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 : Neural Nets - A tool for the 90's -- Ignore unavailable to you. Want to Upgrade?

To: F Robert Simms who wrote (703)	12/11/1999 2:02:00 PM
From: Larry Livingston	Read Replies (1) \| Respond to of 871

All, Here is an excerpt from an interesting article in Barrons about a professional research consultant who has developed a neural net screening tool that looks at price/volume relationships. "Enter Mark Scott, founder of the Volume Investor, an Atlanta-based brokerage and research boutique, who has harnessed the power of neural networks to a computerized stock database. The result is a powerful software program that screens for statistically attractive issues under incipient accumulation. The purpose, says Scott, is to home in on stocks that are just about to have a "growth spurt." Ten names that presumably fit that bill popped up in the accompanying table, which Scott prepared for Barron's last week. The Volume Investor's proprietary system works best at identifying trading opportunities with a three-to-four-month horizon. Over time, however, more than a few picks have turned into "buy-and-hold" investments. First things first. Just what are neural networks? Simply put, they're computer programs that can be trained through trial and error to recognize patterns in a given collection of data. In this case, Scott and some colleagues developed and trained a neural net that discovered several price/volume relationships that tend to predict stock-price performance. Out of these findings grew the Stealth Accumulation Monitor, or SAM, a program that algebraically measures the spread between the slope of multiple money-flow indicators and a stock's price. The higher the SAM score, the less the correlation between a stock's price and the amount of money flowing into its shares. In other words, SAM at its best pinpoints stocks that are under accumulation but have not yet reacted to such buying pressure." "To begin with, Scott screens for nine fundamental and technical characteristics, each of which the computer weights and expresses quantitatively. Topping the list is a stock's recent earnings-revision history, which -- no surprise -- has proven highly correlated with future price performance. The system examines analysts' estimate revisions over the coming quarter, the current year and the following year; assigns a weighting to each revision and then divides the net upward revisions by total existing estimates to arrive at a ratio between zero and one (the EPS estimate-revision index in the far right column of our screen). The computer also sorts for "group alpha," an industry group's performance relative to its past history. It compares each stock to its industry group on the basis of various price-to-earnings and price-to-book-value measures. And it screens for percentage changes in quarterly and annual earnings, as well as for companies that have a history of upside earnings surprises. "You're just trying to stack the deck in your favor," Scott explains." The full article is available at: interactive.wsj.com