Saturday, November 29, 2008

RSI2 Above 99 "System"

Before I get into the RSI2 thingey, I wanted to mention that Pristine is having a free three day look at their Gold Suite, Dec. 2 to Dec. 4, next week. This includes a look at all their services, I am not affilliated with them, and I only "do" the free stuff (magic words), although I have been a member in the past, also, I have never taken any of their courses. Who knows, you might make some change. I know why they are doing this, they really had a soap opera going there, as the former owner came in and took over the company, in a power play, and Greg and some others left, blah blah blah, it lasted about two to three weeks, before Greg regained control, I can imagine they probably pissed some member's off, and are trying to suck them back in, eeeeerrrrr, regain their confidence. If you aren't a member of their free stuff, you can sign up at .

I looked at that 60min chart below on wednesday, and I noticed the RSI with a 2 setting was at 99.6, and I thought, wow, I bet that dosn't happen very often. Well, surprise surprise, it's happened 11 times this year, and 157 times since 1998. So, being the gulible dumb ass I yam, I thought, well I bet that if you short this thing every time it gets over 99, and buy it back when it hit's 90 or below, you'd probably get a 100% win rate! WRONG! Actually, you get a 60% win rate, 94 winners and 62 losers. Don't get me wrong, you make money on it, if you shorted 1,000 shares of the SPY since 1998 you are up about 11 grand, but the thing has been performing poorly since about October of 04', although it's picked it up lately.
Some other stats are that it's maximum consecutive wins were 12, from 8/00' to 3/01', and the max consecutive loser's were 7, from 10/05' to 1/06'. Now, those time periods are not surprising, as the winners came during a bear market, and the loser's during a short bull market. Most of the signals came during bear markets, because you get these short, violent bear market rallies, and then they don't last long, and eventually dump. And of course, the draw downs come in bull markets, as the stupid market can stay over bought a lot longer than you can stay solvent. Another interesting stat, is the draw downs, as the maximum draw down waaaaaaay exceeds the maximum profit, 4100 to 1270, that of course being because the damn thing can stay over bought for long periods of time.
When I checked it on a daily time frame, I was a little surprised to find this happened 15 times, and your win rate went up, 12 winners and only three losers, with about the same amount of profit, and draw downs, accept your maximum run up increased to 3200 rather than 1270.
When I increased the buy to cover parameter to 95, I was REALLY surprised to find you have the same number of winners and losers, 94 to 62, I thought it would increase the number of winners, but what it did, is increase the total profits a couple of thousand bucks, the max draw down stayed the same, but it increased the max run up to 1870 rather than 1270.

Anyway, since I have no life, this is what I do on holiday saturdays. I would imagine that you could really improve the results if you would set a STOP into the testing parameters, which I can't do, like 20 cents above the highest candle when the above 99 trigger to short it hits, but I can't do that yet, but I'm learning. I mainly do this kind of stuff, because it's interesting to see the stat's, but also it's interesting to try and grasp the "CONCEPT" of when and why it works or fails, mostly of how it fails, so when I sit here looking at that RSI above 99, I have to know that the market is going "UP", and just because it's over bought, dosn't mean I can just blindly short the stupid thing, and expect to make money.


blogger templates | Make Money Online