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 : The 56 Point TA; Charts With an Attitude -- Ignore unavailable to you. Want to Upgrade?

To: Doug R who wrote (27434)	3/17/1999 2:54:00 AM
From: Craig DeHaan	Read Replies (2) \| Respond to of 79187

Doug, They're a dynamic polar pair. The SMI-plex is bdog's original setup concoction with your corrected 4.2.3 component (I read past the debut typo; not that it made a hoot of difference in the composite indicator) and a 13ema trigger. After this "core" setup I perused lotsa charts to realize that when things set up on that trigger they usually looked close to the way I like things to appear for a long trade. What a novel concept; filter the db on just those relative positions with a trigger crossover. Backtested the 34TSF on it first and optimized across many to arrive at the conclusion that the 13ema was better for entry, but 34tsf was safer for exit trades and resulted in an average 4-6 day position trades which has been my preference for the past couple years. Then I started testing derivatives using the original component parts: SMI13.25.2, Stoch(13.8.8), SRSI(14) pivots, SMI4.2.3, 5.3.5 crossing over and under 40/-40 arbitrary thresholds; some tests using 5.3.5 and 4.2.3 w/ 2.1.2 or 3.2.1 triggers; some using 4.2.3 with 9TSF triggers (unoptimized) after backtesting across an even wider base. Finally to balance these oscillator systems with trending factors I worked in the MACD13.34.89-0 crosses and the SRSI21 derivative Grace signal from gonz G. In the end I have a series of about 14 system tests to run on initial M-plex trigger hits - usually about 50-60 each day. Based on backtest % ratios and w/l #'s I look for any major improvements or deviations from the b/h norm and especially any multiple hit extremes on concurrent systems. The key is to look for unoptimized consistencies (or excesses) within any of the subsets; the theory being that the more systems that key off at once with better btest results, the better the chances of repeating results, the power of numbers and all that (I used to use OmniTrader 3.5). Hey, there are always some duds for sure, but the risk/reward factors after backtesting each hit helps me get a feel for past system performances in order to adjust responses for the trades. I haven't always been waiting for the exit 'system' signals on these average 5 day setups tho, usually ditch in 2 days if the % is there, so I assume I'm only bettering the system test results if the trades match the norm for that time period and no fixed system can take all the play-by-play action into account anyway. Since mid Feb it has been acting well. Exemplary result hits since the debut have been BLSI, ASPX, SIGM, FAXX, NEON, VVTV, SEEK, & last night LVCI, IFLY. Don't know if its just an alignment with a temporary idiosyncrasy of the market or a breakthru. But, I'd be glad to send you the battery sequence (I think bdog has about 4 or 5 of the MSWIN tests already). Would appreciate your input for the inevitable continued tweaking. By no means is it 100% mechanical, but it appeals to my prevailing doubtful statistical nature and a laziness to look at 100's of charts each night. CPtestr