In the first two months of 2015, the CBOE VIX of VIX Index (VVIX) had an average daily close of 100.2, a higher level than in any of its previous eight full calendar years from 2007 through 2014.
On Thursday at the 31st CBOE Annual Risk Management Conference in Carlsbad, expert presentations on Volatility of Volatility (VOV) were delivered by Benn Eifert, Ph.D., Portfolio Manager, Mariner Investment Group and Kambiz Kazemi, Portfolio Manager, Picton Mahoney Asset Management.
INTRODUCTION TO INDEXES
Before highlighting the remarks of the speakers, here is an introductory overview of two key indexes that measure 30-day expected volatility – the CBOE Volatility Index® (VIX®) and the CBOE VIX of VIX Index (VVIX). The VIX Index measures expected volatility that is conveyed by prices of S&P 500 (SPX) options, while the VVIX Index measures expected volatility as conveyed by VIX options. On Dec. 12, 2014, the VVIX Index closed at 138.60 its highest daily close in more than four years.
In analyzing the weekly returns for the period of June 13, 2008 through Feb. 27, 2015, the VVIX Index had a correlation of negative 0.47 versus the S&P 500 Index, and a correlation of positive 0.76 versus the VIX Index.
TOPICS COVERED BY THE SPEAKERS
Dr. Eifert and Mr. Kazemi covered these topics —
- Understanding the distributions of realized and implied volatility
- Key differences between volatility of volatility and volatility of spot
- Structuring trading and hedging strategies using a variety of instruments
- Impact of trading flows on volatility of volatility levels
PRESENTATION BY BENN EIFERT
Dr. Benn Eifert noted that –
- Basic considerations for volatility-of-volatility include —
- Future volatility itself (both implied and realized) is uncertain
- The market prices uncertainty of both implied and realized volatility
- Some recurring questions on volatility-of-volatility —
- “How does vol-of-vol ‘perform’ over time?”
- “Does implied volatility-of-volatility predict future realized volatility-of-volatility?”
- “How persistent is volatility-of-volatility over time?”
- “VVIX is low, should I buy some VIX options?”
- “Is vol-of-vol high/low?”
- “Is vol-of-vol cheap?”
- Realized volatility varies dramatically over time, and more so the shorter the trailing window
- Realized volatility measured over shorter windows has fatter-tailed distributions because extremely low or high volatility is more likely over a month than a year
- How do the VIX Index and other implied volatilities move over time?
- Implied volatility of shorter-maturity instruments is more volatile than long-dated implied volatility
- Changes in implied volatility over longer periods of time have a larger range but relatively thinner tails
- The distribution of changes in implied volatility grows more slowly than the square root of time
- This is very different than for spot prices, where the annualized volatility of daily returns and longer horizon returns are much closer
- Trends in implied volatility of implied volatility –
- VVIX has moved towards the higher end of its range recently, and generally has moved higher since 2008
- Normal implied volatility-of-volatility (in vegas per day) has had some modest spikes but is still relatively low
- Reflects some differences of perspective… is a move from 12 to 15 a very large move in implied volatility?
- Can implied VOV be used to help predict future realized volatility of implieds?
- This depends on how you think about and measure volatility-of-volatility…..
- Lognormal (%) implied V.O.V. does not predict lognormal realized volatility well
- The normal-normal prediction is much better
- How persistent is realized volatility-of-volatility over time?
- This depends on how you think about and measure volatility-of-volatility….. lognormal (%) realized VOV is very quickly mean-reverting, while normal (vega-per-day) realized VOV is more persistent over time.
- How persistent is implied volatility-of-volatility over time?
- This really depends on how you think about and measure volatility-of-volatility….. Lognormal (%) implied VOV is very quickly mean-reverting, while normal (vega-per-day) implied VOV is more persistent over time.
- Should an investor consider buying VIX options when VVIX is low?
- If VVIX is rapidly mean-reverting, and so is realized vol of implied vol, it’s plausible that the level of VVIX would be a good signal for buying and holding VIX straddles. However, the historical data is less clear… all the big wingy payouts from owning VIX straddles came when VVIX was at moderate levels.
- We’ve recently been in a regime of high implied and realized volatility-of-volatility, compared to other risk factors. There are some structural reasons to think it may remain in place until dislocated by a large and persistent equity market selloff. If the regime shifts back, vanilla equity puts are a great value right now.
- As with S&P volatility, it’s not as simple as “just buy the vol when it’s low.” However, this is driven in part by major crises, where the big selloff happens after an initial period of nervousness.
- Optionality on implied and on realized volatility are distinct and fit different needs. In the S&P, implied volatility has had more violent moves than realized volatility in the recent regime (not necessarily true of SX5E and NKY though). Investors’ objectives factor in here too.
PRESENTATION BY KAMBIZ KAZEMI
Kambiz Kazemi noted that —
- Every option book has exposure to vol of vol (i.e. gamma of gamma or 4th moment)
- Vol of vol is the sensitivity of out-of-money options in relation to those at-the-money
- Why we increasingly hear about vol of vol —
- Because vol of vol is directly related to “convexity”
- Because vol of vol is directly related to “tails” (post-crisis everyone cares about “tails”)
- Because VIX options are now being used by a wide range of market participants, and vol of vol matters in pricing and trading these options
- Vol of vol is a direct way to hedge or get exposure to “convexity”
- The advent of volatility products has made vol of vol accessible to many participants and is efficiently tradable
- The liquidity in the VIX complex has made many previously theoretical strategies a possibility
- Trading instruments with vol of vol exposure requires some preparation: understand pricing and sensitivities
- More products will (hopefully) emerge and complete the market (such as Mid-curves, spread options, etc.)
ABOUT THE SPEAKERS
Benn Eifert, Ph.D., is a Principal and Portfolio Manager for Mariner Coria, an investment team which focuses on relative-value strategies in equity, foreign exchange, and commodity derivatives markets. Mr. Eifert joined Mariner Investment Group in 2014. Prior to Mariner, Mr. Eifert was at Wells Fargo / Overland Advisors in San Francisco, where from 2010 to 2011 he was Head of Quantitative Research and from 2011 to 2013 a trader for the Derivatives Relative Value strategy. From 2003 to 2005 he was an economist in the Office of the Chief Economist at the World Bank in Washington, DC. He holds Ph.D. and M.A. degrees in Economics from the University of California, Berkeley, where he also taught in the Masters in Financial Engineering program, and B.A. degrees in Economics and International Relations from Stanford University.
Kambiz Kazemi is Portfolio Manager at Picton Mahoney Asset Management. After an engineering-oriented career, Kambiz joined the investment industry in 2001. Prior to joining Picton Mahoney Asset Management, Kambiz was a portfolio manager responsible for volatility arbitrage and derivative-based strategies at Polar Securities. His previous experience also includes developing derivatives-based strategies and trading ideas as a Senior Vice President at Newedge (formerly Fimat). He is a graduate of Ecole Nationale Superieure d’Electricite et de Mecanique de Nancy. He received an MBA from McGill University and is a CFA Charter holder.
To learn more about the CBOE VIX of VIX Index (VVIX) and more than 30 other volatility indexes, please visit www.cboe.com/volatility.