3 ms·
Quant hedge fund guy here. Astrologers look at the positions of the stars and planets. Astronomers look at the positions of the stars and planets. Over the pa
by lordnacho 5y ago
Quant hedge fund guy here.
Astrologers look at the positions of the stars and planets. Astronomers look at the positions of the stars and planets.
Over the past centuries, astronomers have discovered such amazing things like the distances to other galaxies, the paths of quite small rocks in our solar system, the age of the universe, redshift, cosmic background radiation, and so on.
TA, basically looking at the numbers that come out of the market, is the same. You can do silly unscientific things with it, or you can apply a falsifiable method and make real progress.
If you go down the scientific route, you will find the same data takes you down a rabbit hole of low latency coding, linear algebra, infrastructure, financial theory (time value of money, optionality), economics, and so forth.
- csee 5y agoHave you seen traditional TA methods applied profitably within a quant framework?
- lordnacho 5y agoWhat I've always missed with "Traditional TA" is a rigorous accounting of how to do it, and what the results are. You can do this yourself on a spreadsheet: - Get some data - Calculate the 50 bar moving average - apply some rule like "trade when it crosses this or that stat" - See what the PnL is. - Think about whether this is a fluke and whether it will keep working. Somehow none of the ads and articles I see do this completely. But yes, you can take the same data and use it effectively. Needs more than a spreadsheet, admittedly.
- csee 5y agoForgetting MA crossovers (which don't work), what are your thoughts on backtesting trendlines and support/resistance lines. Have you seen anyone try that?
- lordnacho 5y agoI have actually. It's not so easy to keep your degrees of freedom down, so it's hard to conclude anything. Or it was for that project.