### Investment Thesis

By examining the real returns of the S&P 500 (SPY) (IVV) (VOO) and the P/E 10 ratios of the index at various points throughout history, we can determine that high valuations have a negative impact on future long-term returns. Despite this, 30-year returns on the S&P 500 are surprisingly steady, so if you have a long enough investment horizon, it is correct to continue investing in an expensive stock market. If your investment horizon is 10 years or shorter, though, you might wish to moderate your investing.

### Overview

It is hard to miss recent warnings from academics and investing professionals alike about how expensive the stock market has become, how there may be a bubble, or how a crash might be imminent. As long-term investors, how much weight should we put on these dire predictions?

I am 41 years old, have been working in the same industry for 20 years and I hope that I have at least another 20 years of work ahead of me. I have two pools of money that are invested in the stock market. One pool of money is in IRA accounts. At the age of 70 1/2, I'll be forced to start withdrawing from these and I hope to put it off until then, so I consider that money to have at least a 30-year horizon. The other pool of money is in taxable accounts. In an ideal world I would not touch that money for 30 years as well, but I have to take into consideration the possibility that I might lose my job, have a serious medical expense, or experience some other large event that alters my financial situation. I decided to place a more conservative 10-year horizon on the taxable pool of money.

10-year and 30-year investment horizons are a luxury that individual investors have that investing professionals and corporations rarely do. They are judged on quarterly and annual results, and that is the focus of most of their research and prognostications. So as I tried to investigate what impact our expensive stock market had on future returns, I had a difficult time finding answers.

Vanguard put out a paper in 2012 detailing their findings on predicting future returns, and the research has been replicated elsewhere. They came to three conclusions:

First, stock returns are essentially unpredictable at short horizons.

Their research indicates that the correlations of most metrics with 1 year forward returns was close to zero.

Second, many commonly cited signals have had very weak and erratic correlations with realized future returns even at long investment horizons.

Poor predictors of long-term returns included, but were not limited to, dividend yield, economic growth, corporate profit margins, and past stock returns.

Our third primary finding is that valuation indicators-P/E ratios, in particular-have shown some modest historical ability to forecast long-run returns.

Their findings were that while there was no clear winner between the various P/E ratios, the valuation metric explains about 40% of future 10-year returns.

P/E 10 is a metric championed by Robert Shiller, also known as cyclically adjusted P/E (or CAPE) or simply as Shiller P/E. Instead of looking at the past 12 months of earnings like the traditional P/E metric, or the estimated future 12 months of earnings like the forward P/E metric, it looks at the last 10 years of earnings in an attempt to smooth out the cyclical ups and downs of the market.

While P/E 10, like pretty much everything else, turns out to be a terrible predictor of short-term results, it is a reasonable predictor of long-term results. Since long-term results were what I cared about and because Shiller hosts a website with easy to access data on the metric going back to 1871, I decided to use it.

The S&P 500 is not a perfect proxy for the US stock market, or for the world's stock markets. That said, it is the most replicated and researched index in history. Because of this, I decided to examine the impact of high P/E 10 ratios on the future returns of the S&P 500 to help me decide whether to keep investing in an expensive stock market. To aid in this, I used an S&P 500 return calculator that allowed me to view returns of the index over different time periods with dividend reinvestment and while taking inflation into account.

The P/E 10 of the S&P 500 reached 30.49 at the start of August, which is over two standard deviations above the long-term average of 16.78. Should I continue adding money each year to my stock investments in my IRA? Should I sell the investments in my taxable accounts?

### 30-Year Real Returns

I examined 107 rolling periods of 30 calendar years beginning in 1881 (the first year that Shiller begins providing P/E 10 ratios) through 1987 (the last date for which we have full visibility of the future 30-year returns). What I discovered was truly remarkable. Keep in mind that I looked at annualized real returns for the S&P 500, returns that include dividend reinvestment and adjustments for inflation. This is what I care about and what most investors should care about, not nominal returns.

- Average annualized returns: 6.47%
- Median annualized returns: 6.43%
- Standard deviation: 1.67%
- Highest annualized return: 10.24%
- Lowest annualized return: 3.24%

The mind-blowing part of this for me was that not only has there never been a 30-year period with negative returns, but the lowest annualized return was 3.24% for the period beginning in 1892 and ending in 1922. This period of time includes the Panic of 1893, the Panic of 1896, the Panic of 1901, and the Panic of 1907. Even with this especially active time for economic depressions and market crashes, investors with a 30-year horizon could have earned three and a quarter percent per year above inflation if they remained invested.

There is too little data to draw a conclusion about instances of P/E 10 that are two standard deviations above normal, so I decided to set the bar lower. I looked at the 15 of 107 rolling periods that began with P/E 10s of greater than 20, and in these years, the subsequent 30-year real returns averaged 5.62% on an annualized basis.

Similarly, when looking at the 18 of 107 rolling periods that began with P/E 10 of less than 10, we see 30-year real annualized returns averaging 7.97%. While increased or reduced returns are not fully explained by P/E 10, we can see the correlation between high P/E and lower long-term returns, as well as low P/E and higher long-term returns.

What this tells me is that I should expect lower than average 30-year returns for money that I invest in my IRA today, but that I should expect positive returns. I have no way of knowing precisely what those returns will be, but it is pretty clear to me that I should keep adding to my IRA each year and keep investing that money in the stock market regardless of valuation. The only changes that I intend to make are to moderate my expectations for the growth of those investments, and to consider directing more of my investments to stock markets with lower valuations, like non-US developed markets and emerging markets especially.

Could there be a once in 200 year event that produces negative real returns over the next 30 years? Certainly.

Would it be wise to bet that this event will occur in my lifetime? Certainly not.

### 10-Year Real Returns

With the question of what to do with the pool of money in my IRA accounts solved, I turned to the money invested in taxable accounts. While my IRA investments are designed to give me exposure to the total market (mostly index funds, but also one active emerging markets fund), the investments in my taxable accounts are in individual securities. Even though the stocks that I have chosen to invest in have lower P/E ratios on average than the broader market, they are still elevated by historical standards. They look attractively valued in comparison to the rich market, but the rich market still impacts them.

I examined 127 rolling periods of ten calendar years beginning from 1881 to 2007. While the average ten-year annualized returns for the S&P 500 were nearly identical to the average thirty-year annualized returns, as one would expect, the dispersion of results was much larger.

- Average annualized returns: 6.45%
- Median annualized returns: 6.15%
- Standard deviation: 5.19%
- Highest annualized return: 18.33%
- Lowest annualized return: -4.41%

These numbers were far less encouraging than the 30-year returns. 17 out of the 127 ten-year periods (13.4%) had negative real returns, with the lowest return occurring if you invested at the start of 1999 and caught both the dot-com bubble bursting and the 2008 financial crisis.

If we examine the 30 of 127 periods that began with P/E 10 ratios of greater than 20, we get a ten-year average annualized real return of only 2.50%. 9 of these 30 periods (30.0%) resulted in negative real returns.

If we examine the 18 of 127 periods that began with P/E 10 ratios of less than 10, we get a ten-year average annualized real return of 10.56%. 0 of these 18 time periods resulted in negative real returns.

While the decision to remain fully invested in my IRA accounts and to continue contributing annually was an easy one to make, the decision of what to do with my taxable accounts was not as clear cut. Complicating this decision is the large uncertainty surrounding that pool of money. While I'd like to keep that money invested for 30 years, a major life event might require me to access it much sooner, whether it be in 20 years, 10 years, or in 3 years. In addition, there is probably some correlation between a global or US economic crisis and the loss of a job, one of the events that might necessitate me accessing this money earlier in than expected.

On the other hand, instead of investing this money in the broader market, I have it invested in a dozen or so individual stocks. Only a few are trading at P/E 10 levels of 30 or above, in line with the S&P 500. The majority trade at much lower valuations. Would this insulate me from poor returns over the next 10 years? It is hard to say with any certainty.

In the end, the best plan of action with investing is often simply to stay the course. It is pretty clear from my research of P/E 10 ratios and the subsequent 10-year and 30-year returns of the S&P 500 that valuation matters, but that even in times of high valuations, betting on the stock market is a good bet if you have a long enough investment horizon. While there is no guarantee that future results will mirror past results, our investing decisions should be informed by them.

I had decided to continue contributing as much as possible to my IRA and investing those contributions in the stock market. As for the taxable pool of investments, I'm not going to take any money off the table. Even though I can expect relatively low returns over the next ten years and there is the very real chance of negative returns, the expected return is likely still positive.

Instead of adding additional taxable investments at these valuations, however, I plan on building up a bigger cash cushion. It seems to me that the way to maximize the returns of my taxable investments at this point in time is to lengthen the investment horizon. If I can reduce the likelihood that I'll need to tap into those investments and push out that time horizon from five or ten years to twenty or thirty years, then I can dramatically increase my odds of coming out ahead.

### Conclusion

Despite the historically high valuation of the US stock market and the unpredictable nature of returns over the next one to three years, it remains correct to invest in the stock market if you have a long investment horizon. With a 30-year horizon, the stock market is still a very safe bet. With a 10-year horizon, it is a dicier proposition investing at historically high P/E levels because of lower expected returns and higher variance, but returns are still likely to be positive on average.

**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 (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.