Tuesday, May 4, 2010

Sell in May... Don't Go Completely Away

There is always lots of "Sell in May, Go Away" chatter this time of year. Marketwatch's Mark Hulbert (via Abnormal Returns) details why:

There are surprisingly strong statistical reasons for "selling in May and going away," many nevertheless find it difficult to actually do what the strategy calls for -- sitting on one's hands from now until Halloween.

It is especially hard to do now, in fact, with the bull market defying all odds and continuing to chug along. Money managers tell me that the worst crime that they can commit, at least in the eyes of many of their clients, is being out of the market when it is rising.
Agreed... and while EconomPic in no way, shape, or form advocates investing in the manner described below, it is interesting none-the-less. Rather than "Sell in May and Go Away", the chart below details the results of the "Secret Sauce" (i.e. Sell S&P 500 in May and then invest in the Long Government / Credit bond index, rather than sit in cash) vs a buy and hold S&P 500 strategy (note that these returns include reinvestment of dividends).

Why does this work?

I have no idea...

Source: S&P / BarCap


  1. Perhaps it worked because the bond market went through a long-term bull market from the interest rate highs of the early 80s to the ultra low levels of recent years. Going forward, with interest rates starting at such low levels the strategy may cease to produce such stellar returns, at least during the "go away" months.

  2. easy answer... its all detailed in Sy Harding's book "Riding the bear" ... anyway there are a dozen seasonal factors he lists - generally by buying in end of october and selling early may you avoid most corrections.

    as to why, again, tons of anecdotal reasons -- year end tax selling meets january reinvestment, global focus during the back to work/back to school winter cycle, more uncertainty around summer business cycles, money managers dont want to have money at risk when they are on August holiday... blah blah... bottom line is Sy Harding backtested and this works.... OVER LONG HAUL.... no guarantees it works one year versus the next...

  3. Because they pick Halloween. Adjust sell in May and go away to buying equities in September and your results change. Then you get exposed to the 1929 crash, 1987 crash and the 2008 crash.

    I remember reading something comparing returns when one missed the 10 best days in the past X years, missed the 10 worst days in the past X years and just always invested. Missing the 10 best days did best, where as missing the 10 worst days did worst, if I remember correctly.

  4. Hello
    You will find additional figures sue thereon
    http://weinstein-forcastinvest.net/investissement-les-periodes-propices-aux-achats-d% e2% 80% 99actions /

    Thank you to you

  5. Don't know why it works but today was a good day for it!

  6. Well, considering interest rates were in the high 6s and then went to 16 in the early 80s before going down to the 3s now, I could come up with some idea of why that happens. You've been long fixed income for almost half of the year during 25+ years of steadily dropping interest rates. Of course your credit portfolio appreciated.