Booms, Busts, and Recoveries Message Board

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 : Booms, Busts, and Recoveries

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next

To: shades who wrote (68643)	9/24/2005 11:17:13 PM
From: Moominoid	Read Replies (1) of 74559

We ain't at October yet. I don't pretend I can predict any actual value or time frame, but turning points are a bit easier and it is all a matter of probabilities. On gold, I have consistently been sending savings to Australia whose dollar is closely gold linked since I got here in 2002 on the basis that it was cheap and so was gold and it should go up. That was till this year when the USD started to rise and Igold go straight. Now waiting for Aussie to fall a bit... I've told you the model before and you paid no attention, so I didn't think you really wanted to know. I think you just want to criticize. But here it is again: You just estimate the following regression on the logs of stock prices: y(t) = a + by(t-1) + u(t) You do this for a variety of time frames (usually I use 8,13, 21,34 etc. days the Fibbonacci series). And you continually move the window. Excel does all this simply and easily. The parameter b varies over time. For very long samples it is close to the expected value of 1 (random walk). But for shorter samples it varies a lot. In the shorter sample when it gets to one it is often a turning point. There are false signals so confirmation with another method helps. There is at this stage a lot of art to using the method.... I also look at both daily and weekly data... Using multiple methods certainly helps. I also use Bollinger Bands with Fibbonacci periods. Checking the way different moving averages (with Fibbonacci periods) stack up is a method I discovered this summer. When they are all stacked up in the right order you are likely at a major bottom or top. Then MACD, stochs and those methods are helpful for confirmation and determining E-Waves. For the major indices I use the McClellan Summation, and the Bullish percentage. Highly useful. And I try to compare across markets and asset classes to get a consistent picture. None of these methods can predict when something will reach a price target or what level. All that I think is guess work, though there are often strong hints it is still just guessing. The methods above have a consistent basis to them and tend to coincide. What they do is break a series into a stochastic trend and stationary noise around the trend in different ways. It makes sense from a time series perspective. E-Wave is about fractals that's all. The basic insight to distinguish corrective and impulsive waves is valid. Wave counts sometimes though really need to be stretched to fit the theory.... There are plenty of TA methods out there that don't make sense and plenty that make sense but aren't useful. So all this modeling is just time series analysis, of a strange kind, but nothing more or less. When exogenous shocks hit the series there is a limit to what they can do and that varies depending on the past history. So it doesn't matter what market makers or rednecks do. It is all in the time series. Just like I can apply similar methods to the climate or the macro-economy.

Report TOU Violation

Share This Post

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next