
Among our favourite reads are the thought provoking vignettes on a wide range of topics published in
Deus Ex Macchiato. The
most recent entry, dealing with the statistical unpredictability of markets, provides us with this succinct description...
Think about it like this. Mostly in finance we assume that we have the equivalent of a standard dice. That is, while we assume we don't know what number will come up next, we think that we know the distribution of numbers perfectly. In fact the real situation is much more akin to throwing a dice where we have imperfect knowledge of what numbers are on the faces. They might be 1 to 6; but they also might be 1 to 5 with the 1 repeated; or 2 to 7; or something else entirely. Worse, the numbers are changed by the malevolent hand of chance on a regular basis. Not so often that we know nothing about the distribution, but often enough that we cannot be sure that the current market will be like the past.
The problem of how to make money, by necessity, suffers from the same drawback. Punters, for example, who try to suss out repeating patterns in stock prices, however simple or complex, make the assumption that future distributions will bear some resemblance to past and that the risks also will have been previously delimited. Typically, it is thought that the more examples one has of a given relationship`working out well, the more 'robust' a trading system based on one's method will be. Events of the last year and a half, following a five year period in which mechanical trading dominated the floor, should have put that illusion to rest.
Consider that there might not yet be enough data available despite that an investigator may have found hundreds of examples to prove the hypothesis - this because, thinking half rationally, the hard-wired trades that work when the numbers on the dice are 1 to 6 are not suitable when that distribution changes to 2 to 7. Betting on the height of the next person you come across in the Sixers' dressing room on the basis of data collected at a meeting of the Dwarfs Benevolent Society is not a winning plan. The safest assumption you could make under the circumstances would be that all midgets sum to 1, the same value applying to a profitable run during a stretch of market coherence.
To the end of perhaps further obfuscating the matter, the writer dedicated a bit of time on one of the recent rainy days that have prevented him from attending to his true destiny as olive farmer to compiling the chart at the top of the page. The cause of the confusion on the horizontal axis is that it is a record of the value of the DJIA with the four data points of each candle corresponding to the open, high, low and close of that index when it is, alternately, above or below its own 40-week moving average. The result is that each bar is of a time duration not predetemined. For example, point 1 of the 234 displayed begins on October 25th, 1929 and ends on the 19th of August, 1932. The bar upon which we currently reside began in January of this year. We chose 40 weeks as the averaging period because it corresponds the 200 days that many, rightly or wrongly, consider vital in determining investment possibilities.
Our pedantic suggestion is that a collector of statistics using above the 200-day average (or anything else) as a filter, upon finding that over 20 such periods in the past some system turned a profit on 150 long trades, count that as only 20 positives. Add to the mix that the index has crossed above that line 117 times since 1929 and the reader might let us know if he or she wants to play the game with the rent money.
*Note the very extensive periods in which buy-and-hold was utterly the wrong strategy, by the way. They are marked by the vertical lines.Click on the chart to make it full size.
-------------------------