Persistence in Volatility

4-volatility persistence


The VIX has been very high last few months and at 35 today. Some talking heads on CNBC have been suggesting that this is the time to buy momentum given elevated volatility. I tested the data and find no relationship between past volatility and near-term future returns. I do find volatility clustering effects, which is very high, but this in itself is non-sequitur to a suggestion for entering into equity momentum trades.

Historical volatility measures can forecast future volatility, albeit with some error. Volatility clusters when periods of high volatility followed by periods of high volatility, and periods of low volatility tend to be followed by periods of low volatility. We also tested with VIX, a  forward- looking measure, to find materially similar outcome. Constant volatility exposure can be arrived by de-leveraging when past volatility was high and leveraging when past volatility was low. No relationship between past volatility and near-term future returns.

Author: Sameer_Jain

Partner. Sameer Jain is founder of FinTech, the world’s first portal that seamlessly integrates traditional, illiquid and alternative investments within portfolios. Prior to this he was Chief Economist & Managing Director at AR Capital. Before that he headed Investment Content & Strategy at UBS Alternative Investments. At UBS, he served as a non-voting member of the Wealth Management Research investment committee, and as a capital allocator was responsible for all illiquid investing including fund manager selection and due diligence across the platform. Prior to UBS he headed product development & investment research at Citigroup Alternative Investments that managed over $75 billion of alternative investments across hedge funds, managed futures, private equity, credit structures, infrastructure and real estate. Here he led a team that developed proprietary models for portfolio strategy and asset allocation with alternative investments, provided investment support and research to pension plans, sovereign wealth funds, endowments as well as internal clients including Citi Private Bank. Before this he was with Cambridge Alternative Investments and SunGard (System Access) where he travelled to over 80 countries for work across Europe, Asia, Middle-East and Africa. He has written over 30 academic and practitioner articles on alternative investments with thousands of downloads at SSRN, presented at over a hundred industry conferences and has coauthored a book, Active Equity Management. Mr. Jain has multiple degrees in engineering, management, public administration and policy and is a graduate of Massachusetts Institute of Technology and Harvard University. He is a recipient of the Alfred Sloan Fellowship and subsequently was a Fellow of Public Policy and Management at the Harvard Kennedy School of Government for a year. He holds Series 7 and 66 securities licenses.

One thought on “Persistence in Volatility”

  1. Hi Sameer — I think any prediction based on a relationship between two variables is doomed to fail. The market is clearly more complex than that regardless of the “fine structure” (volatility clustering) superimposed. Moreover, such relationships would have been identified long ago. Machine learning may (although I have my doubts) be able to overcome the first limitation, but is still susceptible to the second. Am I agreeing with your conclusions?

Leave a Reply

%d bloggers like this: