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: Ken Adams who wrote (8180)	6/10/2000 12:45:00 PM
From: Dan Duchardt	Read Replies (1) \| Respond to of 39683

Ken, Your reply to Sam: You said you used $19.95 per trade. Do you mean round turn? If that's actually per ticket it's about double what I'd pay. That would make your results look much better, too. I took a look at Sam's files. You can deduce from his data that he used $19.95 per transaction, not the round trip. For example, on ADAP he has 20 turnaround trades at a cost of $798. Cutting that in half would certainly make things better. Cutting it by 90% would be WAY better. Wishful thinking?? Actually it's not. I'm doing it now, though I'm not in a position to really take advantage of the BTT-MAX strategy. That's a limitation of resources, not the nature of the account. I'm not here to hype a broker. I have nothing to gain by it, but if anybody cares to check where I post on SI you can find it, get an honest summary of the performance, and ask questions if you like. Dan

To: Ken Adams who wrote (8180)	6/10/2000 12:47:00 PM
From: smchan	Respond to of 39683

For grins I re-ran the backtests using $10 per trade, or $20 per roundtrip. The only thing I'm not certain of is whether the software considers reversing a position 1 or 2 trades. (I believe some brokers will let you close long and sell short at the same time.) I wonder about backtests too b/c they don't always reflect market reality. My backtests are simple; they do not account for spreads, etc. My rules are as follows: "establish initial position" if long signal, buy 200 at closing price if short signal, sell short 200 at closing price "reverse position on signals" if long signal, cover short and buy 200 long at closing price if short signal, sell position and sell short 200 at closing price If you like, I could build a huge database and run the system on EVERYTHING week to week. :-) Sam

To: Ken Adams who wrote (8180)	6/10/2000 12:57:00 PM
From: Ken Adams	Read Replies (4) \| Respond to of 39683

All... Here are some initial thoughts about backtesting, and why I am somewhat skeptical of the results. I'm going to refer to a "timely" chart on BVSN for some examples. It might be helpful to print the chart and then follow my comments. 207.61.23.99 Begin by looking at the signal to short on Wednesday, May 31. It's my understanding that these programmed signals result from the close of the signaling bar. On May 31 the bar closes about 35.875. This would result in a losing trade. If, however, the signal could come when the actual MAX occurs (something like 37.375) the trade would have given a small profit. Just one more. The next signal comes the following day (June 1) to go long. However, the signaling bar closed about 39.5, about a point and a half above where the actual MAX happened. Granted this resulted in a very nice trade, but it could have been even better. I see examples of this on every chart. Now, I'm not naive enough to believe we could actually trade like this. But, couldn't we trade somewhat better? My point is that the backtesting software doesn't take into account the way I might FEEL about the trade. I believe this skews the results away from the better results that are attainable. I am perfectly happy with the MAX idea for an aid to better trading. I'm just not sure of the value of backtesting. For my own "backtesting", I like to look at what might have been if I'd acted on the last 2 or 3 signals the chart shows. I don't ever go much farther back than that. Bottom line! I think the idea of scanning for these volatile issues is very worthwhile. After that, I think it's best to eyeball each chart. Throw out the least volatile immediately. Anything that had nice moves last week will probably do so again this week. At least that's an assumption I make. With a big enough stable, you can throw away anything that stops producing for you. A new scan next week will turn up a whole bunch of new candidates. As can be seen on the results of yesterday's scan, many old favorites keep showing up. Comments certainly welcome here! Ken