Valid seasonal price patterns
The first of a series of meetings entitled “Marketing In a New Era” was held in Geneva, Nebraska, on Wednesday of this week. It is based on a similar computer simulation game similar to the original “Winning the Game” software developed by the University of Nebraska-Lincoln” back in the 1990’s.
At the Geneva meeting, a farmer questioned if my seasonal strategies continue to be relevant in the new era where a one day price move can be greater than the range of average prices over a whole year. My response was that even with the huge swings in recent years, prices follow relatively reliable patterns through the year and from one year to the next.
There are several ways I evaluate seasonal patterns, to be sure that the information I share has a reasonable probability of success. The first is that I am careful to study an adequate number of years. When I began researching grain prices, I started with 1980 and added the most recent year every year. That approach worked well, but as the number of years accumulated, I suspected that those early years became less relevant. Because of that, when I was doing research for this winter meetings, I decided to look at only the most recent 20 years. Therefore, my current charts show average prices for November soybeans and December corn only from 1992 through 2011. I never give any credibility to price patterns for less than ten years.
The second way I test my seasonal patterns is by comparing groups of years to the long term averages. For instance, I might compare the years 2002 through 2011 to the years 1980 through 1989. This comparison should tell me if the recent price action was accidental or if it followed a reliable, long term pattern. If I discovered a drastic difference between the two time periods, I would need to study whether the difference resulted from some fundamental factor or whether the pattern in early years was simply random movement. In the most recent version of my charts, I compare the last 20 years to the latest five years, to see if the current higher price level has eliminated the seasonal patterns. My opinion is that it has not. Prices continue to go up and down in a somewhat predictable pattern but at a much higher level.
The third test I employ is to make actual trades in the futures market when the charts indicate a good probability of a trend change. In other words, I test the system with actual money. I have been doing this since August of 1997. In that time, there have been 97 trades. The most profitable was $2.60 in October 2011. The biggest loser was -$3.15 in February 2008. The year 2008 was bad for my trading system. Since then it has been up and down, but generally profitable. This is clearly not a speculative strategy for the faint hearted. However, it proves to me that the seasonal patterns are still valid. It also proves that prices go up and down roughly the same number of times. The profit comes from the trading profits being larger than the trading losses.