Kub's Den

MUGWUMP (noun | mug·wump)

: A person who is independent (as in politics) or who remains undecided or neutral

Presidential elections are assumed to have some effect on the markets, or on the stock markets anyway, and during these two summer weeks of flag-waving convention furor, I thought I should examine the grain markets' relationship to elections, too. But first, I want to assure you that I myself am a mugwump. I'm no longer a Republican or Democrat (and I won't tell you which one I used to be), so I can't be accused of any political motivation in bringing up this topic. The Libertarians already had their convention back in May, and I promise I won't write any names starting with C or T or J in this column at all.

Okay, so the presidential election cycle as experienced by the stock markets goes like this: Stock prices tend to move lower during the first two years of a presidential term, then move higher during the last two years of a presidential term. Theoretically. It was described using three decades of data in a 1995 study published in the International Review of Economics & Finance* and has been confirmed off and on by various industry reports and other academic studies.** As one researcher summarized: "The presidential cycle effect prevails even after controlling for the party in power and the incumbent versus nonincumbent presidents."

And yet. It has rarely worked in the past 10 years. During the first half of George W. Bush's second term (January 2005 to December 2006), the stock market didn't move lower as The Cycle would have predicted. The S&P 500 Index actually rose 16%. Again it failed to follow the pattern when it proceeded to move lower during the last half of his second term. Nor has the cycle worked well as a predictor during Obama's presidencies: The S&P 500 Index went up 33% from January 2009 to December 2010, the first half of his first term when, according to the presidential election cycle, it should have been falling. The stock market did obey the cycle during the last half of his first term. Then it failed again during the first half of his second term by rising again. We have yet to see if the S&P 500 Index will post a gain or a loss (above or below 2058.9) during the last half of his second term.

That is a potent demonstration of the dangers of using infrequent data sets to predict future outcomes. When one team of researchers considered market data from 1965 to 2003, that was only 9 1/2 cycles. If a statistician is trying to draw a decently legitimate trendline on any given chart, she typically wants at least 30 samples, as a rule of thumb. So these academic descriptions of a theoretical presidential election cycle simply don't have enough data behind them. But they can't really go back farther in time to get a robust number of samples because whatever the stock market was doing in 1880 when Winfield Scott Hancock was running against James A. Garfield, it just does not hold much wisdom for today's world of sovereign wealth funds and algorithmic traders.

Note that this is the same problem faced by political commentators trying to predict the actual outcome of an election (or a referendum, in the case of our friends across the pond) when there can never be any truly comparable samples of historical data to draw upon. It's unlikely there will ever be any set of two or three exactly comparable candidates facing an exactly comparable population with exactly comparable sentiments and economic situations.

We see that in the grain markets, too, when we analysts sometimes stretch too hard to say that the predicted La Nina weather in later 2016 might resemble the La Nina weather of 2010 (or whenever) and so maybe prices will do the same thing they did then. Sure, the weather might be comparable; the yield trend might be comparable; but the prices? I'm doubtful. Global economies are different now than they were then; outside markets are different; industrial usages are different; investor tendencies are different; lots of things are different. If we had a lot more crops facing a lot of different scenarios every year for a long series of years, we'd have a lot better chance at isolating or controlling for other variables, and then maybe we could be more confident when making cyclical prophecies.

Ultimately, do commodity markets (especially grains) behave according to a four-year presidential election cycle? Pretty much no. In Kub's Den columns from 2008 and 2012, I already demonstrated that grains didn't have a reliable presidential election cycle, even when the stock market did appear to (back before the last few presidential terms so blatantly broke the pattern). Commodities and stock markets used to have a reliably negative correlation -- when raw materials got expensive, corporate profits tended to suffer and vice versa -- but even that relationship has drastically changed in the last several years as huge investment funds have moved their money in and out of all asset classes at roughly the same times.

Let's hope that economic sentiment for the next several years will be driven less by which half of a presidential term we're in, and more by the policies that the government is enacting. In any case, there is something to be said for being a mugwump -- not about the election necessarily, but about grain prices themselves. If you remain as unconvinced by the bulls and bears as you do by the elephants and donkeys, then you should keep your marketing program flexible and always have the risk of a worst-case scenario covered.

Elaine Kub is the author of "Mastering the Grain Markets: How Profits Are Really Made" and can be reached at elaine@masteringthegrainmarkets.com or on Twitter @elainekub.

*Gartner, M., & Wellershoff, K.W. (1995). Is there an election cycle in American stock returns? International Review of Economics & Finance, 4(4), 387-410.

**Wong, W.K., & McAleer, M. (2009). Mapping the Presidential Election Cycle in US stock markets. Mathematics and Computers in Simulation, 79(11), 3267-3277.

(CZ/SK)

Copyright 2016 DTN/The Progressive Farmer. All rights reserved.