Test of FundzTrader Sector Rotation Strategy
This is a test of a simple sector rotation strategy from FundzTrader (now defunct) that takes advantage of the no-load/fees offered on Fidelity Select sector funds held longer than 30 calendar days.
[logarithmically-scaled, growth of $10,000]
The graph above shows strategy results (red) versus the S&P 500 (grey) from 1987.
Strategy Rules: (1) Invest in the Fidelity sector fund with the highest % gain over the last 25 trading days. (2) Hold that fund for at least 30 calendar days. (3) After 30 days, if that fund is still the best performing over the last 25 trading days, continue to hold it; otherwise exchange it for the new best performer. Rinse and repeat.
I’m considering the entire universe of 40 Fidelity Select Portfolios. These funds do not carry a load, or incur a per-transaction or early redemption fee if held longer than 30 calendar days.
Also note that I’m assuming a one-day lag in execution because mutual funds are only priced once-per-day. For example, after Monday’s close I calculate price changes over the last 25 trading days. On Tuesday I place my order. My order is executed at Tuesday’s close (a one-day lag in execution).
Numbers for the number-lovers…
Thoughts
I’ve wanted to add a sector rotation strategy as a component in my own trading for a while now. At the moment I’m heavily weighted towards extremely short-term trading, and since retiring the original MarketSci strategies I no longer trade anything that plays in this middle space between aggressive trading and lazy trend-follower.
This is a very simple example of such an approach.
Two big issues with this particular strategy:
1. Because the strategy is picking the top % gainer (rather than using volatility-adjusted return or the like) it will tend to pick higher volatility sectors, meaning some of the return is simply the result of more volatility/beta and not good picks.
2. The strategy is always long the market (note atrocious drawdowns) and has no mechanism to step out of the market in bear trends.
In a follow up post I’ll show a couple of tweaks to fix the two issues above. As always, more to follow.
[Edit: click for a summary of posts related to this simple sector rotation strategy]
Happy Trading,
ms
Geek note: this strategy is based on counting calendar days so it will produce slightly different results depending on when the strategy begins, but because it can hold the top performer longer than 30 days, it doesn’t take long for those differences to “synchronize”. In other words, I don’t think that issue is an issue here.
. . . . .
To stay up to date with what’s happening at the MarketSci Blog, we recommend subscribing to our RSS Feed or Email Feed.
Filed under: Stock Market Sectors, Trading Strategies | 7 Comments
Not related to the subject of your post, but curiosity got the best of me.
You have mentioned several times that you have retired your old strategies. Which ones are those ? Why you retired them ?
I will understand if you cannot/will not discuss this issue.
Thanks
eber
RE to eber: hello sir. I can always discuss – we are an open book.
The original MarketSci strategies performed just fine – I preemptively retired them because I didn’t feel they were as robust as our newer strategies. You can read more here:
http://www.marketsci.com/strategy.MS.html
We retired the RH strategy for different reasons. The strategy suffered a sizeable drawdown followed by a long period of poor returns. RH was a product of our Timer Seeds program, so much of it was a black box to me, so I couldn’t get a good read on the strategy’s future potential. Rather than expose folks to a question mark, I dropped coverage on the strategy. Read more:
http://www.marketsci.com/strategy.RH.html
P.S. you can find summary stats and audited results for all of our retired programs at http://www.MarketSci.com and we continue to carry them in our combined performance graph.
Hope that helps.
michael
1) Being long “beta” is fine so long as you have a mechanism for recognizing when to get out. By increasing or decreasing exposure to beta (which is a big part of what this strategy is doing), you can generate alpha.
2) Agree on this point. At the very least you should add a cash option (like SHY ETF). Even if you aren’t actively trading it, you can still go to cash if your cash proxy is the best performer.
RE to BB:
I agree that long beta is not a bad thing if the strategy steps down exposure during bear markets. The point of #1 is that readers shouldn’t be fooled by the higher return into thinking that’s necessarily superior results – it’s all relative to beta/risk taken.
RE #2: good idea (moving to cash if it’s the best performer), but there have been very, very few times when the top performing sector hadn’t beaten cash over the previous 25 trading days. I think a better approach is going to be to go with the trend of the broader trend (which will be the subject of the next post).
michael
I don’t disagree, but I am not sure about the cash idea. In theory it is sensible, but Michael Carr performed some tests in his book:
Amazon.com: Smarter Investing in Any Economy: The Definitive Guide to Relative Strength Investing (9781934354056): Michael J. Carr: Books
He tested very thoroughly (I suggest to read the book) the RS ETF rotation (as you start to do it now), but he used weekly review, not monthly.
His test on Fidelity funds usually returns about 20% CAGR, but
he found this:
”
-portfolio selection matters: using 20 Rydex funds (instead of Fidelity) with the same method only returns: 7% per year. His reasoning: it contains a money market fund (cash?): this doesn’t allow the ‘winners to run’. The less volatile money market fund replaces the better gainers.
”
Personally, I am reluctant to accept his reasoning, but because
The book is the summary of 3 years research, I guess he really tried to investigate why the Rydex funds didn’t work. So, I think every programming bug was corrected.
My message here is then than please try different sets of ETFs, not only the Fidelity.
Hi Michael,
Not sure if you saw this, but Woodshedder over at iBankCoin wrote about this strategy a couple of months back.
The link is here:
http://ibankcoin.com/woodshedderblog/2010/01/20/fidelity-select-sector-rotation-strategy-wrap-up/
RE to Jeff: hehehe…great minds think alike. I was going a similar direction w/ my test as he ended up going. I’ll finish out my series and try not to peek at what he did – should be interesting to see if we come out in the same place. michael