Using strength of recent closes to time SPY

I recently read a post by QUSMA (great blog, btw) that inspired some research. The author’s post was on the predictive power of a stock’s closing price in relation to its daily range. The author defines a measure, CRT DR, that is calculated as (close – low)/(high-low).  It turns out to be a fairly useful mean reversion indicator for stock indexes such as the S&P 500 and Nasdaq.

I’ve been playing around a bit with this measure. One thing I’ve found useful is to create a moving average of daily CRT DR values. Below is a chart comparing 2 strategies for SPY.

Strategy 1: Hold SPY when the 4 day simple moving average of CRT DR is below 50%, else hold cash.
Strategy 2: Hold SPY when the 4 day simple moving average of CRT DR is above 50%, else hold cash.

As you can see, the first strategy captured most of the gains while the second captured a majority of the losses.  Holding SPY when the measure was less than 50% performed remarkably well during both bull and bear periods.  There’s decent drawdown in there but the strategy tended to bounce back pretty quickly.

I’m sure there are ways to improve on this and I plan to continue to search for smart ways to incorporate it in my trading.

Posted on November 26, 2012, in Other. Bookmark the permalink. 11 Comments.

  1. Very interesting idea, Mike. Are you planning on adding it to your list of strategies? And, what is the CAGR and max DD during the test period?

    • CAGR is 12% with max DD of 21% and average DD of 2.1%. SPY had a CAGR of 6% with max DD of 56% and average DD of 15.8% during test period. Stats are for trading cash index only. I don’t take dividends into account here for the strategy or for buy and hold. Results also don’t take into account commissions, slippage, etc.

      Not planning to add this to the list of tracked strategies yet. Still exploring ways to use it.

  2. Hi Mike! Really great idea the moving average! I think this is completely innovative.. I passed some numbers on excel this afternoon working on it. Just to improve the odds, consider more selective values CRT DR, like <30 for longs and above 70 for shorts. The results are great.. I tried to reach a trade-off between the number of occurences and profit factor. I made different backtests for:
    today's night performance, tomorrow's opentoclose performance and from today's close to tomorrow's close. The results are really impressive. I will use it like a filter for my nightly leverage trading.

  3. Interesting stuff Mike. A simple long-only strategy with real promise. Love it!

    Brian Johnson of recently posted on a very similar (albeit slightly more complex) system ( Worth a gander if you haven’t already seen it….


  4. I was able to reproduce the results. Total return ~800%, CAGR ~12%, Max Drawdown ~21% etc ONLY when using a 4 DayMA of (C-L) / (H-L).

    If the moving average was 5 weeks – big looses.
    If the moving average was 3 weeks – no advantage.

    Seem way to specific – form fitting !!!
    Why is the 4 DayMA magic ?

    • 5 weeks is a much different time frame than 4 days. I would check out 1-7 day moving averages to get an idea for how sensitive the results are to that parameter. 4 days is somewhat optimized but 3 or 5 day averages also work pretty good.

  5. Don’t know about that. I would suggest you do the 1-7 day moving averages and post the results as an overlay chart on your web page.

  6. Mike in the equity curve above, what is the assumed investment size per trade? $100?

    • There’s no fixed trade size in the chart above. It shows how $100 would grow over time if every trade profit was used towards next trade. The scale level (i.e. starting at 100) is unimportant.

      You will get the same curve shape if you assume a fixed investment per trade and remove the logarithmic scaling.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your 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


Get every new post delivered to your Inbox.

Join 246 other followers

%d bloggers like this: