More on the Double 7 Strategy

15Jan13

This is a follow up to the Double 7 Strategy originally described by Trading Markets and recently covered by System Trader Success.

Strategy rules: go long the S&P 500 at today’s close if the S&P 500 will close both at a 7-day closing low and above its 200-day moving average. Hold the position until the S&P 500 closes at a 7-day high. Results from 1993:

20130115.01
[logarithmically-scaled, growth of $1]

Strategy results are in red, and for comparison, days when the strategy is out of the market are in grey.

Entries/exits are based on the S&P 500 price index, but strategy returns are adjusted for dividends. I’ve ignored return on cash and transaction costs/slippage.

Of course, you can’t trade the S&P 500 directly, but I’ve used it here for simplicity’s sake. Similar results would be had with the ETF SPY or other broad market indices.

. . . . .

Both Trading Markets and System Trader Success began their respective tests with the launch of the ETF SPY in 1993, and clearly the Double 7 Strategy would have done an excellent job sidestepping the major bear markets over this particular 20 year period (no stats necessary – the graph speaks for itself).

But taking a longer view, we see that prior to 1993 the strategy was much less impressive. Same strategy, same assumptions, from 1970:

20130115.02
[logarithmically-scaled, growth of $1]

That doesn’t necessarily mean the strategy is a bust today as the market (and the effectiveness of indicators) clearly evolves over time.

. . . . .

I’ll make a long story short (only because this strategy is a non-starter for me).

The strategy has done a good job reducing the impact of long protracted bear markets (ex. 2000-02). That’s partly the result of the 200-day moving average requirement (basic trend-following), but also partly because the strategy is only stepping in after recent weakness.

But it’s also done a poor job capturing enough of the market’s returns, particularly during protracted bull markets, which makes sense given that (again) it’s only stepping in after recent weakness.

I say Double 7 is a non-starter for me only because I think that there are more effective ways to play what I call “intermediate-term mean-reversion”, but as always, this is but one nerd’s opinion and I appreciate the opportunity to put the strategy through the paces.

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.



6 Responses to “More on the Double 7 Strategy”

  1. 1 MarketSci

    Quick comment forgot to mention in post:

    Slightly better results (when looking at the entire 40+ years) using longer lookbacks…say 9 or 10 days. Doesn’t change my own conclusion on the strategy, but worth a look for anyone testing the strategy.

    michael

  2. Michael, could you touch on those [more effective ways to play what I call “intermediate-term mean-reversion”]? Thank you, Jozef

  3. 4 Alex Argyros

    Michael,

    Here’s a system that seems to do rather well from a “looking at the chart” perspective: Go long or short $SPX (SPY) based on crossovers of the 50/200 EMA on the daily chart, as long as they’re confirmed by the Aroon (230): long if above the zero line, short if below. The Aroon indicator seems to filter out a lot of whipsaws.

  4. 5 arco2

    Hi, I agree that there are better ways… furthermore the idea that exits are only at 7 days high might yield some nasty drawdowns. Using historical cost model on equity taking 5bps for last 10yr, 10 for last 20 and 30bps for last 30 (as costs were indeed higher) shows a complete different result with a less nice equity line… 7 days holding vs 200 days trend analysis are way too many trades that dont cover the costs. It does look nice in the last 10 years though…

  5. 6 Rick

    Michael,

    Here is another analysis that includes more symbols. What do you think? I think more or less the author makes the same conclusions as you: http://bit.ly/SNQptH

    In another post he studies three similar strategies: http://bit.ly/W69VAr


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