Back to Community
Poor man's VIX/VVIX

Now that you can't use fech_csv anymore, I'd like to develop a simple approximation
to cboe.com's VIX/VVIX indices, using only the Q available data (Morningstar?).

This may take me a while, but I'll publish code here when it's done.
Any contributions welcome.
alan

3 responses

I am curious, how?

WelI, I don't know how...maybe not possible...this is a research topic for me...sorry for the confusion...

I've started looking at the papers pointed to at:
http://quant.stackexchange.com/questions/1951/how-to-price-a-volatility-index-option

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1524193
http://www.iaeng.org/publication/WCE2011/WCE2011_pp433-438.pdf

I've included my initial version of a "Poor Man's VIX Peak Indicator".
It ends up using data from the ETN VXX security which I believe obeys the current rules of the Quantopian contest.

For some background, when the Aug 26th mini-crash happened, my contest entries lost a lot, and it made me want to create
a "mitigation" indicator that told me when to liquidate all my positions, as the market is just too risky to trade in.
This would be using the same reasoning used when betting horse races at the track or climbing mountains...just don't do it if seems too risky.
This type of reasoning results in not betting most races at the track, and in not climbing most days due to weather or physical conditions of the climber.

So, I created a very simple indicator using VIX, whose data I downloaded using "fetch_csv", whose purpose is to tell me when to have a "fire sale" and liquidate all positions, and stay out of the market until the indicator turns the other way.

Then the contest rules changed so that you couldn't use "fetch_csv" anymore...but I liked my indicator, so wanted to find a contest-legal way to get an equivalent indicator. This was the goal of the included backtest, and I've made a stab at that, with some validation.

In the process, I've learned a lot...as Simon pointed out...getting a VIX equivalent isn't that easy, so I reverted to try and use the data from the VXX to get at the local peaks.

Comments welcome!
alan

Clone Algorithm
32
Loading...
Backtest from to with initial capital
Total Returns
--
Alpha
--
Beta
--
Sharpe
--
Sortino
--
Max Drawdown
--
Benchmark Returns
--
Volatility
--
Returns 1 Month 3 Month 6 Month 12 Month
Alpha 1 Month 3 Month 6 Month 12 Month
Beta 1 Month 3 Month 6 Month 12 Month
Sharpe 1 Month 3 Month 6 Month 12 Month
Sortino 1 Month 3 Month 6 Month 12 Month
Volatility 1 Month 3 Month 6 Month 12 Month
Max Drawdown 1 Month 3 Month 6 Month 12 Month
# Backtest ID: 5626c154d4b69a110f13b50f
There was a runtime error.