Quant: What Exactly Is It?

What's quant?
Before i get practical, let me rhapsodize on what quant is exactly for a moment. Quant is nothing less than the sheet music of the spheres, the rhythm of the universe, why bees make honey, the way things really work. okay, enough of that- for now- but as you get into it, such thoughts become irresistible.

one scientist, Dr. Bernard Chouet, finally used quant to predict a volcano and save many lives. it was described on the PBS program NOVA (read interview). quant is also used to anticipate adverse cardiac events, and many social phenomena such as population migrations and economic changes. engineers and other scientists often use quant to predict everything from quantum particle positions to the size and locations of as yet unseen astronomical bodies (which later on prove accurate).

but you don't need a PhD to use quant- all you need is some spreadhseet skills. i'm just showing you the kind of horsepower you'll have under you when you pull the trigger on a quant play. here's another example of that horsepower...

in my own early experience with quant, as a General Electric engineering researcher, i worked with the manufacture of diamonds (see wikipedia article). real diamonds, not "zircons" (zirconium oxide), but they're called "synthetic" because they're human-made. i also worked with Borazon (cubic boron nitride, another super-hard crystal). my job was to figure out how to make the maximum amount of a given size crystal per each run. hmmm- compare this to forecasting how many glass particles of a certain size result when you strike a marble with a hammer. kinda depends, huh? - depends on how hard you strike it, how hard the glass is, how hard the hammer is, how hard the underlying surface is, how much follow-through is in the blow, the angle of the blow, the temperature of the marble at the time... that enough variables for you?

now multiply that by a factor of about a zillion and you see how complex it is to forecast diamond particle size distributions, where you have infinitely variable temperature, infinitely variable pressure, infinitely variable time factors, and infinitely variable rates of acceleration of these variables. yes, infinitely variable because, after all, there's an infinite number of specific temperatures just between, say, 1700 degrees and 1701 degrees (1700.0001, 1700.0002, etc.). well, i'm no genius, but i was consistently able to figure it out with quant. now that's horsepower- and with the right tools, you can do it with stocks at the click of a mouse (we didn't have pc's when i was doing this back in the 1960's- all i had was paper, pencil, and a calculator).

enough of the flashdance- have i finally got your undivided attention?

as mentioned earlier, quant and t/a have certain things in common, and certain differences...

1. both rely mainly on OHLC-volume data (open-hi-lo-close-volume), although quant can reach out for other data as well.
2. t/a uses "canned, one size fits all" algorithms, i.e., the same indicators for all stocks. with quant, YOU create fresh algos for each stock, and each algo is "almost" as individual to that stock as a fingerprint- well, not quite that individual, maybe more like being as individual as the combination of hair color, shoe size, and date of birth.
3. with t/a, you might make several trades per day (and probably get "stuck" for weeks 1 time in 4). with quant, you make at most just a couple trades per day and almost always exit profitably within 1 to 3 days (ah, FOREX trading is a bit different, however, with trades lasting minutes, at most).
4. with t/a, all the other t/a traders are seeing the same opportunity you are. quant is stealth trading- it's your algorithm and only you see the opportunity coming, often days ahead of technical traders.
5. with t/a, you have to screen large numbers of stocks every day to find good trades. with quant, you slowly build a personal stable of stocks fitted to algos and just watch those. the rest of the time you go fishing or to the movies or spend time with the family.
6. t/a is a good system, but it takes almost no imagination. quant requires creativity- you have to hunt for relationships between algorithm elements. sometimes it's easy, sometimes it's hard. but in the end, it's super-rewarding, both financially and emotionally.
7. as a t/a trader, you're always asking yourself, "What is?", i.e., "What is this indicator saying to me?" as a quant, you're always asking youself, "What if?", i.e., "What if I compare this value to that value?, What if there's a relationship between this and that?", etc., and then you go try it to see.
   

i hope this gave you a fair idea of what quant's about. next, i'll cover the main software tool you need to do quant.