Thursday, October 19, 2006

Two Systems, Different Outcomes. Why?

From just the last two blogs you can see the power of quant. i could also display the power of t/a, instead of showing how bad it can be. instead, here's my unscientific observation about the two methods: on average, quant's forecasting accuracy is about 85-90%; t/a has a forecasting accuracy of about 60-70%, and seems to be dropping. that's quite a difference.


But why? Both systems are based in exotic mathematics (you don't need to know much beyond how to count to 10 on your fingers or toes to use either one). Both systems tend to rely mainly on price and volume figures. So why the dramatically different success outcomes? And why, when you Google for quant trader (it'll open a new window for you), do you see so many big banks and investment houses willing to pay such astronomical salaries to hire quant traders?

Here's what's up. simply put, t/a is being used by too many people and the really big fish are starting to use that to jiggle the table up and down, and up and down, as discussed earlier. they make money on the runup, then make more money on the short as it comes down. they read the t/a, and they use quants (like me) to figure out how to slaughter the t/a traders. oh, no- it's nothing personal- it's just business. so the best defense here is to use quant. and i use it like a chainsaw. when i go into the market, it's to steal a sizeable chunk from the big guys and run like crazy.

well, i gotta get a few hours' sleep before the market opens. i'm not hunting today but you never know. sometime this weekend i'll explain more about exactly what quant is. meanwhile, if you want to see a very simple quant algorithm that you can use (the ones shown in this blog are very exotic), you can get a free download of the R algorithm at whattolearn.com . Just click the Free Download at top of page.

No comments: