Sell in May and Go Away? (Historical performance)

Here's the historical performance at The Big Picture



We can see that the Nov- April periods have been relatively volatile, too. A more important question remains, will this pattern stay in the future?

Dangers
It is dangerous to assume that this pattern would persist into the future, without a robust fundamental theory. Statistically speaking a sample size of 50~ means nothing. From my experience, there's no such thing as "statistically significant" with financial time series. A tradeable edge requires solid fundamental theory.

What this pattern CAN be used for
Sales at brokerage firms in November.

Free historical economic data

Wikiposit has made available a wide range of historical (daily, weekly, monthly) data freely available. Despite the lack of intraday data, having daily settlement/term structure data for futures contracts makes it very convenient for basic mathematical finance research, or estimation of trading edge.

There aren't too many places with these data sets freely available. Free data is good!