Causation and Correlation

Causation and Correlation

This should be a short post this evening, because I am tired.  Always be skeptical of analyses that reason like this: A will lead to B, B will lead to C, C will lead to D, D will lead to E, E will lead to F, which is a (horrible disaster / incredible success).

Causality is problematic.  Things don’t always move the way we would expect, because the system is more complex and robust than most anticipate.  There’s more than one way to skin the cat.  Also, we often get spurious correlations, which fools amplify noise into signals.  Correlation is not causation.

Also, there are expectation effects that vary over time.  Sometimes a market effect that commonly occurs has been overinvested in, it is true usually, but overpriced for now.  Example BBB bonds out perform in the long haul, but in May of 2002 as spreads were crashing in, it was right for me to throw those bonds out the window for AAA, AA, and A-rated bonds for what was a small yield give-up.

When you reason about investing, write out the chain of causality, and recognize that things that are certain are less than certain.  The more links, the less likely the whole argument is.

It is better to keep things simple, and avoid extended reasoning.  Most good investing involves simple judgments on broad issues — try to get the big things right, and the little things will follow.  You can do well in investing by avoiding situations that are highly levered; yes there are some big gains with high leverage, but there are more big losses.

So keep things simple, invest simply, focusing on relative value and income.

By: alephblog