Much of my work, however, has been inspired by, reinforced by, expanded upon, or confirmed by, numerous other bloggers and traders who also understand the usefulness of this simple indicator.
For more ideas and backtests on RSI(2), you can read my posts, Woodshedder's, Fast Eddie's, Marlyn's, and IBD Index, among many others.
While previous posts have focused on using RSI(2) to trade stock indexes or individual stocks, tonight I focus on using RSI(2) for timing trades into sector-specific ETFs. Specifically, I used the following:
XLY - Consumer Discretionary
XLP - Consumer Staples
XLE - Energy
XLF - Financials
XLV - Health Care
XLI - Industrials
XLB - Materials
XLK - Technology
XLU - Utilities
XHB - Housing
Obviously, this is not a comprehensive list, nor does it include leveraged ETFs, but these sector ETFs have multi-year histories that allow for more reliable backtesting, so I think the results from this test may be extrapolated to other sector-specific ETFs with much shorter trading histories.
The rules of the backtest were simple. Buy when the RSI(2) was less than 10, sell when RSI(2) was greater than 80, or sell when a 10 percent stop loss was hit. Maximum open positions were limited to five.
And here are the results for 2008 year to date:
The results are pretty good and the 74 percent win rate should be psychologically comforting for traders who need more winners than losers.
On the downside, there have only been 38 trades year to date, which means a lot of cash is sitting around doing nothing a great deal of the time. To address that problem, a trader could add other ETFs to the watch list to increase the number of potential trading opportunities.
As for immediate trading opportunities, three of the above ETFs have a RSI(2) below 10, they are XLF (financials) at 9.47; XHB (homebuilders) at 6.71; and XLV (health care) at 5.34. And here are their charts:
At this point, I do not have any positions in these ETFs, although I'm seriously considering a position in one or all of the above. I'll let you know what I plan do before I pull the trigger Monday morning.
3 comments:
good stuff dog
Interesting - I just tried to replicate your results in Amibroker using TC2007 data and Yahoo data and couldn't get to the winning percentage. Here's my question for you - what are the buy/sell rules in terms of open/close. Do you buy the open or close, and the same for the sell?
I also notice that a stop loss was triggered 5 times - what were the rules for the stop loss?
Also is ROI in this case the same as % return? I'm guessing the answer is no because I don't see how you can get to 44% return from 26 trades with an average return of 1.14%.
Thanks!
Buy at market the next day at the open. Sell at market the next day at the open. This system uses end of day prices, so any action takes place at the open the next day.
In other words, if the entry trigger is hit Monday, buy Tuesday morning. If the exit trigger is hit Wednesday, sell Thursday morning. However, if you know an exit trigger is getting close, you could exit the trade during the day if you wanted to watch and wait.
Stop loss is 10 percent.
The ROI number is annualized, so no, the system has not produced a 44 percent return in 26 trades. Still exploring the difference between StockFetcher's ROI and annual percentage return when looking at a full-year backtest.
Post a Comment