What math formula do algo/maths-based traders use?

[MrGecko]

EDIT: I thought I should use this opportunity to say that I am with GJ here. TA, IMHO, is a bunch of waffly mumbo-jumbo. That is to say, the only difference between quantitative methods and TA is the quality and rigor of the analysis, nothing more.

Out of interest, if you were developing a rule-based technical strategy* (or, say, someone approached you with one), at what point would you be satisfied that it was worth a pop?

And pursuant to your first point... I would agree that the rigor threshold for TA is... well non existent for the most part (if someone believes it, then it's legitimate TA apparently), but I would balance that with the point that most sensible proponents of TA consider it more qualitative information than quantitative, and treat it as such.

all IMHO.


* not stat/pure arb, HFT, but the type of strategy you read about here for example.

 

[arabianights]

1 + 1 = 3, or thereabouts

 

[FX Bandit]

1 + 1 = 3, or thereabouts

Christ, ya dontarf talk outta yer arris.

1 + 1 = 11.

 

[MrGecko]

1 + 1 = 3, or thereabouts

Close, but no cigar... Bill Gates, for example, built his wealth upon this:

 

[Martinghoul]

Out of interest, if you were developing a rule-based technical strategy* (or, say, someone approached you with one), at what point would you be satisfied that it was worth a pop?

Well, as it happens, some of the analysis I do for certain trades can be thought of as similar to what you describe, as it's based on some very bog-standard time series analysis techniques. The point there is that I understand the theories and the assumptions behind them relatively well, which allows me to have a degree of confidence in analysis that's based on these theories. That's actually the point of the two equations I have quoted.

So, I suppose, the most accurate summation of my view would be that I believe, have studied and have a wee understanding of mathematics. That implies that I believe in analytical tehcniques that are based on fundamental mathematical concepts and axioms. Most TA methods that I have encountered couldn't be further from that: either their proofs violate simple principles of mathematical logic or they are too vague to be formulated in the first place.

My Z$2c...

 

[MrGecko]

Well, as it happens, some of the analysis I do for certain trades can be thought of as similar to what you describe, as it's based on some very bog-standard time series analysis techniques. The point there is that I understand the theories and the assumptions behind them relatively well, which allows me to have a degree of confidence in analysis that's based on these theories. That's actually the point of the two equations I have quoted.

So, I suppose, the most accurate summation of my view would be that I believe, have studied and have a wee understanding of mathematics. That implies that I believe in analytical tehcniques that are based on fundamental mathematical concepts and axioms. Most TA methods that I have encountered couldn't be further from that: either their proofs violate simple principles of mathematical logic or they are too vague to be formulated in the first place.

My Z$2c...

Well I can't really disagree with any of that, but that is not the gist of my question...

... lets say I approach you as a potential investor. I have a systematic trading strategy, but I am not going to tell you what it does. Ever.

All I have to convince you are the results - condifence intervals, sample sizes, in-sample and out-of-sample testing, real returns, correlation with other asset classes, Sharpe ratio's etc...

What does it take for you to bite?

 

[Martinghoul]

Well I can't really disagree with any of that, but that is not the gist of my question...

... lets say I approach you as a potential investor. I have a systematic trading strategy, but I am not going to tell you what it does. Ever.

All I have to convince you are the results - condifence intervals, sample sizes, in-sample and out-of-sample testing, real returns, correlation with other asset classes, Sharpe ratio's etc...

What does it take for you to bite?

To be totally honest, before 2008, I would have said that, if I were happy about all those metrics you mentioned, I'd bite. However, I have become horribly skeptical after the crisis (and it's got nothing to do with Madoff) and would have a very hard time entrusting anyone with my hard-earned money.

I suppose that I'd do it if I had some spare cash and were willing to treat it as a cheap punt, but that's about it.

 

[GammaJammer]

1 + 1 = 3, or thereabouts

Christ, ya dontarf talk outta yer arris.

1 + 1 = 11.

Wrong, wrong, all wrong

1 + 1 = 10

 

[andrewdb]

O'Brien says 2 + 2 = 5

 

[Dommo]

Wrong, wrong, all wrong

1 + 1 = 10

so 10 + 10 = 100 = 10 * 10

 

[GammaJammer]

so 10 + 10 = 100 = 10 * 10

Deep man, deep!!

 

[Dommo]

Deep man, deep!!

Thanks - I took a moment to reflect on your arithmetic while I was in drawdown of (approximately) £10000000000