My last post on the options forum went south so lemme try another one, hopefully with different results.
I've some experience with options, mathematics and programming. In reverse order as enumerated
Lately I've refactored the self-made software I was using to backtest trading strategies (with options, that's why I post in the options forum) and in the process caught some bugs and got some new ideas. Damn having a reliable system is hard, when you support thousands of use cases (underliers) and "bugfixing" experience looks exactly like updates on your phone: fix a thing with one underlier and have dozen others fail, while previously they were working just fine.
Therefore I'm relaxing my demands on the system from "top-notch handling of everything" to "top-notch handling of something". Especially when there's enough liquidity in "something", like maybe 100-1000x every other crappy thing.
So I'm concentrating on a few high liquidity use cases and noticed the following. I'm busting my head at the moment to produce a solution for a trading problem which is probably already solved, but:
1) Googling for it doesn't yield results, neither immediate nor after lots of digging.
2) I can't just go on Wilmott or some quant forum and ask "how do you solve this crap?" because in trading 99% of the edge that you have is formulating the damn problem. Solving it afterwards it's indeed "simple".
Why I'm saying it's probably already solved: because there's a liquid market for it. But on th eother hand there was a liquid market for options even at the turn of the last century. At 1900, most of the usage of the trans-Atlantic underwater cables was to send messages between London and New York stock exchanges in order to trade options. Yes, HFT. But it was only at 1900 that https://en.wikipedia.org/wiki/Louis_Bachelier laid out the foundations of the rigorous treatment of options, then came two wars, then it took a Japanese guy named https://en.wikipedia.org/wiki/Kiyosi_Itô to lay te foundations of stochastic differential equations ( https://en.wikipedia.org/wiki/Itô's_lemma ) then about another 20 years till Black, Merton and Scholes figure out with razor-sharp precision how shit works.
Why is this relevant? Because even though there was a very liquid options market for 100 years, it was "inefficient". Armed with knowledge of the true way things work you would have made a fortune. It's like playing dice before the invention of probabilities: there was a liquid market for gambling dice for thousands of years, and they somehow made it work based on intuition and plain experience, but a math guy would have made a few castles worth of money of the ignorants playing them (and actually did: Chevalier de Méré https://en.wikipedia.org/wiki/Antoine_Gombaud ).
So I the fact there is a liquid market for something doesn't in any way imply the thing is perfectly efficient and known. And as I said:
1) Fact I don't readily find a solution to my trading problem on a very liquid market, might be a sign the problem is not fully solved if at all.
2) Problem's I can't ask other guys for the solution to the problem because knowing there's a problem in the first place is a competitive advantage.
So sorry guys, can't ask you for a solution to my problem since just formulating it means someone with more resources will solve it long before me.
But I can formulate a helper problem. That was the whole idea of my previous post before I blew a gasket. How do I find a trustworthy associate willing to associate on research problems. Have I said that trust is of the uttermost importance? See (see my "Nursiki" thread on Chit-Chat).
I've some experience with options, mathematics and programming. In reverse order as enumerated
Lately I've refactored the self-made software I was using to backtest trading strategies (with options, that's why I post in the options forum) and in the process caught some bugs and got some new ideas. Damn having a reliable system is hard, when you support thousands of use cases (underliers) and "bugfixing" experience looks exactly like updates on your phone: fix a thing with one underlier and have dozen others fail, while previously they were working just fine. Therefore I'm relaxing my demands on the system from "top-notch handling of everything" to "top-notch handling of something". Especially when there's enough liquidity in "something", like maybe 100-1000x every other crappy thing.
So I'm concentrating on a few high liquidity use cases and noticed the following. I'm busting my head at the moment to produce a solution for a trading problem which is probably already solved, but:
1) Googling for it doesn't yield results, neither immediate nor after lots of digging.
2) I can't just go on Wilmott or some quant forum and ask "how do you solve this crap?" because in trading 99% of the edge that you have is formulating the damn problem. Solving it afterwards it's indeed "simple".
Why I'm saying it's probably already solved: because there's a liquid market for it. But on th eother hand there was a liquid market for options even at the turn of the last century. At 1900, most of the usage of the trans-Atlantic underwater cables was to send messages between London and New York stock exchanges in order to trade options. Yes, HFT. But it was only at 1900 that https://en.wikipedia.org/wiki/Louis_Bachelier laid out the foundations of the rigorous treatment of options, then came two wars, then it took a Japanese guy named https://en.wikipedia.org/wiki/Kiyosi_Itô to lay te foundations of stochastic differential equations ( https://en.wikipedia.org/wiki/Itô's_lemma ) then about another 20 years till Black, Merton and Scholes figure out with razor-sharp precision how shit works.
Why is this relevant? Because even though there was a very liquid options market for 100 years, it was "inefficient". Armed with knowledge of the true way things work you would have made a fortune. It's like playing dice before the invention of probabilities: there was a liquid market for gambling dice for thousands of years, and they somehow made it work based on intuition and plain experience, but a math guy would have made a few castles worth of money of the ignorants playing them (and actually did: Chevalier de Méré https://en.wikipedia.org/wiki/Antoine_Gombaud ).
So I the fact there is a liquid market for something doesn't in any way imply the thing is perfectly efficient and known. And as I said:
1) Fact I don't readily find a solution to my trading problem on a very liquid market, might be a sign the problem is not fully solved if at all.
2) Problem's I can't ask other guys for the solution to the problem because knowing there's a problem in the first place is a competitive advantage.
So sorry guys, can't ask you for a solution to my problem since just formulating it means someone with more resources will solve it long before me.
But I can formulate a helper problem. That was the whole idea of my previous post before I blew a gasket. How do I find a trustworthy associate willing to associate on research problems. Have I said that trust is of the uttermost importance? See (see my "Nursiki" thread on Chit-Chat).
