An excerpt from an email from Hayden to the LinnSoft Yahoo Group on 03/27/02

Toby, Arnie, Duke, Graham and anyone else who would like to jump in, I'm going to respond to a compilation of comments from the latest emails. I printed the emails out and am using them as a basis for responding. It is really helpful for getting the synapses vibrating. And if there is one conclusion that we all seem to be saying in different ways it is Interpretation, Interpretation, Interpretation. TI's are part of the trader's bag of tricks and it's what we make of them that counts.

Arnie talked about the statistics of coin flipping which always remains at 50-50. No argument, but where statistics gets interesting in gambling is in poker and some other card games. I don't know the game but watched a TV show on the world poker championship in Vegas recently, and clearly there are experts who consistently beat out opponents. This doesn't help much in TI's for stocks, but it does tell me that there is knowledge that can beat the odds of 50-50 chance that 2 players will come out even in a hundred rounds.

Personal comment:
As a day trader I love looking over the results at the end of the day. Looking at just the movement of bars with no TI's, it's very easy to see what could have been a good trade long or short, or no decent trade at all. And a simple scan using change greater or less than 1 can provide a quote page with decent trades. In my new project of looking at many technical indicators, I would then cycle through the TI's of those stocks to see how well they indicated the move. (the purpose being to familiarize myself with a wider range of IRT tools).

The word "indicator" is absolutely correct. Broadly defined, I take it to mean "maybe it's in that direction." We want indicators to be as good as road signs, like "Hoboken, 5 miles," but that is our mistake.

Question: do you put more faith in
a) Technical indicators
b) Chat rooms
c) Brokerage analysts
d) News
e) None of the above?

Toby,
Thanks for getting me to expand my horizons about statistics and TI's beyond stocks and IRT. Our world abounds in examples of people manipulating facts to suit their wants and needs. And, of course, if an honest scientist asks the wrong questions in his experiment, makes a mistake in writing a program or is sloppy in gathering the data, plus many other possible gaps, then the results will not be true, and the statistics will lie. In the case of IRT, we all know how hard Bill and Chad work to give us the best in so many areas and yet they can make mistakes. I was assuming that for the existing TI's, the computations are correct.

When I said TI's should corroborate each other I was thinking of looking at one stock in one time frame and using multiple TI's to look at the same results. Agreed that even using the same TI in a 1 minute and 3 minute time frame can give a different reading, but that's logical and up to our interpretation.

Back to time frames:
Drawing on another of your emails where you explained that if you placed your stop too close you would have reduced your winnings by approximately $20k. I'd call that the luxury of long term trading. I'm working on a day trading model of a batch of stocks at 500 shares each. Yesterday, at 4:28 was my total high for the day and 8 minutes later the total was down $700. And it only got worse. That's why day traders are inclined to keep tight stops and not let hopes of an upturn later on influence their strategy. Plus, there is always the possibility of entering fresh positions later.

I keep returning to TI's and their connection to time frames. What a long term trader sees as a great moment to buy based on a daily chart and TI may be disaster for a day trader, and I suspect that short term traders of a few days to two weeks would have other criteria for buying and TI's that would not be appropriate for the long term or the day trade. This is a very specific example of interpretation that I believe we would all agree on (but please feel free to disagree).

Duke,
Enjoyed your statement of numerous lifetimes to thoroughly explore one TI. If I understand you correctly I would compare it to a poll about "Who are you going to vote for President?" and asking every American of voting age. But some bright statisticians have worked out formulas for asking a carefully selected sample to get very close to the total result and scientists don't run tests to their complete conclusion either. And your comment about the mistake of confirming one TI with another TI of the same type is good. It fits in with my idea of knowing the range of TI's well enough to play them like musical notes and finding unexpected combinations that reveal something new. Next, I admire anybody with the math skills and drive to break down the formulas for TI's. As an artist whose math education stopped at geometry, if I didn't have TI's graphically available in IRT with the click of a button (plus the power of tweaking the preferences), I probably would have stopped using the program a long time ago. Nevertheless, between the bars of actual price fluctuations and the graphics of a TI, it's relatively easy to corroborate the accuracy of a particular TI. And then go on to see how accurately the TI is indicating an entry and exit point. Now we get to my personal interpretation and use of a TI: all I really ask is to give me a good buy signal and I'll take care of the selling. And that the buy signal is accurate over 50% of the time - which moves into scans and backtesting.

Risk management: Absolutely. The first thing I go for in a trading system is the stop loss feature. If I can tweak that pretty well the rest is relatively easy.

Comments about Jones Law:
Know your stocks - for better or worse I don't know much about the specifics of any company. I know which sector a stock is in, but I have selected my stocks because they are major names with heavy volume and volatile enough for day trading. So "knowing" seems open to interpretation. Trade outside your comfort zone - wonderful. I think the dominant myth in short term and day trading is to only allot a small portion of a portfolio to such a risky venture. I'm coming to the opposite conclusion. Since you can control your losses each and every day and since the only way to really make impressive financial gains is through trading a large number of shares per trade, pour a maximum of your portfolio into day trading, BUT ONLY WHEN YOU HAVE THOROUGHLY TESTED AND RETESTED YOUR STRATEGY.

And finally I'm going to close with an observation about the last couple of years where I have seen my long, long term portfolio dwindle dramatically along with a few other million people. Why didn't the major brokerage houses research departments catch the trend? Overlooking the possibility that maybe they did see the trend and they were prevented from saying anything for fear of a broad collapse, the metaphor I think of is that everyone saw that it was raining but there were no clouds (clouds being the indicator of rain - see, some indicators are timeless).

Sorry for the length. Good to get this off my chest :-).

Hayden