With six weeks until US midterms, arguments are already heating up as to their likely impact on markets. Should you use any weakness to load up on risk? Or, faced with the prospect of losing at least the House and possibly the Senate, is Trump likely to resort not only to tariffs and $5000 promises but potentially even to constitutionally questionable tactics2 like involving ICE personnel on election day in ways which are potentially disruptive to markets?
To judge purely from the historical record, both arguments have merit. This short piece examines the data and argues it leads to a quite specific implication for trading.
Sell in May, short September, the January rally – each has their adherents, and we’re always surprised just how popular seasonal trading strategies are in general. Perhaps it’s their simplicity which makes them popular, but in general we approach all such arguments with a healthy degree of skepticism. October, for example, has the worst reputation of any month and yet is still positive 59% of the time. For the S&P it contains 1929 (-20%), 1987 (-22%) and 2008 (-17%), and still averages +0.6%. Strip out those three and it averages +1.2%.
