Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Strategies & Market Trends : Befriend the Trend Trading -- Ignore unavailable to you. Want to Upgrade?

To: Triffin who wrote (8068)	6/6/2000 9:52:00 AM
From: Ken Adams	Read Replies (3) \| Respond to of 39683

Jim, I completely agree. What we've evolved into here is a method that works if fed the right sort of stocks. That should be the challenge... finding a constant supply of those stocks. When the market changes its personality to the point that what we're doing no longer works, that will be the time to restart the "Holy Grail" hunt. Ken

To: Triffin who wrote (8068)	6/6/2000 10:59:00 AM
From: Dave Bissett	Respond to of 39683

It's a good .02 worth too. I didn't mean to start in another direction. Regarding your statement <<We should automate screening for suitable candidates for our 'stable' and develop a profit taking rule to maximize the P&L from open trades ..>> Do you know any way to automate this hunt? I don't know of a technology to do that. Dave

To: Triffin who wrote (8068)	6/6/2000 11:48:00 AM
From: Dave Bissett	Read Replies (3) \| Respond to of 39683

After rereading your post to me and Jesso a few times I'd like to take a stab at finding some common ground on this system development/stock selection issue. I've done tons of backtesting and am well aware of curve fitting issues. Results are completely dependent on the particular stock and time frame tested. OTOH, there's nothing particularly magical about any specific MA crossover...a few stocks "fit" that and many don't. I don't know how you search and identify stocks that fit a particular rule (such as a 5-8 crossover) unless you run backtests (at least with eyeballs) on a basket and select those which show good backtested returns with those parameters. As others have pointed out, there WILL come a time when those stocks will no longer "fit" the system and results will become negative, and the rule will have to change or different stocks will have to be found. So the curve fitting kinda actually works both directions. But I have a suggestion that may provide a workable compromise. Let me refer to the PMCS chart that CL posted.... What's most important about the chart IMHO is it's persistent and continuous volatility, NOT the fact that it accidentally "fit" in with the 5-8 crossover rule. Wouldn't it be useful to know if the profits would have been, say, 40% greater using a a 3-9 crossover, or maybe even a different trading rule entirely? If that was shown to be the case, what rule would you choose to trade the stock tomorrow, next week? Of course it's trading pattern could change, but even now no one knows when any of these stocks are going to fall out of the 5-8 pattern. So until whatever pattern the stock's in changes, why wouldn't we want to be using the most profitable rule we can as a starting point? Here's the potential common ground...maybe we should scout for stocks that have shown persistent volatility over a period of time, as long as possible actually, and backtest several "trading rules" and at least experiment with the one that appears to fit the stock's volatility the best. If the goal is only to trade a very small number of stocks, which is all I can certainly handle, this shouldn't be too difficult to do. And posters here certainly have the technology to do this analysis. I'm sure interested in your and other folks comments as well as participation if you feel like it, and don't mean to upset the established order. If this doesn't interest anyone I'll not bring it up again. Dave