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 : Market Gems:Stocks w/Strong Earnings and High Tech. Rank -- Ignore unavailable to you. Want to Upgrade?

To: Peter O. who wrote (10915)	7/1/1998 9:54:00 PM
From: Jenna	Respond to of 120523

Peter..I heard this while I was exercising this afternoon and I was extremely interested. TYC just broke out nicely and WCOM I already own. CCU is a company I've traded during earnings and I'm not surprised it is on the list. ODP is another one that has been on my watch lists and I like a lot. STB I follow and have had on my list as well. CMVT has been on the earnings play list for about 2 quarters now. The others I'd like to check further since I don't follow them yet. I would put them on Last's neural net and also in a "permanent" watch list for general Technical Analysis. I'd also check the earnings dates on all of them and follow their movement particularly at that time.

To: Peter O. who wrote (10915)	7/2/1998 12:26:00 AM
From: LastShadow	Respond to of 120523

The List: I think I need to clarify something before I respond to the request ot check out some stocks with the neural net. The process involved is not simply one in which I enter the symbol and some kind of magic happens. The normal process is to select what one thinks are leading indicators for a stock, and attempt to predict the price a few days out. The net 'learns' the behavior of the stock price against those chosen indicators until it establishes some confidence in its prediction. Then you look at what weight each of the indicators you selected carried in the mathematics of the learning process. Now you rerun the model against the best forward indicators until you are confident that you are within your tolerance for error. Now you introduce data the net has not seen - say the next three months, and have the net walk forward against that data and attempt to predict the future price (or change or % movement, or whatever you want it to forecast). You do this for several periods at some period of time distant from the preceding data. Once you are comfortable that the current prediction is close to reality, you decide on what your trading strategy will be - do I buy on some combination of crossovers or chart pattern formations or if the predicted price is greater or less than the actual, or what. This basically takes some experimentation and reliance on whatever one already knows about the stock's behavior from fundamental and/or technical analysis. Then you paper trade for awhile to determine what the usual slippage would be on entry and exit. Having said all that, I assume you gather that I didn't do all this for the list you posted. What I did do is run the stocks through four nets that are fairly generic for all stocks in terms of profitability and error reduction. These don't give optimal returns, but it does give me a feel for which direction to take further analysis in. Since I use a long or short trading swing (essentially never out of the market unless you are really unsure of what is going to happen or there are no good picks surfacing), I look for anything that generates above 400% return on those nets to pursue. The reason for that cutoff is that if I am not shorting, then the return would be half of that number, and I would rather just pursue the ones for which I can use my existing TA system to confirm without building custom indicators in there for one lousy stock. So now that I have probably bored you to tears, here is what I discovered about the stocks you listed - the average return from the generic nets was about 200%, with a high of 375% for CMVT, and a low of 110% for ILN. Interestingly, ALL of them were indicated long in every net, specifically since the dates listed: AMAT 6/30 CCU 6/23 CMVT 6/29 DHR 6/29 ILN 6/26 ODP 6/30 PGR 6/30 STB 6/23 TYC 6/25 WCOM 6/29 I would caution that most are not stocks you would have gone long on a month ago and been profitable holding, as they swing a lot. The ones that have risen the least and generated a buy signal the most recently are probably the better candidates to pursue. Lastshadow