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 Rational Analyst -- Ignore unavailable to you. Want to Upgrade?

To: HeyRainier who wrote (206)	1/14/1998 10:16:00 PM
From: Galirayo	Read Replies (2) \| Respond to of 1720

[ 8-10 New Price Low] Hi, Rainier. 8-10 NPL and Bol Band Violations both work. Sometimes Both occur simultaneously. But if you remember, Sakata had the Theory. Method 8 to 10. Better than the 3's Method. IMHO I don't play the Larger Caps much. I like the littlen ones. The larger Caps are "Supposed to" Provide Stability. They seem to be the ones taking the Big Hits now. I think Soros makes his in Volatility. I want the Volatile ones and I want to see them on the Bottom or Bouncing. The 8-10 can find them as well as the Bol Band Violation, not to mention the Fibonacci Retracement. Take a Fib Retrace and look at it ... Up Side Down. I think they all work at different times. I'm concentrating on all the above and looking for Price Support on a Major MA. Thanks, Rainier. ... Thomas DeMark ... Will be looking for this one. Remember the Title? NCTI again today. It qualifies tonite as an 8to10 NPL with Support on the 89 and 100DMAs. It's also showing a Tower Bottom. Ray

To: HeyRainier who wrote (206)	1/14/1998 10:23:00 PM
From: Scott H. Davis	Respond to of 1720

[8-10 NPL question] I probably missed the post. But I am also interested in when B Bands, in conjunction with confirmations such as RSI and stochastics indicate an imminent downturn after an overbought condition. Specifically, how to pre-assess the magnitude of the retrenchment. Two companies I own, VICL and UNRG were overbought, I know they were about to have a fallback, I subjectively assessed that, since the overbought condition was not an exteme situation, that I would ride it out, since one piece of positive news would cause a significant pop. In VICL's case since it had several 98 pieces of good news following a lot of tax loss selling & heavy short interest, I thought that the rise ( 11's to 16+ in 98) was due to a positive combination of re-entry after tax loss selling, news, and after a strong up-trend, short covering. With UNRG, I exited twice & re-bought much lower twice in fall 97 when it was clearly overbought then oversold. Since this is very thinly traded, if one of the expected news items hits, it would also fly, so I elected to not cash out two days ago. (no, I'm not crazy, I diversified enough profits into a company with a solid earnings trend that I can afford to ride out a speculative). The Oct 97 UNRG spike may prove a good contrast to modest overbought/sold conditions for your analysis. But I wondered too much. Again, is there a way to objectively pre-assess an expected snap-back after an overbought or oversold condition?

To: HeyRainier who wrote (206)	1/16/1998 12:44:00 AM
From: ftth	Read Replies (4) \| Respond to of 1720

Rainier: Great call on EK! Still looks like there's more to come. On your Bollinger band observations, I think the standard deviation / volatility measure used in Bollinger bands is based on the regular statistical definition of standard deviation of an arbitrary data set. Option pricing models (Black-Scholes) uses lognormal volatility. Regular standard deviation doesn't give the correct result if used as the volatility measure in B/S because there's an underlying assumption that stock prices follow a log-normal distribution (a normal distribution (regular standard deviation) would allow negative prices; other option pricing models use weird empirical distributions, but in any case there is always an effort to skew the distribution so that negative prices aren't allowed). I just wonder if a lack of fit to the underlying assumed distribution has any relation to how well Bollinger bands perform as indicators (not to mention the observation interval). Maybe Peg could address this too. I'm diggin' deep here in the old brain, but I think there's a t-statistic that can be used to validate or invalidate the underlying premise of a normal distribution. I'm pretty sure this is part of the reason why Bollinger himself says never use periods less than 10. Also, isn't there only about a 16% chance that the price will move by more than a standard deviation in one direction over the time period of interest, meaning there is always a tendency to cluster about the moving average, and to swiftly move inward toward the MA if its statistics get out of line. Large cap issues may more closely approximate a normal distribution, with smaller, more consistent swings and less tendancy to gap. My point is (yes there was a point to this rambling) perhaps a measurement of "Bollinger efficiency" could be attached to the indicator so you have a gauge of how much faith to put in the signal??? dh