Testing TM, Rule #5: the VIX 5% Rule

08Jan09

This is the fourth part of this week’s series on TradingMarkets.com’s 10 Trading Rules. In this post, I’ll look at the fifth in TM’s list: the VIX 5% rule.

TM advocates not being long the market when the VIX is more than 5% below its 10-day moving average because this has historically indicated that the market was overbought. Per TM: “if you only follow one market sentiment indicator, it should be the 5% rule”. Ahem.

2009010801
[logarithmically-scaled]

The graph above shows a strategy that is long the S&P 500 from today’s close when the VIX closes above TM’s threshold (red), versus a buy and hold approach (blue), and for comparison’s sake, another that is long when the VIX closes below TM’s threshold (green), from early 1990.

Recall that TM’s threshold is 5% below the 10-day simple moving average of the VIX. This is a proof of concept so these results are frictionless (ignore transaction costs and slippage) and do NOT account for return on cash.

For the number lovers:

2009010802

TM’s 5% rule cuts out about 25% of the time in the market.

During that 25% of days, the market has generally traded neutral to bearish. This wasn’t true in the late 1990’s, but intermediate-term overbought indicators (like this one) in general didn’t work well in those go-go days (a product of euphoria).

I think it’s important to note that trading above the 5% threshold hasn’t necessarily been bullish for the market (just look to the bear market of the early 2000’s to confirm that), but that isn’t the way TM postured the rule. It was postured as a defensive indicator, not an offensive one.

So do I agree that “if you only follow one sentiment indicator, it should be the 5% rule”? No, on two accounts: (a) no one should be following just one market sentiment indicator, one strategy, one anything – that’s just foolhardy, and (b) I think that there are more effective indicators that play in this intermediate timeframe.

Having said that, as a defensive indicator, I think TM’s 5% rule has wings.

[Edit: click for a summary of all related posts in this TradingMarkets series]

Happy Trading,
ms

 

To stay up to date with what’s happening at the MarketSci Blog, we recommend subscribing to our RSS Feed or Email Feed.



One Response to “Testing TM, Rule #5: the VIX 5% Rule”

  1. 1 david

    my issue for this rule stems from the fixed parameter of 5% which is ironic considering that the VIX tracks volatility and hence will vary considerably from quiet to noisy conditions. The bollinger band strategy you mentioned previously on this blog is likely to be more robust because it scales for volatility on the VIX itself. The equity curve looks better. Note that CXO has an article about this TM rule which looks at nonoverlapping periods and concludes it is of limited value. I disagree, as it clearly appears than almost all countertrend strategies are working now on the S&P500 and they are all correlated to some degree. Whether or not there is predictive value is debatable, but like moving averages-which are not supposed to predict– it keeps you out of the roughest periods.
    A second option could be to test changes in ATR (average true range) which has a low correlation to changes in standard deviation (TASC article by Gordon Gustafson). ie buy the vix if it is “n” atr units below the 10-day moving average etc. The same TM rule with ATR should have a much better equity curve.
    Perhaps a measure of ATR and standard deviation simultaneously, or using two systems might work best.
    dv


Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s