Nope, we’re not even calculating any infinity. We’re calculating a growing finite setup of discrete states. I’m not sure if you understand that it’s always finite for all its conditions even if the set itself grows (so long as it is finite at the time we’re calculating).
I’m probably repeating what Autolykos said above about “solutions” and “problems”, I think…
The problem with using high frequency trading techniques based on algorithms is that it is difficult to factor in uncertainty and expectation. How do you quantify something that is hunch, a gut feeling, etc.?
I find this topic interesting since I trade relatively actively.
(Side note: Did I read correctly in a different thread that you play EVE Online, ladyattis?)
Yeah, I’m playing right now. It’s fun when you have enough skills to fit a ship the way you want. I got a Myrmidon which I fitted it with a target painter and projectile turrets (since it has no hybrid bonuses).
Interesting subject that doesn’t get enough play IMO. What I see as the major flaw for HFT is its lack of subjectivity, mathematical algos can compete on an objective basis, but what hinders algo’s from always optimizing trades is subjectivity. The markets need subjective actors who follow their own valuations against the proper objective analysis, in order to utulize trading and profit. Subjectivity is needed to recognize serious flaws of valuation that a purely objective algo will not percieve. since prices are subjecticve, algos will always be a defective tool in trading.
Subjectivity is important. I don’t see the problem with HFT in terms of economics. I don’t think algos are defective for trading. They are just a part of the market and can’t be a complete market. (unless we consider the algos to be AI with goals and think about means to achieve ends)
1)Prices can’t be “wrong”
2)Those who run HFT systems will either learn to make better use of flash crashes (or more likely) create more robust systems that will be less susceptible to flash crashes or become less likely to create them.