Statistics and math behind it. Do I need it?

[globalarbtrader]

Hi everyone,
I'm ready for my first question here... every response is appreciated.

If I want use some statistics/analysis/whatever in my algo trading/strategy development, do I ever need to really know the math behind it (calculus, combinatorics, probability theory, theorems etc.)
OR
it's just good enough (from practical point of view) to use well known high level concepts: distributions, regressions, well known models, pricing models etc.
And use all this stuff like black box.

What do you say?

P.S. I do really well software development and all aspects of software engineering. It's not so hard to close the "math" gap, but avoid to waste a single minute.

Calculus you don't need unless you're trading some exotic hard to value options.

A rigorous and intutive understanding of statistical significance is vital for anyone who is looking at any kind of backtest results. Also some understanding of the modelling of financial time series would help, eg the typical behaviour of volatility.

GAT

 

[Mikkel]

QQQ 0.72; SMH 0.58; IBM -0.06; GLD -0.17

Maybe a noob question, but how can you obtain that data?

 

[DaveV]

I don't know where you can get the daily percent changes, but the End Of Day Historical Close Prices are available from many sites, some of them free. Check Yahoo Finance, Google Finance, or search this website. Once you have the closing EOD prices, then calculating the daily percent change is simple math. The Pearson algorithm, in almost any computer language, is easily found on the internet.

 

[sle]

Pearson Correlation, which measures how closely related two series of numbers are.

Yeah it’s also useful to understand the various artefacts that can be produced by it (which is why it’s sometimes useful to winzorize the data, use digital correlation or use basket vs component variance instead)

 

[truetype]

sometimes useful to winzorize the data, use digital correlation or use basket vs component variance instead)

Sorry, what's digital ρ? Maybe ρ of the signs?

 

[sle]

Sorry, what's digital ρ? Maybe ρ of the signs?

Same thing, yeah

 

[Buy1Sell2]

Hi everyone,
I'm ready for my first question here... every response is appreciated.

If I want use some statistics/analysis/whatever in my algo trading/strategy development, do I ever need to really know the math behind it (calculus, combinatorics, probability theory, theorems etc.)
OR
it's just good enough (from practical point of view) to use well known high level concepts: distributions, regressions, well known models, pricing models etc.
And use all this stuff like black box.

What do you say?

P.S. I do really well software development and all aspects of software engineering. It's not so hard to close the "math" gap, but avoid to waste a single minute.

You do not need the math behind it. Learn to exercise Prudent Risk Management.

 

[Mikkel]

I don't know where you can get the daily percent changes, but the End Of Day Historical Close Prices are available from many sites, some of them free. Check Yahoo Finance, Google Finance, or search this website. Once you have the closing EOD prices, then calculating the daily percent change is simple math. The Pearson algorithm, in almost any computer language, is easily found on the internet.

Okay but in Y! Finance, in the historical data section there are opening prices, highs, lows, etc. How can I separate them for use only the data I need?

Edit: I tried to open the file with a program that I have used on the econometrics subject (EViews) and now I have the data separated.
Thank you and excuse me for waste your time.

 

[Simples]

Hi everyone,
I'm ready for my first question here... every response is appreciated.

If I want use some statistics/analysis/whatever in my algo trading/strategy development, do I ever need to really know the math behind it (calculus, combinatorics, probability theory, theorems etc.)
OR
it's just good enough (from practical point of view) to use well known high level concepts: distributions, regressions, well known models, pricing models etc.
And use all this stuff like black box.

What do you say?

P.S. I do really well software development and all aspects of software engineering. It's not so hard to close the "math" gap, but avoid to waste a single minute.

Yes, the more advanced mathematics you can come over, memorize it, use it well, forget it.
It'll bring zhe $$$ profits and none will understand why, not even you.
Also makes you mysterious and get zhe chicks!