Wednesday, February 25, 2009 1 comments ++[ CLICK TO COMMENT ]++

Secular stock market cycles & Ten year rolling returns

Crestmont Research publishes some excellent macro-oriented documents on their website. I typically hit their site once or twice an year and, I have said this before but, I'll reiterate my recommendation of their 100 year S&P stock market return matrix. The best thing is that all this is free :) If you are macro-oriented, I recommend reading all the documents on their site. One may not necessarily agree with their thinking but it covers the main bases of macroeconomics pertaining to investing.

Given how we are going through a nasty bear market, I pulled some tables/charts that are highly relevant right now. One of them deals with secular bull and bear cycles, while the other is about 10 year rolling returns.

Secular Bull & Bear Markets

The table below, extracted from Secular Bull & Bear Markets presents secular bull and bear markets (as defined by Crestmont) over the last 108 years.

Everyone has their own way of defining bull and bear markets and nothing is ever perfect. This table uses the DJIA average, which some consider seriously flawed and not worthy of study, and excludes dividends, which tends to be a big component of total return. In any case, all this is just a rough reference and it's good enough for me.

I always like looking at this because it presents all the key information--DJIA return, P/E values, price inflation, starting/ending DJIA level--in one easy to analyze table.

You can easily see the worst bear market in history (1929 to 1932) which resulted in the Dow dropping from 300 to 60! Yikes! Another disastrous one, possibly the 2nd worst bear market of all time, is the 1901 to 1920 (I believe this bear market started in the late 1800's but this table only starts in 1900.) In this bear market, the Dow started at a level of 71 and went to 72. In 20 years! Yes, the price index didn't go anywhere for 20 years, although dividend returns, which aren't captured in this table, would have yielded some return (even if you included dividends, real returns would have been horrible given the high inflation in the late 1910's due to WWI--you can see the high price inflation numbers in the table.)

The longest bull market is from 1942 to 1962--the post-war boom. The bull market with the most cumulative return is the one from 1982 to 1999. However, unknown to most people, is the #1 bull market of all time: 1921 to 1928. Not only did you post an annualized return of about 18% (without even counting dividends,) it was with periods of severe deflation so the real return were unbelievable. Think about how big the 1920's returns were: Warren Buffett, possibly the best investor of all time, has a long term return of around 20% per year. Yet, the broad market returned 18%+dividends in the 1920's. If the market was around 18%, I'm wonder what superinvestors of that era were earning. Roaring 20's they called it.

Coming to the present, there is some good news. Our current bear market, which this table assumes started in 2000, is past the average bear market length. If the bear had gone into hibernation in 2002, it would have been one of the shortest in history. But given the huge collapse last year, most of the pain is past us. But--yes there is always a but--the P/E ratios are way too high. Interest rates are much lower and inflation is likely to be low so perhaps a higher P/E is warranted. Nevertheless, the high P/E indicates that we don't have a clear buy signal. This is one of the reasons I keep saying that 'stocks are attractive but not exceptionally cheap.' All prior bear markets have had much lower valuations. Unless you come up with some 'this time it is different' theory (such as arguing that more people invest in stocks and this pushes up prices, or the argument that other assets are less attractive than stocks so people will pay more for stocks) there may be more downside. The presently high P/Es, even after the massive decline over the last 8 years, shows how much of a bubble we had in the 1990's.

10 Year Rolling Return Breakdown

The following chart illustrates the 10 year rolling total return (from Stock Rolling Returns by Crestmont Research):

The 10 year rolling return is the return you would have had if you were invested for the prior 10 years. The average 10 year rolling total return is around 10%. From the top chart, it seems that 10 year returns rise after hitting a return of 3% or less. If you buy when the prior 10 years yielded around 3% or less, you are likely to end up with more than 5% returns in the future.

The good news for us is that the current 10 year rolling return is below 3%. There is no guarantee that this year's or next year's returns will be positive, but it is likely that returns in 5+ years from now will be higher than 5%. This further confirms the point, made in the prior secular market section, that the worst is likely behind us.

(As a side note, you'll realize how powerful the bull market in the 1920's were by noticing that the rolling 10-yr return was actually positive in 1932 or 1933--even after the massive decline from 1929 to 1931! The rolling return only collapsed in 1937 or somewhere around that--this would be the case for those unfortunate enough to start investing in 1928, 1929, or possibly 1930.)

The botton chart illustrates the various components that contribute to returns: change in valuation (P/E in this case,) dividend yield, and earnings growth. A huge chunk of the return in the 80's and 90's was due to an increase in valuation (P/E went up) and this is reversing. The P/E contraction will likely continue for many years (as mentioned in the prior section, the P/E is still way too high compared to the past.) If you look at the chart, you'll also notice how there were always multiple red bars in prior bear markets. It is unlikely that the P/E ratio will stabilze or start increasing this year or next.

The most important question is whether dividend yield and earnings growth will compensate for the P/E contraction. If they can't, we are looking at a nasty bear market where losses will bleed you dry.


I hope all of you enjoyed the information and my commentary (thanks to Crestmont Research for providing the tables and charts.) I have uttered a lot of seemingly contradictory comments but it all makes sense. The stock market is not excessively cheap and therefore we will get conflicting signals. If, however, the market does become really cheap then I'm sure differing macro signals will be consistent with each other. It should also be noted that we don't have enough data points so the results aren't exactly statistically meaningful.

None of what I have presented implies that the market will necessarily decline further. I personally think there will be further declines (already we are down 10%+ this year) but most of the severe bear market correction has likely occured (barring some crazy event like war or something.) Since I tend to be conservative--not with my stock picks but my macro stance--I feel better hanging out with the bears rather than the bulls. If you are a stockpicker, right now is an excellent time to consider buying, or at least researching something. I do think it's worth keeping some capital available since valuations may drop far below what one may think is likely. Come up with and research a bunch of dream stocks that are overvalued and put them on your watchlist.

Tags: ,

1 Response to Secular stock market cycles & Ten year rolling returns

March 7, 2013 at 11:47 PM
This comment has been removed by a blog administrator.

Post a Comment