Archive for March, 2011

Rise and Fall of VIX

Tuesday, March 22nd, 2011

The action in the VIX appears to be unprecedented.  It skyrocketed up +46.4%, then fell -29.9%.

IMAGE  Rise and Fall of VIX

What I want to point out about the chart has nothing to do with what it means as far as the market is concerned, at least not yet.  To me, it’s the perfect portrayal of why we use logarithms in financial mathematics.  The move in VIX carried the measure from 20.08 up to 29.40 and then back to 20.61: almost a perfect round trip.  The percentage moves, however, were starkly different: +46.4% compared to -29.9%.

When you put those percent moves in log terms, the numbers are almost perfectly inverse: +0.381 compared to -0.356.

Bottom line, a 25% gain is not the same thing as a 25% loss.  And a 50% gain is not even in the same ball park as a 50% loss.

Now for some history.  I did a quick analysis with the log numbers to see how frequently you get VIX moves like you did the past six days, and it only happened twice: August 1990 (Saddam invaded Kuwait) and May 2010 (right after the flash crash).  In both instance, the market rally that caused VIX to fall so far ran out of steam.  The major indexes made new lows before rallying.

I then did the same analysis using old VIX, now known as VXO.  It showed the same kind of whipsaw action as VIX did in May 2010.  But there was also a whipsaw in April 1994.  And there were some gigantic whipsaw swings in October 1987, so that month may not be a fair comparison.  At any rate, in 1994 and 1987, the market eventually fell back to the old low before rallying.  Interestingly, stocks did not make new lows.

– Don

Risk and Uncertainty: There is a Difference

Tuesday, March 15th, 2011

FTAlphaville is one of my favorite sources of investment information.  Today, however, they blew it.  They got caught up in numbers at the precise moment that numbers have little meaning.  Here are two articles that I would contend are not just worthless drivel; they’re dangerous because they lead people to the illusion that numbers always matter and patterns are always repeated.

Nuclear meltdowns and the S&P 500

Another market overreaction? 

If there is one thing that I wish I could get through to people, including some very bright people that I know and work with, there is a major difference between risk and uncertainty.

With respect to financial markets, risk is measurable and you can assign probabilities to different levels of risk.

Uncertainty, however, is immeasurable.  It’s impossible to assign probabilities.

Fortunately, for most of our investment lives, we’re only forced to deal with risk.  There are, however, times in which we have to deal with uncertainty.

For the Japanese stock market, I would contend that since Friday, we’ve been dealing with a period of uncertainty.  That is, there is no possible way to quantify the impact of the earthquake and the tsunami at this moment.  To borrow a phrase Donald Rumsfeld, there are known unknowns (we know the nuclear reactors are in trouble, but have no idea what the final outcome will be) and unknown unknowns (will there be any more aftershocks?).

The whole thing is unquantifiable, and anyone who presumes to know is a liar.  They’re either lying to those who are listening (like business news reporters desparate to get in on the story), or they’re delusional and lying to themselves.

I would even make the case that the uncertainty has cascaded to the U.S. market, at least with respect to some stocks that do significant business in Japan, either as a seller to that population or as a source of supplies.

My main point, however, is to provide you with a word of warning: sometimes there are events that make market math useless.  Statistics simply aren’t applicable because statistics depend on relevant history and overly simplified mathematics that are reasonably good estimates most of the time.  But not all the time.

– Don