To: Kirk © who wrote (1255 ) 9/29/1998 9:15:00 PM From: Justa Werkenstiff Respond to of 15132
Kirk: Well thought out post. Re: " Truman. Click on his name and read his last few posts. Then remember that he has been sparing with Dipy and Rill for months if not a whole year yet he only complained about buying too soon at 8650 and he was gone.... My take was all the work he put into the thread did nothing for his investment return above what following a DCA or follow the newsletter approach would yield." Again, I would prefer to let Truman speak for himself if he chooses to do so. Re: "My concern with the model is more an issue of credibility. Bob issued VERY STRONG statements about what his model (and he) was predicting, especially the bottom of the correction." True, he was driving the model close to the curb in calling the bottom with the original benchmarks. He was selling his analysis hard because he believed it. I see no credibility problem for exactly that reason. Re: "Even now, he makes a big deal about some magic closing number of "less than 20%" which is really silly in my view as the only real number that matters is what the market does in real time, i.e. intraday lows and highs. Open and close levels are just snapshots of realtime events and yet he goes on like the closing numbers are so important. Maybe I don't see the important reason the he does with his much larger experience level." So much can happen in a day. Look at last week. There were huge intraday swings. In order to identify the trend, you have to reduce the noise. End of day reduces noise and end of week reduces it further. Although you did not come out and say it, maybe it bothers you that the correction went down more than 20% intraday. Re: "Sure there is some closing out of positions by traders, but it just seems a bit strange to say a model isn't that accurate then post and broadcast how important a few tenths of a percent are for his accuracy." You are going to have to help me here and go into further detail if you so choose. I don't understand. Re: "I guess I am saying he can't have it both ways and I prefer the less accurate method which is more arm waving and less TA." I might understand this point here if you choose to clarify what you said above. Re: " 'He did say that if he got an intermediate sell signal he would be predisposed top go to cash as most bear markets are preceded by an intermediate correction. ' Exactly, just in different words. OK, say his model HAD correctly predicted this intermediate term correction, then would he have had us going to cash rather than buying as the market went from 10% down to 19.8% down? Why or why not? What has he learned?" Fair point, but if this whole episode is still playing out in our minds, it is still playing out in his mind as well I suspect. We are in the middle innings of this correction and I will not draw any conclusions of what lessons should be learned until the game has played out some more. Re: " ' I don't think you can adjust what is essentially a rational model the vagaries of the marketplace...' Actually it is rather simple, you increase your error bars. Every time you get new data that you can't fit to your model, you either widen your error bars to cover the data or throw out the model. You may end up throwing out your model anyway if and when the error bars get too wide to be of any use. To me, it seems that the error bars added 9.8% to the predicted timing model numbers since Bob was so gung-ho on it only being a 10% correction within 3% error. As a "timing model", I think we may be close to this (throwing it out) though it still seems to be an excellent "market direction model"." I disagree for the reasons stated above. No conclusions can be drawn at this time. Re: " Remember, I am an engineer and I have been spending the last year modeling the behavior of photon detectors under bias and extreme power levels so I have experience in what I am talking about....just in a different field." I disagree here too. Market analysis is much more of an art than a science. Re: " It would be an interesting career change to try working on Bob's model..... as the number of variables are many,many times higher so it seems that it has to be much more statistical and thus I would not rely on how exact it is, but rather how well it predicts direction and trends." I do not see how one can quantify the emotions of investors to perceived political and economic risks. Re: "Knowing that the market goes up sure make it easy to buy on 10% dips and even easier on 20% dips (I've made about 10% already on the bargains I scooped up in stocks on the day we hit the low...)" I posted on the day this thing breached the 10% level that the most important thing overall is to be on the right side of the market regardless of the numbers which serve only as guidelines.