I'm not forecasting whether there will be selling in May, but I am saying one thing. I will be going away. At least from trading for the month.
My bond system, the primary system I run, does not trade in May. Nor does it trade in November. Which at first was odd to me.
But eight years of backtesting shows that both of those months are not good for the strategy that I follow. Conceptually it makes sense to me now, but not at first.
Think of it this way. There are certain hours of the day that are better to trade than others for a given strategy. A simple example is the open of the equity cash market for the first thirty minutes. There's a lot of chop, noise.
For some that is the ideal time to trade, for others it is not. For the latter it's better to wait thirty minutes. So right there it shows you, two different strategies are best for two different timeframes.
So if time plays a factor intraday, why not on a larger timeframe? Why not on a calendar basis? Just like the first thirty minutes are choppy, so too are months like May or November.
Sell in May and go away is an adage that will live, likely forever. Even if the historical data does or does not support such a mantra. As a result you'll have traders following what they believe to be a great opportunity to go flat and or short. Contrasted with others who believe they should sit in their position or buy.
The result, choppy markets at times. Just like those first thirty minutes each day. At least that is my rough analysis of why May is period of time my strategy sits idle. But whether that is accurate or not, one thing is.
Historical data shows trading in May for my style, is anything but constructive to my PnL.