Aug 31, 2010

Timeline of Listed Volatility Futures

Update: The timeline has last been updated on Jan 15th, 2013

Right now there are only two indexes with listed volatility instruments - VIX and VSTOXX. There are talks, mainly press-releases from different exchanges about plans to list futures on local vol indexes (VKOSPI futures, India VIX futures, Australian VIX futures, Japanese VIX futures), however I think we're a very long way from seeing another successful volatility instrument. The reason is that historically there just has not been enough interest from traders - just look at this timeline of volatility futures:
  • Jan 1998 DTB lists VOLAX futures on DAX
  • Dec 1998 DTB delists VOLAX futures on DAX
  • Mar 2004 CBOE lists futures on VIX
  • Apr 2005 CBOE lists DJIA volatility futures
  • Sep 2005 Eurex lists futures on VDAX, VSMI, and VSTOXX
  • Feb 2006 CBOE lists VIX options
  • Jul 2007 CBOE lists Nasdaq and Russell 2000 volatility futures
  • Mar 2009 CBOE lists mini-VIX futures
  • Jun 2009 Eurex lists MINI-VSTOXX futures 
  • Jul 2009 Eurex delists VDAX, VSTOXX, VSMI futures
  • Aug 2009 CBOE delists DJIA volatility futures
  • Feb 2009 CBOE delists Nasdaq volatility futures
  • Feb 2010 CBOE delists Russell 2000 volatility futures
  • Mar 2010 Eurex lists options on VSTOXX
  • Mar 2011 CBOE lists futures on GVZ (Gold)
  • Mar 2011 CBOE lists options on GVZ (Gold)
  • Jun 2011 RTS lists futures on RTSVX
  • Jan 2012 CBOE lists futures and options on VXEEM (Emerging Markets)
  • Feb 2012 CBOE lists futures VXEW (Brazil)
  • Mar 2012 CBOE lists options on VXEW (Brazil) and futures on OVX (Oil)
  • Apr 2012 CBOE lists options on OVX (Oil)
  • Feb 2012 HKFE lists futures on VHSI
  • Feb 2012 Osaka lists futures on VNKY

Aug 29, 2010

Week in Volatility

It was a bumpy ride for the VIX (and my portfolio) this week. The index rose above 28 on Tue and Wed, and fell below 25 on Fri. The week to week change in the equity and volatility indexes was completely overshadowed by intraweek action. VIX and VSTOXX futures declined in the near term, however the long-term levels were unchanged. With volatility rising and falling so much in one week, it seems to be a good time to rethink the forecasts. My model still shows Sep expiration forecast of 25.36 (higher than my previous forecast) with volatility of 3.85 (lower than my previous forecast).

Aug 26, 2010

Simple Trick to Convert Volatility

My boss came to me today with "how do I convert this monthly vol to annual?". This is by far not the first time someone asked me to convert x-period volatility to y-period volatility. If you're having troubles doing this, here is a simple trick to remember: figure out the multiplier as if scaling were linear, and take the square root of that multiplier. For example, returns scale linearly with time. If someone gives you a monthly returns to convert to annual, you would simply multiply it by 12. So, in the case of converting monthly to annual volatility multiply it by √12. If someone gives you annual returns and asks you to calculate daily returns you would divide it by 252. To convert annual volatility to daily volatility divide it by √252.

P.S. √252≈16, which is the reason for "rule of 16" for converting daily to annual volatility



Aug 22, 2010

VIX Trading Strategies from Jeremy Wien

Mr Wien published two excellent articles on VIX strategies. Here is the presentation he did for CBOE last April, and here is an article he wrote for Active Trader magazine, published this May.

The strategies he recommends:
  • Short straddle + hedge upside risk with OTM call
  • 1x2 put spread
  • Butterfly
  • Bull spread + long call
Here are some illustrative charts; the discussion of the strategies in the articles is far better than what I can repeat here. The only thing that I can add here is that the first strategy is structurally similar to a fly, since downside risk in the VIX is minute compared with the upside.

    Aug 21, 2010

    Week in Volatility, Forecasts Review

    This was a mixed week for volatility - the short term vol was all over the place, although mostly higher. Long-term vol was definitely on the upswing - the long-term futures being up about a third of a point. VSTOXX futures mostly rose, moving an average of 0.13 points. Overall volatility indexes around the world were down half of a point. 
    Now, about a month ago I made a number of predictions about IWM based on its implied volatility; let's see how they worked out.

    1 Prediction: volatility 2% per day, realized: 1.8% Implied volatility on indexes is usually higher than realized, but this month the difference is not huge. It is not unusual to see all implied risk measures to be higher than the realized. 

    2. Prediction: daily range $2, realized: $1.48.

    3. Prediction: monthly range $10.42, realized: $7.13. Monthly high prediction: $69.19, realized: $67.27. Monthly low prediction: $58.77, realized: $60.14. 

    4. Maximum drawdown: $8.34, realized: $7.13.

    Do these finding mean that the formulas are accurate enough to be used for trading decisions and quick assessment of risk? I think after applying proper implied/historical adjustment the formulas are very helpful, and that is how I use them in my own trading.  

    Aug 19, 2010

    VIX Expiration

    $VRO settled at 24.82, only 2 cents away from my last-month forecast of 24.80, here. Although I wish I could attribute such amazing accuracy to my forecasting skills, luck has played a significant factor. My forecast for September is 24.97 with standard deviation of 5.45, vs market's forecast of 29.44 with standard deviation of 7.45. I have a lot of concern regarding skew trades into Sep, and especially Oct expirations, so plan to proceed more cautiously. Good luck traders, and hedge your deltas!

    Aug 18, 2010

    AlphaShares China Volatility Index

    There is a (relatively) new volatility index, that I think does not get as much attention as it should - the AlphaShares China Volatility Index, ASCNCHIX, or CHIX. I won't pretend to be an economist, but Chinese volatility is an important factor to consider for all traders and investors. The index is also interesting because it is actually composed of two underlying indexes.

    The description on Bloomberg website here is actually somewhat misleading: "The AlphaShares Chinese Volatility Index measures the implied volatility of options on the FTSE Xinhua China 25 and Hang Seng (HSI) indices." In a private email Alphashares representative explained that the index is an equal-weighted average of  ATM volatilities of FXI ETF (US) and HSI (HK) .

    Over the last five years VIX and CHIX moved in tandem, with CHIX about 1.5 times the VIX. However in the last 3 months this ratio fell below 1, with CHIX trading below VIX.



    Aug 15, 2010

    Week in Volatility

    Markets fell this week, with S&P down almost 4% and STOXX falling 2.5%. VIX rose 4.5 points; Jan-11 up 2.1 points, Feb-11 up 1.7 points. The rise in long-term vol occurred overnight from Tue to Wed. STOXX did not fall as much as S&P, and rise in VSTOXX was also less pronounced - index up 2.26, and long-term futures up 1+ point.




    What is really interesting is that while VIX and VSTOXX rose this week, other volatility indexes around the world fell. I did not have the time for thorough data analysis, but NSE India VIX index made an all-time low, closing at 16.74 on Fri, link. Keep in mind that the index was in existence for a relatively short time. I was able to put together two charts of Citi Asian volatility indexes (that have longer history than India VIX) and VIX.
    Over the past five years most of Asian markets had higher implied volatility than the VIX.


    However since the vol spike in May most Asian volatility indexes have declined much more than the VIX, closer to Apr low levels.

    Aug 13, 2010

    Understanding VIX Futures Movement, Part 3

    Part 1 here, part 2 here. In the previous post I described decorrelation of VIX futures over time. In this post I will discuss the the decay, or theta of futures.

    Surlytrader writes "Consider this the cost of holding the long position over time. This is important because we can think about VIX futures trades much like we think about calendar spreads in the options world. In a calendar spread you buy a longer dated option and sell a short dated option. You hope that the short-dated option decays (loses value) quicker than the long-dated option. In addition you hope that your purchased long-dated option covers you against adverse movements on the short position in the short-dated options. "

    Indeed, VIX futures do behave like options. Unlike "regular" index futures, VIX futures are non-linear in the index. That means that besides delta VIX futures also have gamma and theta. The last one is the most obvious, and is indeed observed in the marketplace. As one can see, most of the price decay occurs closer to expiration. The applet below demonstrates VIX daily decay as a function of different parameters.


    If you do not see the applet above you may need to download shockwave player from Adobe (same company that makes Flash player, and Pdf reader)

    Aug 10, 2010

    Aug 9, 2010

    Understanding VIX Futures Movement, Part 2

    Part 1 here, part 3 here. Everyone knows that VIX futures (and ETFs) do not move one to one with the index. Given the complicated dynamics of the index - jumps, multiple regimes, mean-reversion - it is actually not surprising that replicating the index in a tradable instrument is not trivial. However there's very little info about quantifying the relationship.

    In this excellent article surlytrader talks about imperfect correlation between VIX futures and the index, writing "... VIX futures become less correlated to the spot VIX index as you move further out in expirations. ... By investing further out on the curve, we do not capture as much of the movements of the VIX spot index. On the flip side, by investing further out on the curve, we lose less as the futures contract ages." Surlytrader also provides a plot with schematic depiction of correlation declining with time to maturity. In the plot he calls it VIX Beta, but I thought that VIX Delta would be more in line with the convention.

    To explain the relationship I created an applet that quantifies and illustrates this dynamic relationship between futures prices, tenor, and correlation to the index. One of the most intuitive features is the effect of mean-reversion parameter on the price and delta. When mean reversion is large VIX futures converge quickly to their long-term level, and delta quickly decays toward 0. When mean-reversion is small (e.g. like it is in the stock market) then more familiar dynamics emerge - futures are "connected" to the index level, and correlation is high.



    If you do not see the applet above you may need to download shockwave player from Adobe (same company that makes Flash player, and Pdf reader)

    One of the practical applications of VIX futures delta is trading calendar spreads on VIX futures /options / ETFs, and measuring position risk.

    Aug 7, 2010

    Week in Volatility

    S&P and STOXX were up this week 1.8% and and 1.3% - pretty much non-event except for the roller-coaster on Friday. VIX and VSTOXX futures moved lower by about 0.75 points. Meanwhile I came across this article indicating that prices for long-term vol derivatives are historically very high. If anyone knows where I can get more data on this, please email me.





    Like I wrote earlier the term-structure of VSTOXX futures is much flatter than VIX, and I believe it provides an excellent opportunity for relative trade.

    Aug 6, 2010

    Understanding VIX Futures Movement

    This is the first post in the series; part 2 here, part 3 here. I recently came across this blog post from Lawrence McMillan:

    "The advantage of buying, say, September VIX puts is that the huge 6.87 point differential between VIX and the Sept VIX futures will have to disappear by September VIX expiration (in eight weeks). Remember that the Sept VIX puts are priced off of the futures, not off of VIX. So, if that 6.87 differential shrinks by the September futures falling from their current price near 31 to the current VIX price near 24, those puts would profit handsomely."

    He's right that the future is most likely to decline and meet the index than other way around, but his explanation is ambiguous. Considering mean-reversion effect by itself we should actually expect the opposite - VIX index wobbles around all over the place but eventually comes back to some steady level. So in our case we should expect VIX to rise up to the futures level. This is a valid hypothesis, however most of the time VIX futures display an upward slope, so mean reversion cannot be the only factor.

    The explanation why VIX futures are most of the time higher than the VIX, is in convexity premium. The distribution of VIX levels is very right-skewed, and that skew demands premium. Every day that premium is getting smaller, and futures get cheaper. Meanwhile if VIX is below some level it will move up, if it is above that level it will move down. This interplay between mean-reversion and risk is what drives the complicated dynamics of VIX futures.

    When I have more time I will try to post some charts, and more quantitative description of the dynamics.

    On a separate note here is a very thorough VIX introduction from Mr McMillan.

    Aug 2, 2010

    VSXX - VSTOXX ETF Performance Very Different From VXX

    Edit Dec 7, 2010: See my updated post on VSXX here.

    VSXX, an ETF based on VSTOXX Short-Term Futures Index is trading for two months now (summary here, official Braclays page on VSXX is here, and here is the explanation of the index the ETF is based on). The VSTOXX futures trade in €, but as far as I understand European traders have two options as ETF trades in London in £, and in Germany in €. Below is the plot of VXX ($) vs VSXX (€)



    I've read the description on both indexes and calculation algorithm seems to be the same for both. However the most interesting nugget of information is found in the summary on Deutsche Börse website here, page 5 figure 2. The chart compares performance of two volatility indexes - VIX and VSTOXX vs. their short-term futures indexes. The cost of carry, that eroded VIX futures index to about a third of its starting value in 4.5 years had almost no effect on VSTOXX - based index. Below I reproduce the charts from my data. SPVIXSTR is S&P 500 VIX Short-Term Futures Index, and VST1MT is Euro STOXX 50 Volatility Short-Term Futures Index.





    I was going to type up some stats based on the data above, but picture's worth a thousand words. Over the last year and half VXX declined precipitously, however such decline is probably not in the cards for VSXX. While the term stucture of VIX futures is rather steep, making the roll costly, VSTOXX term structure appears to be relatively flat.

    Honestly, I cannot say that I know the reason for that. One possible explanation is that I made a mistake , and there is a critical difference in the calculation of two indexes that escaped my attention. That seems unlikely, as Barclays' publication obviously implies that the indexes are comparable. Another explanation can be that there is a greater institutional demand for VIX futures than there is for VSTOXX futures, however to confirm that I would need to separate the effects of risk premium from statistical effects that also produce upward-sloping futures curve. Accomplishing that would be anything but simple.

    Leaving theoretical issues aside it appears that smart traders can take advantage of different futures slopes by rolling into the contracts with the cheapest cost of carry, possibly financing the trade with selling the expensive roll. For example consider an investor who is long a broad portfolio of US stocks and wants to hedge against volatility spike. Holding VXX or rolling VIX futures can be very expensive in the long run; doing the same thing with VSXX / VSTOXX futures may be cheaper, while providing a proxy for VIX in the time of volatility spikes.

    Edit Dec 7, 2010: See my updated post on VSXX here.