Sep 26, 2010

Week in Volatility

Beside market weakness on Wednesday and Thursday volatility indexes remain almost unchanged for the week. We're only few trading days away from October, but futures and skew in vol options remain cheap. The 22.5/25/27.5 fly I put on a while ago is not working well - even though the futures approached the sweet spot, marks have barely moved - probably because of the skew. Now I'm trying to roll the fly somewhere lower.

I am still on vacation, so there probably will be no other posts this week.

Sep 21, 2010

Crude Oil & Gold Volatility Futures to Trade on CME

After announcing new commodity volatility indexes back in March, CME announced they're actually launching futures and options on NYMEX WTI and COMEX Gold volatility in the fourth quarter. This is to be followed by launching volatility derivatives on Soybeans and Corn in the first quarter of next year.

There is no exact schedule of product launch - whether CME will launch futures first, and then introduce options, or launch futures and option simultaneously. I hope that the new products will succeed, although I have my reservation given many failures in volatility derivatives. What makes these contracts different is that GVX and OIV would be the first commodity volatility derivatives to be listed.

CME product webpage here, including historical data for all new indexes. Press release here.

Australian Volatility Index S&P/ASX 200 VIX

"The S&P/ASX 200 VIX will be an end-of-day index that reflects investor sentiment about the expected volatility in the Australian benchmark equity index, the S&P/ASX 200." The exchange will start disseminating the index (ticker XVI) on Sep 23, 2010. The index follows the same methodology as the VIX.

Exchange page on the index here, including historical data for the new index . Press release here.

P.S. Bloomberg symbol for S&P/ASX 200 VIX is SPAVIX.

Sep 17, 2010

VIX as a Commodity

There is an excellent article on commodities modelling that limns many of common features between VIX derivatives and commodity derivatives:

VIX - underlying not tradable / not possible to practically replicate
Commodities - all liquidity in the forwards, not spot

VIX - no pure arbitrage plays between different months
Commodities - limited arbitrage opportunities depending on storage

VIX, Commodities - stable long maturities, volatile near maturities

VIX, Commodities - volatility is backwardated

VIX - call skew, dominated by consumers
Commodities - call skew in products like NG

VIX, Commodities - spiking behavior, difficult to model with markovian models, requires extreme mean-reversion (much greater than in historical data), or regime switching (VIX seem to be well-partitioned into 3 regimes)

Sep 16, 2010

VIX Expiration, Interesting Action in VIX Binaries

$VRO printed at 22.97 - vs my forecast of 24.97, and market forecast of 29.44, here. September was a tricky month to trade - the swings in the beginning of the month, and rapid decline of VIX/VIX Futures made for a lot of excitement, but I had my best month in terms of PL. My forecast for October expiration (in 34 calendar days) is 22.62 with standard deviation of 5.18. Market forecast this morning is 25.55 with std of 6.19. I'm still trying to figure out my game plan for October - the sweet spot is at 25, significantly above my forecast, but about on par with the futures. On the other hand if volatility picks up (and there's plenty of time and fundamental weakness for it to do so) 25 VIX may seem low.

There was some interesting action on Tuesday in Binary VIX Options: Sep 20 Puts were (suddenly?) bid up. I saw bid at 0.07, to 0.10 200 contracts, which is big for binaries. VIX made a brief dip below 20 that day, but given there were just few hours to go these options seemed hugely expensive. VIX closed at 21.56 and puts close at 0.05 bid. My boss said that perhaps someone was taking profit, as there was 200+ open interest on the puts, however I don't think that was the case. I don't have historical data for options, however knowing VIX history I am pretty sure that that was the highest price for the contract.

Sep 11, 2010

Week in Volatility

This has been a short (US holiday) and rather boring week for volatility. VIX was practically unchanged, with futures falling slightly. VSTOXX was unchanged, however futures fell sharply with front three months declining by about 2 points, and bank months by about half point.
With front month futures risk premium decaying rapidly into expiration, and with my sweet spot at about 25 I had trouble re-hedging, and trying to shift everything lower. With only two trading days until expiration it is a fait accompli that September won't work out exactly as I planned. October futures closed at 27, and with October historically being a volatile month, October skew seems cheap. Is there a way to be short delta and long skew?

Sep 10, 2010

Weekly VIX Options

According to the Bloomberg report CBOE will start trading weekly options on the VIX. I was very excited about the news until I found the details on CBOE website here:
  • The options will trade on CFE, not CBOE
  • The options are American-style, exercised into near-term futures
  • Expire on Fridays, not Wednesdays
I think these points create unnecessary complexity, and I doubt the contracts will attract significant volume. This is very unfortunate, as I think short-term VIX plays could provide really interesting trading opportunities if the weekly options were the same as regular ones in terms of exercise and trading platform.

Sep 3, 2010

Volatility Cones, Support/Resistance for VIX Futures

These are maximum, average, and minimum values for VIX futures as a function of days until expiration. The left-most values are the values that correspond to expiration settlement values (VRO) , and the right-most are historical values of futures that expire in about six months.

The plot does not convey all the available data. Here is a different plot that is derived from the same dataset.

Black lines are like error bars - representing highest and lowest values of VIX futures. The red band inside is the middle 50% of the data, i.e. form first to third quartile. Since VIX futures were listed most of the trading action took place somewhere between 15 and 30. Most of the lower quartile action happened in 2004-2006, and most of the upper quartile happened in 2008.

With longer term VIX futures trading at over 30, and recent realized volatility at about 20% I think the futures will come down a bit to I personally have bearish bets in January - although the reason for trade was based on my forecasting model, and not on the analysis above.

Now here comes what I think is the most interesting part: same plot, now every historical futures value is a dot.

The VIX futures can be partitioned into three regimes. If that is not clear from previous plot - here is a histogram.

These three regimes are not present in the VIX index data, they are only seen in VIX futures. Using k-means of log values the centers are:
  • Low 14.55 +/- 1.60
  • Middle 24.31 +/- 2.83
  • High 37.19 +/- 6.45
There is (as expected) a positive correlation between VIX level and vol of vol. These centers, or regimes are a superior alternative to support and resistance levels derived from VIX index.