What if you could have your capital exposed to the market just 16% of the time, make a handful of trades a year, and still beat the market?
I really like breadth indicators. I fool around with them often. Here is a new one (nothing earth-shattering, but I haven’t seen anything written about it by others) that fulfills my desire to keep things simple, reduce drawdowns, and beat the market. It also has a high percentage of winning trades, which may make it easier to trade, in my opinion.
The system uses an indicator which calculates the percentage of all major exchange listed stocks that are trading above their 20 day moving averages (PctAboveMA20). All stocks, excluding OTCBB, trading above $1 are used to calculate the percentage. The calculation includes de-listed, non-surviving stocks, so the percentage 10 years ago should be very close to what it would have been in real-time, 10 years ago.
- Buy SPY or QQQ at the Next Open if the PctAboveMA20 < 20;
- Sell SPY or QQQ at the Next Open if the PctAboveMA20 > 60;
No commission or slippage included. All SPY (starting 2.1.93) or QQQ (starting 3.10.99) history used.
- Compound Annual Growth Rate: 5.44%
- Winning Percentage: 77.42%
- Max System % Drawdown: -32.30%
- Exposure: 16.66%
- Number of Trades: 31
- Sharpe Ratio: 1.86
- Compound Annual Growth Rate: 8.31%
- Winning Percentage: 80.77%
- Max System % Drawdown: -33.11%
- Exposure: 16.01%
- Number of Trades: 26
- Sharpe Ratio: 1.98
Take a Look:
Click on the chart to enlarge…
As you can see, the indicator triggered a long entry last Wednesday morning.
Note that the indicator misses some nice setups, while waiting for the extreme < 20 reading. This could perhaps be improved, possibly by using Bollinger Bands around the PctAboveMA20 and using them to trigger a buy or sell. Bands may allow the system to catch a dip when volatility is low and the the market is in a strong uptrend – see September 2010 through March 2011 in the chart above. Just thinking out loud…
Equity Curve for QQQ
In the next post, I’ll be describing another, more interesting breadth system compared to the one above.
You had me at “beats the market”.
Up 7.7% this year on Qs, and 6.9% on SPY.
love this study.
It seems like the system has some nasty drawdowns – prior to spiking higher. Any way this can be improved? Sitting tight watching your trade go against you (badly) is not fun.
Marshall, it will be difficult to improve the drawdowns without curve fitting it. It is not really supposed to be a hot shot system. Instead, just a way to beat the market with much lower exposure and volatility in the returns.
I think this is a similar idea to some of the “T” indicators in Worden’s TC2000. Specifically, T2108-T2116 plot the % of stocks 1 and 2 channels above/below their 40/200 dma. I tend to watch the % 1 Channel Below 40dma and when this gets above 65 it is close to good entry point. Your version is using 20dma so a different spin, but nonetheless, these “T” indicators give a good sense of where the market is. I don’t think the “T” indicators are available on the Freestockcharts.com site, only TC2000 version.
Really nice… Pretty good long term results. Thanks for posting it.
A couple of things stand out – the Sharpe ratio is really high for such a simple system; there are at most only a couple of trades a year; and the drawdowns, while significant, would discourage widespread adoption.
Any chance of testing this on something like the Japanese market?
Hmmm. You mean use a country-specific ETF, rather than SPY or QQQQ?
Since the PctAboveMA20 is created from stocks trading on our major exchanges, I’m not sure how it would apply to other countries. Interesting idea though.
Do you have any insight about the sensitivity to the microcaps? I.e., if you confined the breadth to, say, the S&P 1500 stocks or the Russell 3000, does it work as well? I ask because with SPY and QQQ you are trading only the top 500 by cap or top 100 ‘techs’.
Looking forward to your next post!
Don’t know about sensitivity to micro-caps. I could add some criteria to ensure that micro caps are excluded but I do not have access to a survivor-free S&P database. I would have to try and recreate it by assigning minimum price and volume thresholds.
Very nice. Thanks.
Interesting study. I love breath indicators. A much misunderstood and under-used group of market metrics.
Good stuff. Thanks for sharing.
Is this a long/no position system? Did you look at going short over 60?
Banx, shorting does not improve it. Yes it is a long/no position system.
Thanks for the interesting posts. In the chart providing signals over the past year, it appears that Long Entry signals are better placed than the Sell signals. Perhaps using the Long signals only, and using the Sell signals to tighten up a trailing stop would be a simple modification that improves results.
Of course if you posted a picture with a longer history of signals we might know quickly if this observation is true over a longer period.
Larry, that is a good idea. I will try it out when I get my computer setup running again.
It’s a little hard to tell from the charts but it looks like most of your drawdown is happening near the beginning of the trade. I think you’d get better total returns if you only buy if PctAboveMA20 >= 20 but was < 20 yesterday. I'd also be inclined to exit positions that cross back below 20. Think of it as a stop. I'd expect the number of trades to go up and the successful trade % to drop, however I also think you'd see a drop in max drawdown and exposure with an increase in CAGR.