As noted earlier in the week, the VIX's recent fall from its intraday peak of 56.74 on July 24 has been cited by many observers as evidence of renewed bullishness among investors. But things aren't always as simple as they first appear. "In no way do I see bullishness rising because of the fall of the VIX," said one options trader, who requested anonymity. The trader observed that a higher VIX, such as what occurred in late July, happens as mutual funds look to "dump stocks in a capitulatory manner." Brokerage firms with which the funds place those sell orders commonly buy puts ahead of those sales, he suggested. Hence, the VIX goes up. "Huge moves in the VIX only occur because major brokerage firms are taking down some size of stock," the source continued. "As the market lifts they sell their inventory and it is only natural for them to unload the puts as well." Without quibbling about the morality or legality of brokerages effectively front-running their clients (nor explaining why a proprietary trading desk might not use cheaper alternatives such as futures to hedge big inventory), he summarized thusly: "A rising VIX suggests only that the market has been going down. A spiking VIX suggests only that the market is experiencing massive but unsustainable amounts of selling. In short, don't overanalyze the VIX. It simply reiterates what is going on in the market." Maybe, maybe not. Either way, there's an ongoing debate about the meaning, or lack thereof, of the VIX's recent fall from its July heights. ( Tuesday, the CBOE Market Volatility Index rose 1.4% to 32.73 as stocks slipped.)
Nothing to Fear, but...
The VIX has become commonly known as an indicator of fear, and its recent decline has compelled some (bears) to conclude complacency is returning, even if indicators such as the Investors Intelligence survey and levels of short interest suggest otherwise. The VIX, however, literally measures the implied volatility of a theoretical 30-day, at-the-money S&P 100 index (OEX) option. Because it's a measure of implied volatility, not the actual stuff, the VIX is an imperfect indicator, critics say. Because it's a measure of theoretical contract, they also charge it's basically an extrapolation of trends in the two or three OEX contracts nearest to expiration -- extrapolation being an imperfect science for sure (which is why the track record of many economists is so "dismal," but that's another story). Furthermore, the OEX is "not nearly as active as it used to be," said David Lerman, a former options trader and current associate director of equity index products at the Chicago Mercantile Exchange. "All the volume, all the money and all the open interest" are in options and futures of the S&P 500, as well as the Nasdaq 100 Unit Trust (QQQ). Lerman was being a bit dramatic but OEX options are pretty far down the CBOE's most active list. While not advocating that investors ignore the VIX, Lerman also advised against reading too much into it. He also observed that while the VIX is down from its recent peak, it is "still high, statistically speaking." In March, I examined this argument, which basically states the VIX got inflated along with stocks during the 1990s bubble, and is going to revert to its long-term mean in the high teens to low 20s. Citing this, many optimists suggest the VIX's recent decline is thus not a sign of complacency's return. But that's "precisely, absolutely" what the bulls said in July and August 1981 and more recently in April and May 2002, according to Bernie Schaeffer of Schaeffer's Investment Research in Cincinnati. Schaeffer passed through San Francisco this week and (over lunch at the -- ahem -- Four Seasons Hotel) we chatted about the VIX, and why its recent history makes him skeptical the 5-week-old rally in equities is anything more than a bear market dalliance.
Bear Says Hogwash
Since the mid-1990s, the 20-week moving average of the VIX has ranged (broadly speaking) from 25 to 30, Schaeffer observed, agreeing a VIX in the low 30s is relatively high. But the mid-1990s were a decidedly bullish period for stocks, he recalled, while the more recent period has been far less kind. So shouldn't we expect the VIX -- if it really is a good measure of fear -- to be in a higher range? Using similar logic, "the July 24 low was not credible as the bottom because of the less-than-climactic fear levels it generated," Schaeffer recently argued at his firm's Web site. "Despite all the media trumpeting of a high VIX-based July bottom, the July VIX peak fell short of its Sept. 21 high" of 57.31. Arguably, there was more abject fear in the immediate wake of the Sept. 11 terrorist attacks; Schaeffer's Research concluded fear was higher in September but that true OEX volatility was higher in June and July. Furthermore, Schaeffer observed the September lows -- which were supposed to be the lows according to many observers -- were violated last month. Therefore, it's surprising that fear, as measured by the VIX, was not as high as in September. "This is of particular concern" given the now commonly cited argument that the July lows were the lows, he continued. "In other words, this is not a 'wall of worry' market; it is an 'accentuate the positive and ignore the negative' market. And this strongly implies that we experienced a bottom on July 24, but not the bottom." In sum, Schaeffer doesn't believe the VIX's recent peak was an extreme event nor is he so quick to dismiss the VIX's recent decline as insignificant or the result of technical factors. As for the VIX itself, the message that investors should take from this is that while the VIX works well as a contrarian indicator at market extremes, it's far from the Holy Grail of sentiment.