Correlation, Causation, and Expiration
Mon, Nov 24, 2008 | Frank C.
We were prepared to let this one go—that is, until yesterday after hearing the latest news report attributing Friday’s final-hour rally to Obama’s choice for Treasury Secretary. Let’s just say there’s a limit to the amount of gross oversimplification one can take. A Reuters item from Friday evening typifies what Joe the Investor was getting from the media:
Markets shot higher around 3 p.m. when NBC news reported that Timothy Geithner, president of the Federal Reserve Bank of New York, would be nominated as U.S. Treasury secretary, driving the Dow and the S&P up more than 6 percent.
Maybe the news triggered some buying, but all we really know is, as TheStreet.com’s version of the story put it, the rally “coincided [emphasis added] with reports that New York Federal Reserve leader Tim Geithner is likely to be nominated as the next secretary of the Treasury.” We’ve pointed out that correlation is not the same as causation, but what if we give the benefit of the doubt and say that the cause of the rally was the nomination news—does that 6% climb represent the market’s estimate of the “value” of that news? Hardly.
On any given day, short-covering can magnify the effect of any unexpected reversal—but Friday wasn’t just an ordinary day. Imagine you’re a market maker in SPY options and you have large short positions in November calls, which expire Saturday but, essentially, are settled at Friday’s closing price. The S&P 500 index ran into selling around 770 (SPY 77) twice earlier in the day, and just before 3:00 pm we’re 20 points below that. You know that recent sessions have been volatile in the last hour, so you’re hedged for a rally up to 77.50, maybe 78, into the bell. And after that bell, you’re off the hook, and you’re ready to go home.
But at 3:00, speculators suddenly start buying—it doesn’t matter why—and push SPY up more than a dollar in just 5 minutes. . .maybe you’d better cover more of your short 77 calls. In doing so, however, you’ve helped drive SPY up to 78 by 3:15, and now you’re having to hedge more of your 78 short position. The market reaches equilibrium at 78 and pulls back, but a bounce off 77 makes it look like 78 could be broken. . . and then 79, and then 80, fueling more and more buying to cover calls that minutes before looked like they were going to expire out of the money.
The point of this little thought experiment is that options expiration makes any correlation-based “analysis” even more unreliable than it already was—and, therefore, it’s especially important to take expiration-Friday reportage with more than a grain of salt.


0 Comments For This Post
3 Trackbacks For This Post
November 26th, 2008 at 7:09 am
[...] Condor Options simultaneously tackles a pet peeve of mine, Bubblevision Induced Lazy Correlation, and a favorite topic, the possible outsize magnitude of an expiration day market move. On any given day, short-covering can magnify the effect of any unexpected reversal—but Friday wasn’t just an ordinary day. Imagine you’re a market maker in SPY options and you have large short positions in November calls, which expire Saturday but, essentially, are settled at Friday’s closing price. The S&P 500 index ran into selling around 770 (SPY 77) twice earlier in the day, and just before 3:00 pm we’re 20 points below that. You know that recent sessions have been volatile in the last hour, so you’re hedged for a rally up to 77.50, maybe 78, into the bell. And after that bell, you’re off the hook, and you’re ready to go home. [...]
December 1st, 2008 at 12:09 pm
[...] tend to pay a lot of attention to expiration week, and for good reason. There are also some post-expiration phenomena worth noticing: in particular, unsophisticated [...]
January 5th, 2009 at 11:24 am
[...] Correlation, Causation, and Expiration [...]