Studies - Posted on Wednesday, March 25, 2009, 10:43 AM GMT +1


Mar Wednesday 25

Weekday Seasonality of the VIX

My name is Frank, and due to the fact that this is my 3rd post after I’ve started the blog on Tuesday, March 24, 2009, a few introductory notes.

I have traded as an individual investor for more than 20 years now. I take a statistical approach in combination with historical market data to profit in the US equity and future markets, focused on intraday and swing trading opportunities (regularly using E-mini futures for intraday trading, and -leveraged- US equity index ETFs for swing trading) .

Please accept my apologies for not being a native speaker, and please take respect to the fact that that blog is (and will probably be for the time being) under construction (there is always room for improvement, especially right after the start, and especially focused on getting the tables more self-explanatory; your suggestions -e.g. for different figures and/ore setups- are always more than welcome).

My email:



The VIX® (CBOE Volatility Index) is an index that infers 30-day expected (implied) market (S&P 500) volatility from S&P 500 index options (usually in the first and second month, and averaging the weighted prices of puts and calls over a range of strike prices).

For example: the VIX closed at 42.93 on March 24, 2009. This represents an expected annualized (!) volatility of 42.93% over the next 30 days; thus one can infer that market participants expect -with the assumption of a 68% likelihood (one standard deviation)- that the magnitude of the S&P 500’s 30-day return will not exceed 42.93%/SQRT(12 month) = 12.4% over the next 30-days.

Regularly there is an inverse correlation between the VIX and the S&P 500 due to the fact that market participants expect market volatility to be higher on downward movements in the market. But there are periods when this correlation only applies to a much lesser extend or none at all (such a non-confirmation -an ‘uncorrelated behavior’ of the VIX versus the S&P 500-) sometimes gives an early clue and warning that the then current movement in the S&P 500 may not be sustainable.

But a special kind of such an ‘uncorrelated behavior’ of the VIX regularly happens in the Friday to Monday (weekend) time frame, and regularly intraday as well (mean-reversion tendency). The table below shows the VIX′ historical probabilities and odds for a higher and lower open, a higher high and lower low (than the last sessions high/low) and a higher and lower close with respect to the days of the week (since 01/02/1990, VIX on dates before September 22, 2003 compiled by the CBOE to the new methodology).



(click on image to enlarge)

It is especially notable that

  • since 01/02/1990 (long term only, for other time frames the results may significantly deviate), on average (NOT to be mistaken for ‘always’) and at-any-time (and applicable to all days of the week as well) the VIX closes below its opening quotation in the event it has opened higher, and above its opening quotation in the event it has opened lower (intraday mean-reversion tendency of the VIX). So market participants always seem to be overly optimistic or pessimistic concerning (overestimating ) the real-time impact of the S&P 500′ opening direction and magnitude of change on the expected 30-day market volatility (VIX), regularly calming down as the session develops (almost independently from the then developing intraday trend concerning the S&P 500),
  • the intraday mean-reversion tendency of the VIX mentioned above is also evident in the respective profit factor. If one would have implemented (theoretically, and for statistical purposes only) a simple mechanical trading system like ‘buy the VIX on open in the event the VIX opens higher; sell the VIX on open in the event the VIX opens lower; close the trade on market close‘, the respective profit factor (sum of all profits divided by the sum of all losses, calculating daily profits and looses by simply subtracting the VIX’ close from the open) would have always (at-any-time, for every day of the week and for both long and short trades as well) been below 1 for a negative expectancy,
  • although the at-any-time probabilities for a higher and lower open of the VIX are about even (applicable for Tuesdays, Wednesdays and Thursdays as well), there is a huge historical variance concerning Mondays and Fridays. On Mondays the probability that the VIX will open higher and/or close higher more than triples the respective probability for a lower open and/or a lower close, while on Fridays -but to a lesser extend- the reverse rule applies. It is almost twice as probable that the VIX will open higher on a Monday than on a Friday, and almost thrice as probable that the VIX will open lower on a Friday than on a Monday.

As Adam on the Daily Options Report already pointed out: “The *cash* VIX loses some steam ahead of holiday’s/weekends as traders lower bids to account for their weekend/holiday decay.”

But since 01/03/2007, the intraday mean-reversion tendency of the VIX has been almost completely reversed to a trend-confirming tendency, means the opening direction and magnitude of change of the VIX is regularly at least confirmed if not intensified until the close of the session (additionally evident by looking at the respective profit factor which now makes a -theoretical and for statistical purposes only designed- system ‘buy the VIX on open in the event the VIX opens higher; sell the VIX on open in the event the VIX opens lower; close the trade on market close‘ profitable.



(click on image to enlarge)

Best regards

VIX® is a registered trademark of the Chicago Board Options Exchange

Comments (4)


  1. Bill Luby says:

    Hi Frank,

    I have enjoyed your comments over the past year or two at VIX and More and am glad to see you striking out on your own.

    Congratulations on the new blog!

    Best wishes,


  2. Anonymous says:


    Looks like a very interesting blog and am excited to start reading your posts. GL. One small note about your charts, it might make it easier on the eye to discern up and down periods by varying the background color btw two different colors.

  3. […] P.S. Frank at Trading the Odds also did a nice write up on VIX DOW seasonality here. […]

Leave a Reply

Your email address will not be published. Required fields are marked *