April is the month when the talk of 'Go Away in May' is revived every year. It appears from historical data that it may indeed be possible to get substantially higher returns (or at least decrease volatility) for some ETFs by buying them on the last day of November, and exiting the position at the end of May in the following year. The improvement becomes quite impressive if the strategy also includes investment in a long term treasury bond fund held during the remainder of the year (buy, say VUSTX or TLT on the last day of May, and exit the position on the last trading day of November).
Here are the results of back testing of the enhanced strategy using VUSTX for some SPDR Sector ETFs for the period beginning on the date of inception thru 2011.
|Date of Inception||1999||1999||1999||2007||2007||2007|
|CAGR (Buy &Hold)||10.2%||5.3%||4%||6.3%||0.2%||-13%|
|CAGR (Sell in May)||20.2%||13.7%||12.5%||28%||34.5%||20.7%|
|Min Return (B&H)||-39%||-44%||-39%||-38%||-59%||-47%|
|Min Return (S in M)||6.3%||-5.5%||-3.6%||11%||13%||4.8%|
If the strategy is applied to an equally weighted portfolio of all six of these ETFs, the results are quite remarkable (for 2007-2011):
CAGR (Buy and Hold): 1.8%
CAGR (Sell in May) : 24.3%
Minimum Return (Buy and Hold): -43%
Minimum Return (Sell in May): 18%
(Alpha and Beta not meaningful due to short time period).
Obviously 2007-2011 is not long enough a period to make any significant judgments for practical investment decisions. However, a similar improvement would have also occurred for a portfolio of Fidelity Select Sector Funds for which data is available for a longer period. Indeed, for an equally weighted portfolio of seven such funds (FSESX, FSENX, FSRPX, FSRFX, FSCGX, FSELX and FSCHX) the following results are obtained for the twenty one year period 1991-2011.
CAGR (Buy and Hold): 13.3%
CAGR (Sell in May) : 20.4%
Minimum Return (Buy and Hold): -44%
Minimum Return (Sell in May): 2%
Sharpe Ratio: 1.46
One Factor Alpha: 16.3%
One Factor Beta: 0.19
Disclosure: I am long XLE.