years of history showed how wealthy one could become by investing in the stock market. How could anything go wrong?
For those investors who bought in at the beginning of 1999, there was a little more upside ahead, as the S&P once again roared for a 21.5% return. Then, things started to go downhill quickly after that.
In 2000, the S&P saw a 5.8 loss, followed by even larger losses of 12.8% and 20.2% in 2001 and 2002, respectively. The three year trailing return on the S&P 500 was 26.2% in 1999 and quickly fell to -13.1% by 2002. If we were to examine the NASDAQ index or tech stocks, in general, we'd see an even more dramatic decline.
The high returns from 1995 to 1999 convinced more and more investors to jump into the market. These investors then proceeded to fall off a giant cliff. Even with the companies that experienced spectacular growth, investors still got burned because they paid too much for that growth.
We're accustomed to examining the stock market from the perspective of past (i.e. trailing) returns. Let's shift our perspective a bit and examine the 5-year and 10-year forward returns on the S&P 500. The two charts below will show you the total return based on the day invested. What you might quickly notice is that in many time periods, it's almost the reverse of the trailing return chart.
If one invested in the S&P 500 in January 1999, the 5-year forward return was -0.5% and the 10-year forward return was -2.0%. This is a far cry from the 29.7% annualized return over the prior four years, but it was the far more important metric to analyze at the time.
Here's the 10-year forward return chart, as well. Note that it cuts off in May 2004, since that was the last month I could conduct the analysis for.
Click to enlargeFrom this perspective, the attractiveness of the stock market bottomed in periods such as the late 1920's, the mid 60's, and the late 90's. Of course, this is all hindsight, but in many cases, it could have been reasonably foreseen by savvy investors.
How to Analyze Forward Returns
In an article last December, Ronald Surz provided a great formula for trying to predict future returns.
Forward Return = Dividend Yield + (1 + Earnings Growth Rate) * (1 + % of PE expansion or contraction) - 1.
This formula estimates forward returns by looking at earnings growth and P/E expansion or contraction. Of course, this requires us to set assumptions, and it's simple enough to be wrong on these. Nevertheless, there are a range of outcomes that, over a long period of time, are much more probable than others.
The trickiest part in that formula may be the P/E ratio bit. It's difficult to predict the future P/E ratio and we're best served by basing it on historical averages. Unfortunately, historical P/E ratios are a complicated subject.
The average P/E ratio of the S&P 500 since 1900 is in the 15x - 16x range. However, it has been significantly higher in recent decades, ranging from 17x - 20x since 1950, depending on where we begin the analysis. Though, there are major problems with the data.
For one, the more "abnormal" periods we see (i.e. bubbles and crashes), the less reliable the data becomes. Earnings crashes (such as the one created by the US Financial Crisis of 2008) particularly skew the data, since P/E ratios can skyrocket upward as a result. The Tech Bubble is also problematic, because the S&P 500 sold at historically high multiples, which was then followed by an earnings crash that created even higher multiples. So it's not clear whether the "average P/E" ratio since 1990 is that predicative of a measure. Instead, I veer towards trusting the 15x - 16x figure as a true "average P/E", but it's certainly a debatable issue.
The historical average earnings growth rate for the S&P 500 is a bit more straightforward. Since 1950, it seems to consistently come in around 6.0%. I tend to trust this number a bit more, because the P/E ratios seem to radically shift from era to era, whereas the earnings growth rates seem a bit more reliable since the mid 20th Century.
Running this formula for a 5-year investment horizon, I came up with the results below. I highlighted the regions that I would view as "most probable" outcomes with a 15x P/E and a 6% growth rate being right at the center of this. Using that scenario, we should expect forward returns on the S&P 500 to be a paltry 3.6% for the next five years.
Click to enlargeWhen you consider that the yield on the 10-year US Treasury bond (NYSEARCA:IEF) is about 2.6% and investment grade corporate bonds can yield 3.0% - 3.5%, it does appear that the stock market may be a bit overheated. In fact, to push our 3.6% expected return back up to a more reasonable 8%, the S&P 500 index would have to experience a 20% correction, falling back down to around 1,575.
There are also reasons why even the 3.6% expected return may be optimistic. We're in the midst of a multi-year earnings boom, with record high corporate profits. If we see a reversion to the mean, it's not unreasonable to expect an earnings growth rate closer to 2% - 5% over the next five years, suggesting forward returns on the S&P 500 could be even lower. With a 2% earnings growth rate and a 17 P/E ratio in 2019, the S&P would provide annualized returns of 2.3% over the next five years.
Let's also take a look at the 10-year chart. The longer we stretch out the timeframe, the more "normalized" the returns tend to be. Using our base case scenario (15x P/E and 6% growth), the 10-year forward return on the S&P 500 jumps a bit to 5.8%. This is bit higher, but still well below the historical averages, which have tended to be in the 10% - 11% range.
Click to enlarge
Still, no matter how you slice it, US equities (in the aggregate) do not appear that attractive right now, unless one has above-average growth expectations for the next 5-10 years. To achieve a 10% return (closer to the historical norm), we might have to see 8% earnings growth over the next 10 years and an 18 P/E ratio in 2024. It's possible, but seems unlikely.
Modeling the Past
It's worthwhile to run this formula for a few historical scenarios to see how our model would've looked at a few critical points. Let's go back to January 1999 and January 2007 and run our analysis.
In early 1999, it would have probably been fair to have expected an above-average earnings growth rate and above-average P/E ratio moving forward, given what was going on at the time. Assuming a 5% - 7% growth rate and a P/E in the range of 16x - 22x, you can see the range of outcomes (highlighted in green). The box closest to the actual outcome is highlighted in blue at -1.0%.
The actual outcome was 5.6% earnings growth. The P/E ratio also surprised a bit to the high end at 22.7x in January 2004. A more average P/E ratio of 18 with a 6% growth rate would've projected a much more dismal -4.9% return. However, you can clearly see that with the 1999 projections, it's very difficult to come up with many scenarios with positive forward returns, without relying on excessively high P/E ratios or growth rates. This chart was flashing a warning signal!
Now, let's jump into January 2007. With a P/E ratio of 17.36x, the stock market was not as obviously overvalued preceding the Financial Crisis as it was with the Tech Bubble Crash. Nevertheless, earnings growth did look a bit unsustainable and fueled too much by excessive leverage.
If we assumed a slightly subdued earnings growth rate of 3% - 4.5% and go with the historical average P/E ratio of 15, we would've projected below-average returns of 1.8% - 3.2% on the S&P 500. Even going with the average growth rate of 6%, we would have only projected a 4.7% return. For comparative purposes, the 10-year US Treasury bond also yielded 4.7% at the time, suggesting that treasuries were undervalued relative to US stocks.
Click to enlargeThe actual result came almost spot-on with the P/E ratio. However, earnings growth surprised to the downside at 1.5%, creating a five-year forward return of around 0.3%. While the end result was worse than expected, you can see that the forecast above was not particularly positive. The Jan 2007 chart does not have the same sea of negative returns as the Jan 1999 chart, but it does suggest below-average returns, unless one were predicting a shift to even-higher growth from 2007 - 2012.
There are always a lot of unknowns in investing. It's important to ignore the noise of past returns, and instead focus on risk and reward moving forward. Once we examine the US stock market from that perspective, we see a picture that suggests that it's going to be difficult to achieve strong returns over the next 5 years.
Realistically speaking, in order to approach historical averages for the S&P 500 (around 10% returns), we'd likely have to see earnings growth around 8%+. That's a pretty high hurdle to climb given the strong growth we've already seen over the past five years. More realistic earnings growth scenarios in the 3% - 6% range suggest S&P returns of about 1% - 6% over the next five years. This suggests the stock market is somewhat overvalued in the aggregate. The S&P would need to fall about 20% to hit an 8% expected return over the next five years.
My advice is to tread cautiously. While the market is certainly not as overheated as it was in 1999, there's still a very strong case to be made that equities, on the whole, are somewhat overvalued. This, however, does not mean there are not attractive stocks out there; it merely means that they are more difficult to find.
Overall, I view now as a good time for a defensive posture towards the market. Find some reasonably priced dividend-paying blue-chips and hold a bit of cash on the side to cushion against a market correction. Don't worry about "chasing returns."
Disclosure: The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it. The author has no business relationship with any company whose stock is mentioned in this article.