Mechanical Investing
Hedging in September

Related Links
Discussion Boards

By mungofitch
August 20, 2009

Posts selected for this feature rarely stand alone. They are usually a part of an ongoing thread, and are out of context when presented here. The material should be read in that light. How are these posts selected? Click here to find out and nominate a post yourself!

The US stock market has historically not done very well in September. Sometimes yes, sometimes no.

Similarly, the US stock market tends to do best around the end of the month, and worse during the rest of the month.

Putting these two effects together, arguably the worst time of year to be fully exposed to price movements is the middle section of September. Now, I wouldn't want to bet my last button on the market going down in September, but if you were ever going to hedge, this wouldn't be a bad time to consider it.

FWIW, here is an example: hedge using short Russell 2000 futures for the period from market close on the 4th trading day of September through to market close on the 2nd-last trading day of September. Rather than the superior six, you can think of this as the inferior 14.

Here are the returns since 1979 for just the hedge alone. Remember, this is solely the return for shorting the small cap index for the middle section of Septembers, with zero return for rest of the year. The figures use futures for the length of their history back to 1983,  and for the years 1979-1983 the actual index levels are used as an approximation. Futures trading is extremely cheap so I have not bothered with friction costs.

1979   -1.769%
1980    1.756%
1981    6.690%
1982   -2.143%
1983    0.312%
1984    0.269%
1985    5.513%
1986    6.602%
1987    0.942%
1988   -1.278%
1989    0.640%
1990    9.980%
1991   -0.648%
1992    0.368%
1993   -2.324%
1994    1.025%
1995    1.236%
1996   -3.055%
1997   -3.601%
1998   -6.127%
1999    4.393%
2000    4.070%
2001   11.276%
2002    7.928%
2003    3.008%
2004   -1.546%
2005    1.824%
2006   -3.484%
2007   -4.199%
2008    9.432%

Of more interest is that the hedge as defined above has historically had a negative net cost. i.e., it's like buying insurance where the company pays you the premium.

The average year above is +1.57% (average without 2001 is 1.235%) Median is +0.79% The return is positive 62% of these years.

So, what good is it? Well, if you add some multiplier of this hedge to your MI screens, the expectation might be both increased returns and lowered risk. For example, take the recently posted CashWhileYouWait variant. For the January 1989 to June 2009 period, this backtests as 39.4% CAGR and 73.8% of the risk of buy-and-hold using my DDD3 downside deviation metric. (magnitude weighted probability of a losing quarter, or year return under 10%)

If you add in 100% of this hedge amount to each September's return, the  overall portfolio result is much nicer: 41.7% CAGR and 64.7% of the  risk of buy and hold. Worst rolling year goes from -45.3% to -39.6%. So, 2.31%/year better CAGR and much reduced risk, all simply by being market neutral for 14 days per year.

If you solve for the lowest risk, the solver recommends going very net short based on history: 1x long stocks and 2.65x short the Russell. This raises CAGR to 44.8% and drops risk to under 60% of buy-and-hold in backtest. But I wouldn't recommend that because you are then betting that the market will drop, rather than merely declining to bet that it will rise (or if you prefer, betting merely that your screen will beat the small cap index).

Will it work this year? Beats me. But it has worked sometimes, and it has saved me a lot of pain.