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)

There are some principles that people need to learn. One is to look widely. Be open-minded about every form of investing in every nation. Don’t let yourself get misled by trying to limit yourself to one area.

Our second principle would be diversification. Of course, nobody knows what tomorrow will bring, so it’s not wise to put too many eggs in one basket. Nobody should ever have over half of their total investments in one nation, over half their investments in one industry, and perhaps no more than one fifth of their investments in one corporation, so a worldwide search is better than a local search. Diversification is a safety factor that is essential because we should be humble enough to admit we can be wrong. — John Templeton (source)

That gets me to the point about the relationship between gambling and investing, that is, what you learn from one helps with the other. Gambling games are, for the most part, an area where you can calculate the odds, the probabilities, in detail and get them rather exact. There are some exceptions, like sports betting and so forth, that are more like social or financial markets. But you can actually come into a gambling game with known probabilities and get answers. You have the advantage, like you have in the physical sciences, of so-called repeatable experiments. You can simulate a gambling game a million times if you want because you know the probabilities. It’s much more difficult in the securities markets because we just have one history from which we have to infer what’s going on, and the probabilities that we have are not exact—they’re just estimates. We’re in a world that’s controlled by people, which evolves in complex ways that we don’t fully understand. So you don’t have the same simple rules you have in the physical sciences and in the calculations that are behind gambling games. Nonetheless, the things you learn about gambling games carry over, in large part, to the investment world. — Ed Thorp (source)