Preview

In this issue, I explore the concept of asset class rotation as a long-term investing thesis. Most of the newsletter issues so far have described specific asset classes as being favorably or unfavorably priced relative to fundamental value. Below I try to present the broader theory that I have in mind for long-term investing strategies.

Long-term asset class rotation means adjusting one's financial portfolio every 2-5 years in response to broad trends that determine which investment classes are expensive or cheap. Fundamentally, I believe that this strategy is effective because investors tend to overpay for following conventional investment advice and underpay for being a contrarian investor.

This framework is useful for understanding why U.S. Stocks and U.S. Real Estate are in long-term bear markets and why gold and commodities are in long-term bull markets.

Asset class rotation

First let me emphasize the distinction between investing within asset classes versus investing across asset classes. By asset class, I mean any broadly defined group of similar investments, such as stocks, bonds, or real estate. Investing within an asset class means picking specific stocks or real estate investments that will outperform the market, whereas investing across asset classes means picking types of assets that will outperform other types of assets.

In this newsletter issue, I focus on long-term differences across asset classes over time. I hope to convince you that asset classes go through long-term cycles of underperformance and outperformance, and that a smart investor can benefit from paying attention to these cycles.

Wait, but what about the efficient market hypothesis that says prices should reflect fundamental value at all times? The academic consensus regarding this question is best summed up in the following quotation by economist Kenneth French:

In "the old days," many people confused the efficient market hypothesis with the idea that the stock market follows a random walk. Then, in 1988, Gene Fama, my colleague at the University of Chicago, and I showed that long-term market returns are surprisingly predictable, and that simple measures like the market's dividend yield and earnings price ratio predict much of the variation in one to five-year stock returns. We then applied the same logic to the performance of individual stocks. Using variables such as the book-to-market ratio and the earnings price ratio to sort firms, we found that value stocks tend to have higher average returns than growth stocks.

-- Kenneth French, Dartmouth University

In other words, markets are *sort of* efficient. It depends a lot on the time interval of interest and how one defines efficiency. As French points out, much of the variation in long-term stock returns can be predicted by a few simple measures of value. This would not be true in a truly efficient market where price always reflects fundamental value.

Instead, I like to think of the market price as a reflection of broad investor sentiment and investment convention. Sometimes the price is too high, and sometimes the price is too low. Over time, the market price fluctuates around fundamental value depending on whether the investment is popular or in-style. Just as you would overpay for a Tickle-Me Elmo doll at the peak of a Christmas frenzy, you might overpay for a technology stock at the top of the Internet Bubble.

It takes years and sometimes decades for financial institutions to develop and for an asset class to become part of the conventional investment wisdom. The paradox is that by the time an asset class becomes widely accepted by investors, the price is already too high to produce an adequate return.

The Internet Bubble is an extreme example, but it fits into the framework of long-term asset class rotation. Technology stocks were cheap in the early 1990s because financial institutions were not pushing these stocks and very few investors owned any. By the late 1990s, it had become widely accepted that investors could get wealthy by buying into the Nasdaq. At that point, however, the prices were already too high. With the benefit of hindsight, we know that a much better time to buy U.S. stocks would have been when Businessweek published its famous "The Death of Equities" article in 1979 after stocks had spent almost 15 years underperforming inflation.

You've got to think to yourself, "if everybody else around me is investing in asset class X, doesn't that mean I'm probably overpaying?" I would say that the answer is generally yes. But generations go by and memories fade, and the same asset classes move in and out of favor over the years. Understanding this framework of long-term asset class rotation in mind is a good tool for analyzing broad market trends over time.

Historical performance of various asset classes

Below I examine the history of some of the major asset classes, starting with U.S. stocks. For an overview of the cyclicality of the U.S. stock market, please see the March newsletter.

In the chart below, I present S&P/500 returns from 1900-2008 using data from Robert Shiller of Yale University. The graph shows the total return including dividends adjusted for inflation. The thin line presents a 2-year return and the thick line presents a 10-year moving average.

Looking at the 2-year return line, it's hard to see any real trend. But once the average is extended to 10-years, one can easily see the long-term cyclicality of the market. My interpretation of this graph is that the market will continue to produce less than average returns in the near future.

Now let's move onto Bonds as an asset class. As with stocks, there is substantial variation in performance within this asset class, but on the whole, bonds do well when interest rates are falling and inflation is low. I won't go into all the details, but the chart below bases returns completely on the path of long-term interest rates. Once again, all returns are adjusted for inflation.

Bonds did very well in the 1920s/1930s as well as the 1980s/1990s, during falling interest rates with lower than average inflation. As I argued in the April newsletter, I do not expect bonds and cash equivalents to do well in the current low interest-rate environment.

Real estate is an asset class that is well known for boom and bust cycles, but the cycles occur with slightly higher frequency. For this reason, I display the 5-year moving average for real estate.

The real estate data comes from Robert Shiller's NYTimes chart of historical home prices, which I use to generate annual returns adjusted for inflation. Note that real estate appreciates roughly in line with inflation over time. The recent price boom was certainly an anomaly and will lead to a correction in inflation-adjusted terms.

The next asset class (if you want to call it that) is gold. Gold is interesting because it tends to be counter-cyclical to the stock market. But this should make some sense. Why would anyone recommend gold when stocks are returning 10% a year? However, gold can be so undervalued during a stock market boom that it actually becomes worthwhile to hold it during a stock market bust cycle.

Commodities have a similar cyclicality as gold. Commodities do well when cash and paper assets do poorly. In the longest of horizons, holding commodities and gold is certainly worse than holding stocks, but one can do quite well holding commodity producing companies during the boom cycle. The chart below displays the inflation-adjusted returns for the CRB Index, a broad-based index of commodity prices.

Note that the absolute level of returns is not directly comparable from one asset class to another. For example, the real estate returns have not included rents collected from owning property. In the image below, I "normalize" all returns by asset class into the "Best" and "Worst" 10-year period.

Click on the image above for a sharper version

I have mixed feelings about how many more years the gold/commodities bull market will continue. Certainly some commodities have become overbought in the near-term. But I'm more positive on hard assets than stocks, bonds, and real estate because they spent so many years in a bear market before the year 2000. However, no asset class is a screaming buy right now since we seem to be transitioning between extremes. I think perhaps international stocks are another good alternative - more on this in a future newsletter issue.

The contrarian in all of us

As much as we would all like to be contrarian investors, it's not as easy as it sounds. I often find myself reading smart arguments on both sides of the issue, and the "correct" contrarian viewpoint is never so obvious. Is it contrarian for financial companies to be investing in subprime mortgages right now? It's hard to say. It is difficult to predict just how bad sentiment will get before going against the conventional wisdom makes sense.

When evaluating investment advice, you should try to figure out where the provider of such advice fits into the investor distribution. Are they constantly reporting about investment trends after they happen? Or do they look ahead and try to figure out what the future holds? Over time, I hope to end up in the early portion of the investor distribution, which by definition is limited to a small fraction of the population.

Click on the image above for a sharper version

There is such a thing as being too early with an investment, but you certainly don't want to be too late, especially in the bubble environment we have experienced in the last two decades. Long-term asset class rotation may not help much in figuring out where the next boom will occur, but at least it can help you avoid getting caught up in the current convention.

Conclusion

Markets will always contain a high degree of unpredictability. I don't claim to know the future of anything with certainty, but historical patterns of performance cyclicality between asset classes can be a useful way to think about investments. I wouldn't rule out the possibility that broad U.S. stock investing moves back into the investing spotlight, but I'm not willing to bet on it right now.

Experts in a particular asset class will continue to do well in any environment based on within asset class performance. Successful stock pickers, for example, will always find quality stocks, even in a poor overall stock market outlook. But as a general rule, I'd rather be finding value in those asset classes which are currently not part of the conventional investment advice.