Saturday, April 7, 2012

The VIX as an Equity Hedge

Back in January I posted a pair of VIX / S&P 500 tables that outlined the performance of the VIX and the S&P 500 given recent changes in the VIX, as well as the current level of the VIX. In response, reader Nazumo asked:

In service of the perpetual quest to find cheap and reliable hedging vehicles, I'd be curious to see a third table: (change in SPX / change in VIX) against VIX.
That table is below, but first I'll try to explain in a simple manner exactly what we're looking at.

The x-axis shows how much the VIX has changed over the last month (as of Thursday, this would read a '0 to 2.5 point drop' as the VIX reads 16.70 vs 18.05 a month back), while the y-axis shows how much the S&P 500 has changed over the last month in percent terms (as of Thursday, the S&P 500 would read '2.5% to 5%' as the S&P 500 was up around 2.6%).

That gets us to the 8.4% average rise in the VIX over the next month based on these two historical factors (note that I am not saying this will happen going forward); certainly a nice hedge if you can get it should equity markets sell-off.

So, is the VIX a good S&P 500 hedge? Based on the above table and the previous tables which incorporate starting levels the VIX may be a good hedge when:
  • Markets are calm
  • The price of volatility, as a form of insurance, is cheap
In other words, the VIX may be a nice equity hedge.... when you don't think you need it. By the time you KNOW you need(ed) it, after markets sell-off or when the VIX index is rising, it is likely too expensive to be of value.


  1. Jake: thanks for indulging me. I suspect you now have the raw data for what I think might be the most useful analysis--putting aside the backwardation issues which get in the way of effectively implementing a hedging strategy. Say you already have a long equity position, and you observe a level and a rate of change for VIX (could be a month, could be some oscillator, or whatever). For a given level and rate of change, how effective would some kind of long VIX position be at hedging equities? I think what this would look like in one of your tables is VIX level on one axis, VIX change on the other, and the values in the table would be something like the ratio of change in S&P to change in VIX--capturing the effectiveness of the hedge for each level and rate of change of VIX. The first table of your post almost provides this: I can divide the table entries by the x axis, but I'm afraid the approximation is too rough.

  2. I dig it.

    I think you'd want to scale the VIX to make the hedge equal from a risk parity perspective. any thoughts on how to scale? The VIX has historically been about 6-7x more volatile, but less in normal markets.

    Btw- there is an index that does just this...

    There is also an ETN (not vouching for it), symbol VQT

  3. Curiosity may just kill the catfish. I'll need to get my hands dirty with some data and get back to you. My go-to would normally be daily OLHC data for indices and individual securities (including ETNs/ETFs) from Worden Bros' Telechart software, massaged by spreadsheet. If you have a better data source and an analytical tool which assumes no residual programming skill, I'll be most grateful. Barring that, please be patient while awaiting my contribution. No guarantee any practical scheme will come of this, let alone anything PG-rated for a typical retail investor.