In a fascinating paper published in 1989, Benjamin Friedman and David Laibson hypothesize that all investors function with two models in mind. The first model approximates the efficient market notion that prices are so close to true value that a movement up is as likely as a movement down and that those movements will be lognormally distributed. The second model is that all hell could break loose at any moment.

Many years ago, an older partner taught me to distinguish between outcomes that are unlikely and outcomes that are catastrophic. The latter are to be avoided even if the odds on them are tiny. Rational investors respond to this type of problem by operating with anchors to windward. They diversify and avoid total commitment to any one bet. But the fear we all share — that what looks like a market today just might not be there for us tomorrow — makes today’s market less than perfect and less than totally liquid. — Peter Bernstein (source)

It was the publication of E.L. Smith’s little book entitled, Common Stocks as Long-Term Investments. His study showed that, contrary to prevalent beliefs, equities as a whole had proved much better purchases than bonds during the preceding half-century. It is generally held that these findings provided the theoretical and psychological justification for the ensuing bull market of the 1920’s. The Dow Jones Industrial Average (DJIA), which stood at 90 in mid-1924, advanced to 381 by September 1929, from which high estate it collapsed — as I remember only too well — to an ignominious low of 41 in 1932.

On that date the market’s level was the lowest it had registered for more than 30 years. For both General Electric and for the Dow, the high point of 1929 was not to be regained for 25 years.

Here was a striking example of the calamity that can ensue when reasoning that is entirely sound when applied to past conditions is blindly followed long after the relevant conditions have changed. What was true of the attractiveness of equity investments when the Dow stood at 90 was doubtful when the level had advanced to 200 and was completely untrue at 300 or higher. — Ben Graham (source)

Those who believe in the permanence of tempestuous seasons will view life as a succession of short runs, where noise dominates signals and the frailty of the basic parameters makes normal too elusive a concept to worry about. These people are pessimists who see nothing in the future but clouds of uncertainty. They make decisions based only on the short distance ahead that they can see.

Those who live by regression to the mean spend their time entirely differently. They expect the storm to pass, so that one day the ocean will be flat. On that assumption, they can make the decision to ride out the storm. They are optimists who see the signals by which they will steer their ships toward that happy day when the sun shines through.

My own view of the matter is a mixture of these two approaches. Hard experience has taught me that chasing noise leads me to miss the main trend too often. At the same time, having lived through the bond yield/stock yield shift of the late 1950s and the breakthrough of bond yields into the stratosphere beyond 6 percent in the late 1960s — just to mention two such shattering events out of many — I look with suspicion at all main trends and all those means to which variables are supposed to regress. To me, the primary task in investing is to test and then retest some more the parameters and paradigms that appear to govern daily events. Betting against them is dangerous when they look solid, but accepting them without question is the most dangerous step of all. — Peter Bernstein (source)