Seasonality is really useful for detecting profitable opportunities on the markets ? It seems absurd but it is. Even the good serious econometric theory shows without any doubts that markets have a seasonal trend. This is obvious if you think that the meteorological cycle, the human, vegetal and animal life cycles all have an impact over the economy. And moreover human beings get accustomed over time to do and to do again the same acts because repeating the same gestures gives confidence and it gives the impression you are in control of reality: this is clearly not true, reality is completely out of our control but the impression we get from this is comfortable.
I did some interesting research on the inter-months seasonal trends about stock indexes (DAX, SP500, CAC40, etc.) and it seems that there are weeks it is better to buy and weeks it is better to be short, there are months that are bullish and months that are bearish. Obviously tests are not much significant since, let’s tell us the truth, even if you consider all the historical Dax price series you will never have more than some dozen of January months or October months. And we know that just a sample of 100 or more observations is significant. This is the reason by which I keep the systems running on my computers, I check them periodically in order to gauge whether they are making money or not but I am not so bold to trade them. And on this point is centred most of the criticism regarding the seasonality approach: if you have a system that buys in May and sells in September on a US commodity, for example, in any case you will have just 30-40 instances and from a pure statistical standpoint they are really few.
Another critique advanced against seasonal trading is that in order to overcome the significance problem because of a too small sample size if you extend too much the historical period, provided you can find appropriate data, then these data are not any longer in tune with the current economic situation. Western economies for example were much more dependant from oil in the Seventies than now so that also seasonal tendencies should have been much more robust in the past than nowadays. So why should you look for seasonal tendencies that are wrongly supposed to be constant in an every-day changing world ?
The truth is halfway from a pure statistical standpoint and a more practical approach to trading. I esteem that seasonality has a different degree of efficacy based on the different breeds of market you are trading. Let’s take the kings of all commodities, the cereals. Here huge stock supplies are brought from one season to the next so that there is a real economic reason why one season should be in synchronization with the next. We are not talking about figurative numbers, we are talking about real stocks brought forward from one year to the next by the storage industry. Stocks which next year production will take into consideration before deciding what to plant. We are talking about economic agents that have the same real needs year after year, decade after decade. Even if land cultivation techniques have changed a lot in the recent times, you can assume the Corn cycle is more or less the same than 100 years ago.
This is different from the case where you try to find seasonal tendencies in outright markets positions or spread positions that have nothing constant or in common, nor the players and the economic conditions, neither the vegetal seasonal cycle and the production and consumption behaviours. I see sometimes traders that look for seasonal spread tendencies in different and not related commodities markets, and this just because if you would have bought every first day of April and sold every second day of June you would have made a fortune. If you will put a monkey in front of a piano it will eventually compose a sonata, provided enough time is allotted. It is the same story.
So be careful about seasonal tendencies, and listen this from a trader that every week watch his seasonal trading systems making money without indeed trading them with real money. I believe to have all the right cards to have a respectable opinion on this subject.