More debunking. Can’t help myself. Prolific blogger Adam Warner (not shown, and definitely not the one getting debunked) already covered this a bit, but I thought I’d toss in my two cents … well, maybe three or four cents.
Â
Yesterday, Bloomberg/Business Week published an online article about volatility skew. Ominously titled, Options Show S&P 500 Profit Shortage Amid Stock Gains, the article pointed out that:
The options market shows traders preparing for losses. Implied volatility for S&P 500 puts expiring in July is 68 percent higher than calls that pay off when the index rallies, according to contracts priced 10 percent above and below the index. That’s the biggest ratio since June 5, 2008, three months before New York-based Lehman Brothers Holdings Inc. filed the biggest bankruptcy in U.S. history.
The gap between puts and calls jumped to 59 percent on Aug. 16, 2007. Two months later, the S&P 500 began a 57 percent plunge through March 2009. Bearish contracts climbed to 56 percent above bullish ones on Feb. 26, 2008, preceding a 7.8 percent loss in two weeks.
For those of you new to options, the implied volatility of options varies at different strike price levels. For example, with a stock at 100, the volatility of the at-the-money 100 strike options may have a different implied volatility than the 90 strike options, which may also be different than the 110 strike options. This difference is called “skew” or “smile”.
The skew/smile varies for each individual underlying asset. The skew for the overall S&P 500 index options can be completely different than the skew for IBM. Also, the smile for the same asset can vary day-by-day. For instance, the skew for the S&P 500 index options can be completely different for April 2010 compared to April 2009.
That last example brings us full circle back to the Bloomberg article.  The writers note that the implied volatility of deep-out-of-the-money S&P 500 index options is severely skewed. That is, the implied volatility of the July put that is 10% out-of-the-money is much higher than the implied volatility of the July call that is 10% out-of-the-money.
Let’s look at the numbers. The S&P 500 closed Friday at 1194.37. If we go up 10%, that’s 1313.81. The July call closest to that price is the 1325 strike. Let’s take the average of that option’s bid/ask, and then calculate the implied volatility of that average. It’s 12.38. Now let’s look at the put. If we go down 10%, that’s 1074.93. The July put closest to that price is the 1075 strike. The implied volatility of that options average of the bid/ask is 21.72. The implied volatility of the put is 75% higher than the implied volatility of the call.
Here’s a graph of the current Volatility Smile/Skew from a new feature in the upcoming version of ODDS Online:

That 75% ratio is slightly different than the 68% Bloomberg cites.  The minor difference could be attributed to the type of model used (binomial, Black-Scholes, etc.), whether they used just the bid or just the ask, what day they assume the S&P 500 index options expire (remember, they stop trading on Thursday, the values are determined by how the constituent securities open on the following Friday, and the expiraiton actually takes place that Saturday) or how they factored in dividends. The point is, it’s not a significant difference for our purpose here.
Now let’s look at 2009. Here’s the same smile/skew chart from last year:
 
The implied volatility of the July call that was 10% out-of-the-money was 32.15%. The implied volatility of the July put that was 10% out-of-the-money was 40.42%. The put implied volatility was only 26% above the calls back then.
So Bloomberg is right. The ratio is much greater now than then. Here’s were I have an issue with them. They don’t ask why! They didn’t look at the underlying cause of the ratio’s increase. Because if they did, they would have noticed something.
Look at the difference between last year’s skew numbers 40.42 and 32.15, and the difference between 21.72 and 12.38. Last year, the difference was 8.27. This year, the difference is 9.34. That doesn’t seem to be such a big deal, does it? The difference barely budged.
Put bluntly, the entire reason that the ratio has gone up is almost entirely due to the fact that volatility as a whole has gone down! That’s because the denominator in the fraction went from 32.15 to 12.38, while the numerator declined by almost the same amount.
So this high ratio — the ratio that is telling the folks at Bloomberg that the S&P 500 options are showing a profit shortage — is really only telling us that volatility is a lot lower now than it was a year ago!
This is not the first time this month that someone has raised an ominous voice due to the denominator. The ISEE Index, which calculates retail call buying volume divided by retail put buying volume (nice that they scrub out all option selling and all institutional trading), hit a very high level last week. Some claimed that meant there was a call buying binge.
Click image to enlarge
But a look at the details tells us that the call volume really wasn’t that unusual. What was unusual was that put buying dried up (which also helps explain why implied volatility is so low). That is, the denominator dwindled while the numerator remained unchanged. That’s why the ratio spiked — not because call buying increased.
I have to admit that I didn’t bother taking the time to write about it last week, because, well, I was busy. But the fact that we’ve had these two denominator-related issues was too much.
One final thought. None of this is to suggest that the market is in the clear and that it’s smooth sailing ahead. It’s just that these two sentiment indicators do not in any way tell us that the options are showing a “profit shortage” or that retail option traders went on a “call buying binge”.
Instead, they are a reflection that both volatility and volume are shrinking as investors shun options.
Hey, maybe that is an ominous sign!
– Don








Don,
The problem is that Jeff Kearns knows little about options, is unwilling to learn (or even discuss matters), and blindly goes on his way with articles that contain misinformation.
Why Bloomberg allows that to happen is a mystery.
[...] This post was mentioned on Twitter by Jared Woodard, Raj Gupta. Raj Gupta said: RT @CondorOptions: Don Fishback with some more debunking of that half-baked Bloomberg $SPX skew article. http://bit.ly/b44QKK $$ [...]
Once again another superbly crafted set of insights, Don.
Cheers,
-Bill
Good article. Kearns article june 7 says that the 10 day average skew has exceeded 50 percent 34 times since 1996 and in those cases the 500 has gained 7.2 percent in 6 months. Does this have meaningful value in timing?
The best and worst calls are dramatic.