New Indicator Trumps the VIX?

The current stock market danger meter is the VIX. Simply put when the VIX is high you buy when it's low you go (sell). Mianyville put out an interesting post on the subject during the week, in case you missed it here it is.

The index is calculated using 3-month SPX option prices to answer the following question: "If I sell a 10% out of the money call, what put can I buy with the premium I received?" If investors are very worried about a crash lower, you would expect to be buying a much-farther-out-of-the-money put in exchange for the call, while -- if investors are confident in a market rally -- you would expect to buy a closer to the money put.

For example, as of the close on Friday (April 9th), CSFB closed at 13.68, which means if you sold a 3month 10% out of the money call, you could buy a 13.68% out of the money put for zero cost. This index is effectively a 'tradeable' approximation of a zero cost collar.

It's essentially a measure of skew, and the degree to which that skew biases to the downside. In theory, the more someone wants to pay for downside relative to upside, the more fear. And that's something to fade.

Barron's interviewed the creators of this number:

Barron's: But upside and downside risk is elemental. That is why zero-cost collars are always packed with information. So what is CSFB saying, and how do we use it?

Tom: Say, for example, the S&P is at 855, and CSFB is at 13.68%. [If SPX changes, that percent changes.] That means if you are long the Standard & Poor's 500, you can initiate a zero-cost collar by selling a July 940 Standard & Poor's call that is 10% out-of-the-money, and buying a July 738 put that is 13.68% out-of-the-money. This implies investors are expressing relatively low levels of downside fear for the next 3 months.

Barron's: OK, but 13.68% still seems random.

Tom: Think of 13.68% (or any CSFB level) as an insurance deductible. It is how much the put is out-of-the-money. The market has to drop 13.68% before it kicks in.

Barron's: So if CSFB registers a high level like 20%, investors are afraid, and insurance is more expensive. If CSFB is low, insurance is cheap, and investors aren't.

Davitt: Right. We backcast CSFB to 1998. It is ranged from an October 2008 low of 10.5% to a March 2007 high of 30%."

Upon quick glance, I like the idea. But as augmentation to the VIX, and perhaps put/call as well. I think if you can correlate both volatility AND the degree to which that volatility implies piling into puts, you could get some nice info.

As with all numbers, though, I suspect the "absolutes" ultimately won't matter. Don't forget that, in 2008, the VIX doubled on numbers that once seemed extreme. If you created and ran this CSFB number before 2008, you might have thought something like an 18 or 20 showed extreme fear. And that if the CSFB kept flying and the market kept tanking like it did in 2008, Bloomberg would have run a piece about how the CFSB did not work any more.

So yes, I think this is an intelligently created index. Perhaps if it becomes popular, we'll find instances where it tells us some info that the VIX can't provide. But if there's a grand takeaway, it's that they're all just stats, best used together to create the broadest picture possible.

No one index, in and of itself, will ever be gospel.