Premises
Now then let’s get to work. If any of you have seen the movie pi then you are probably familiar with the “assumptions” that the character Max mentions several times:
1. Mathematics is the language of nature.
2. Everything around us can be represented and understood through numbers.
3. If you graph these numbers, patterns emerge.
Therefore: There are patterns everywhere in nature.
Now while I appreciate the dramatic effect, the logic is fairly poor if not not concrete at best. Later he mentions some examples including:
the cycling of disease epidemics;
the wax and wane of caribou populations;
sun spot cycles;
and the rise and fall of the Nile.
Sure, that’s great. Here are the two major flaws with all this. For starters, these examples have limited predicting power value for translation to something like the stock market (which by the way is really what Max is after). In other words, were you to come up with a formula for sun spot cycles, you would see some accuracy in its ability to predict the next sun spot cycle (seven years, right?), but that’s only because sun spot cycles have been cycling that way for a few billion years. In other words predicting sun spot cycles is a bit like predicting a pendulum’s motion–i.e., predicting something that happens to be very necessarily predictable. Suddenly we shouldn’t feel so cocky. Is the stock market necessarily predictable in this sense? If you look at in on balance, over the long term, it does seem to oscillate or “ebb and flow” like something directly from the scene of ripples moving their way carelessly across the beautiful sun draped Nebraska wheat fields. Ahh…
What if we were to measure sun spot activity for two hours during some seventh year? Might we see no sun spots? Might we see several? On balance we know that during its cycle there are going to be sun spots but is it accurate at a more microscopic scale? By looking at graphs of the Dow over seventy years can we develop a formula that would aid us when the market opens next Monday morning? And what about this: if we had evidence that the sun was on the verge of running out of hydrogen next year, would our model have any merit? That makes it even worse: we can predict sun cycles not only because they are predictable (which is bad enough) but our predicting ability is hinged on the nearly scientific certainty that the sun has plenty of fuel. And this is just first flaw.
The other flaw is this: it’s a good thing sun spot cycles are so heavily influenced by wars between men and other men. Now we can really predict them. Not to mention how thankful I am that the Dow Jones Industrial Average hums along independent of anything as unimportant as say, the European economy. Is our model of the waning and waxing of Caribou populations most relevant just before or just after a large purge carried out by poachers? The stock market is so heavily dependent on human activity (being so purely one itself) that removing human influence so as to be able to study it truly scientifically would be to simply destroy it. We can predict the rise and fall of the Nile only so long as we don’t dam it up. If we can see, over a long time (say the life of the Dow) that it seems to generally trend up, that would indicate that those companies have generally, over those years, simply become more prosperous. Woopdydoo. I think that, over the same period of time going forward, the Dow will continue to trend up. Now you have a savings account. And it will give you a yield over the next seventy years; but only if you buy an equal share of all the companies in it.
So what’s my point? My next entry will explain it.