Kahneman and Long Cycles

Investment Manager's Letter March - April 2012

In November, I, as investment manager, wrote about long cycles in the stock market, wherein the equities markets PE ratios expand or contract in long (15 year plus) cycles, more or less independently of corporate earnings. Since 1929 corporate profits have risen at an annual rate of 6.33% per year. Except for the occasional economic recession this rate has been fairly steady. During the 1929 – 1942 bear market the growth was 6.32% per year, in 1942 – 1966 bull growth 5.17% per year, in 1966 – 1982 bear the growth was 4.49% per year, in the 1982 – 2000 bull growth was 7.66% per year, and so far in the 2000 – 2011 bear corporate profits have increased at a rate 8.89% per year.

It is interesting that while the current cycle (2000 – 2011) which has included two spectacular bear markets and one general worldwide economic meltdown, corporate profits have put in the best performance of the 82 year period. During this 82 year period, we have experienced three long bear markets in which equity prices have stagnated for years even though corporate profits continued in their long term up trend. In all but one case (1966 – 1982), corporate profits performed better during the bear markets than they did during the bull markets. How do we explain these long periods where Mr. Market ignores positive earnings?

Since the early part of the twentieth Century, the history of the stock market is one of long periods of expanding PE multiples (secular bull markets), followed by a long period of contracting PE multiples (Secular bear market). While this pattern does nothing to support the notion of efficient markets, it does find considerable support in the work of the behavioral economists.


Daniel Kahneman in his outstanding new book "Thinking Fast and Slow" Presents an interesting description of the human thought process, when making a decision we tend to rely heavily on our "System 1" behavior, this is the hard-wired tendency to make snap decisions without recourse to the hard work of rational thought, and even if slow careful thought would give you a better basis for assessing probable consequences. System 1 is particularly handy when decision involves something difficult or impossible like trying to predict future stock prices.

Kahneman calls the human brain "a machine for jumping to conclusions". The problem is that intuitive thinking, except with few significant exceptions, tends to be wrong. The thing that makes "System 1" popular is that tends to present quick, easy answers to difficult questions, and makes us sound smart even though our answer may not be correct. System 1 is particularly useful when the question is difficult. This makes System 1 particularly useful for the talking heads on TV since it means they can always come up with quick answer that sounds good.

What makes System 1 relevant to this discussion is what Kahneman calls WYSIATI "what you see is all there is". System 1 is very good at constructing an answer from whatever information is available in your brain and then as he says "jumping to a conclusion". Since our System 1 is focusing on the information that is available in the brain it is by definition focusing on the past. And so, in the market, whatever trend is in place tends to stay in place. In a bear market, our System 1 is powered by negative outcomes. All new information will tend to receive a negative twist. The recent economic disaster makes it much easier to predict a new disaster, and difficult to predict a positive outcomes.

In a bull market, the brain is ruled by positive data from the recent past, so System 1 becomes a momentum machine. The longer the trend continues the more data that is available in System 1 to support the current trend. The housing price bubble is a prime example of the power of System 1 to overwhelm rational thought.


1970's Economist Hyman Minsky said that Capitalism is inherently unstable, by which he meant that during periods of prosperity, when banks and other financial institutions are earning money and their earnings are increasing steadily year after year, competitive pressures from the market place will tend to push these institutions into investments with higher and higher risk.

Investors start the bull cycle with a very low tolerance for risk, having just survived a 15 to 17 year bear market. With their System 1 full of disasters, it is easy to conclude that today's problems will all end badly. By the end of the up cycle a couple of decades later (after years of steadily risings prices and low levels of loan default), investors have lost their memory of, and the ability to understand, risk. Their System 1 is full of only positive outcomes. Time numbs the pain of the past as the investor discovers that the guy across the street is making more money than he is, so his risk tolerance expands.

The irrationality generated by our recent cycle created two great bubbles, the internet bubble and the housing bubble. These bubbles have generated a level of collective insanity that defies rational explanation, and were not in any sense limited to the United States, but spread like a virus throughout the developed and the developing world. Nor was the insanity limited to financial institutions, but spread across the populace of most of the countries involved, as citizens of all stripes sought instant gratification bought with someone else's money.

Michael Lewis' excellent new book "Boomerang" details fascinating tales of this financial mania as it traveled from an exploitative Wall Street to naive and unsophisticated Banks in Ireland, Iceland, and Germany. He details how the financially corrupt were aided and abetted by morally bankrupt political systems in places like Greece and U.S. (His best example in the U.S. – California). Thus we see the folly generated by our System 1.

So we have a long bull cycle where PE ratios expand till they grow bubbles and a long bear cycle where PEs shrink down to single digits. It is apparent that markets are not ruled by corporate earnings so much as they are measuring the investors' current psychosis.

Implications of long cycles

What does this mean for investors? It is important to recognize what secular cycle you are in, and what part of that cycle. While the long cycles have little day to day impact on stock prices it is important to understand that interpretation of PE ratios will be continuously changing and that, in a long term trend, PE ratios will either be contracting (1966 – 1982) (2000 – present) or expanding (1942 – 1966) (1982-2000) irrespective of the general long term up trend in corporate earnings. It obvious from the chart above that the best time to buy stock is toward end bear market cycle. The central irony of investing is that stocks well into a bear market, when they seem the most dangerous (because everyone's System 1 is focused on negative outcomes and are wont to predict disasters, "what you see is all there is'') are not nearly as dangerous as markets at the top of the bull cycle, when the background noise is predominately bullish. This means that it is important to learn to become a contrarian. This is easier said than done, but we need to get help from Kahneman, and like Buffett learn to ignore our System 1.

While tendency to leap to conclusions (System 1) was useful genetic trait for our ancestors (the guy running at you with a bloody sword is dangerous) it is also the foundation for mass hysteria in financial markets. So always "Buy stocks when others are fearful".

This phenomenon may have some interesting implications for investments made in the next two or three years, and we will return to this subject in the future. Even for those of us who accept that these cycles are real have to wonder how long they will continue. For the answer to this question we clearly do not need to consult with stock market analysts, but need to study the nature of human behavior and mass psychology. Probably we can expect these cycles to disappear about the time there is a genetic mutation that eliminates System 1 from the human gene pool.

Zircon - This is a contributing Drupal Theme
Design by