Amidst the buzz about "seasonality" in the grain and soybean markets, I feel compelled to offer a different perspective. While I won't dispute the existence of such tendencies, I find it more intriguing to question if they provide an edge for market participants. I am skeptical that they do.

I hear the seasonality evangelists shouting, "But James, in x out of y years, z happens. That means markets are predictable." That may or may not be true. If we assume for a second that it is true, and even if we forget the well-known disclaimer "past performance is not necessarily indicative of future performance," we could not inherently conclude that there is profitability, or even edge, in knowledge about seasonal price direction.

Did I just claim that there is no inherent edge in knowing which way the markets are likely to go? Yes.

For starters, I want to quickly discuss edge, its likely sources, and one place I consider unlikely to generate it. Edge is an advantage in a marketplace that gives its owner a positive expectation for profit in the long run. Unique knowledge, skills, ability, or capital resources are frequent edge sources. "Unique" is the operative word here, but it does not mean "one of a kind" in this case. Instead, for our purposes here, "unique" means something more like "unavailable to the majority of market participants." On the other hand, historical data available to almost all market participants (especially first-level analysis of that data) is unlikely to yield any edge, at least in my opinion. With all the money on the line and all the remarkably intelligent people chasing that money, it seems an absurd bit of hubris to think that a surface-level assessment of historical price patterns will yield anyone an edge.

So, how can there be no inherent edge in knowing which way the markets are likely to go? Here, the casino gambling world provides a fitting example. An American roulette wheel typically has 38 slots numbered one through thirty-six, zero, and double zero. Which means a roulette wheel is remarkably predictable. If a gambler picks any single number, he has about a 97.4% (37-out-of-38) chance of being wrong. Of course, the reverse is also true; if he bets equal amounts on all but one number, he has about a 97.4% chance of being correct. Well, sign me up; everybody wants to be right, right? Not so fast.

We used to have a saying on the floor, "Do you want to be right, or do you want to make money?" The problem with betting equal amounts on 37 of the 38 available numbers is that, at most American casinos, betting on the correct number only pays 35-to-1, thereby guaranteeing at least a small loss (one unit) and potentially a massive loss (thirty-seven units). Some folks (probably full-freight retail brokers) might balk here, claiming that casinos set the rules unfavorably and the markets work differently. Do they? What about commissions, exchange fees, regulatory fees, and bid/ask spread slippage? "The vig" is nearly as certain as death and taxes.

Even if we changed the roulette payout structure to fair odds of 37-to-1, without an edge, at best, the gambler would break even over the long run. Even with a 37-out-of-38 chance of being right and fair odds, a gambler still does not have an iota of edge. Worse still, on the nth draw, when the number the gambler bets against hits, he loses 37 units. Suppose that happens early in the game or several times in quick succession. In that case, the gambler might lack the funds necessary to paint the board with another 37-unit bet and, thus, be knocked out of the game entirely - so much for breaking even. Consequences (especially risk of ruin) are at least as important as likelihood when dealing with uncertainty and determining edge.

There are at least two potential complications when determining the edge (if any) in seasonality. The first and most important is that we have not seen (and cannot see) the complete distribution. History is always in progress and, therefore, does not fit neatly into the same easily calculated uncertainty of the roulette wheel. The market's future is far more uncertain than the roulette wheel's. Second, it is at least arguable that given the broad access to historical price data and ease of calculation, seasonality should be fully "priced into" the market. Put more simply, don't expect to find an edge in something many people give away for free all over the internet.