Do you remember back in college, or maybe at your first job, when you could take off early on Friday and just not show up at all on Monday and no one gave a damn? Well, it looks like for a lot of investors, 3 and a half day weekends are the key to avoiding volatility in 2008. Note: I wanted to name this entry Monday-Bloody-Monday, but that was before we had a 900 point rally (wtf?) by the Dow.
A look back at history this year (yes, our children will be reading about this in history books someday) turns up a myriad of market changing events that all occurred over a weekend and which were subsequently digested by the markets on Monday: Bear Stearns, Fanny/Freddy, Lehman, Fortis, TARP. I wanted to see what amount of volatility, and more importantly, what returns an investor would experience had they engaged in such investment truancy. Indeed, my analysis shows investors could have outperformed the market by nearly 10% by selling on Friday and re-purchasing the index on Monday's close.
My first step was to analyze the standard deviation of returns on the S&P 500 index. I downloaded year to date closing prices for the S&P 500 index and used daily log returns to calculate an annualized standard deviation of 31.7 (250 day/year). This is already double the 30 year average standard deviation for the index! If I filter by just Mondays, this number goes up to 45 or nearly 3x the 30 year average!
Next, I looked at the compounded returns of the index vs. a dummy index that "sells" on Friday. The results were very surprising. For most of the year, the indexes tracked each other very closely, then in mid-September there is a drastic divergence. While the index is down 30% YTD, the dummy index is 10 percentage points better at down 20%. If we remove today's 900 point wtf-rally (I'm going to continue to call it that until I see another 11% daily rally in my lifetime), the outperformance is even more stark at 17 percentage points for the ex-monday index. On the whole Monday's have produced both higher volatility and lower returns for 2008.
What to make of this? If I said that all investors should categorically sell on Friday and buy on Monday I would be doing a favor to no one but your broker. This SMALL sample of data is bound to revert to the mean when the market calms down. Also, it is only natural that Mondays are busier days in the market as nearly 3 days worth of news must be "priced" by the Market at the open. What I hope to do is expose my own (and your) psychological biases towards market timing and expectations. While this may be a good trade ex-post and before trading costs, I doubt it is likely to yield sustainable alpha for the future. Any attempt to arbitrage this strategy would be akin to picking up pennies in front of a Bulldover driven by Hank himself (Hope you didn't load up on SDS before lunch on Friday). News is random, what is important in times like these is resolve and a good asset allocation strategy.
long: weekends
