MoneyTec 2.0 - Online Forex Traders & Brokers Social Networking Community » Trading Methodologies » Tools of the Trade

[cummish]

Hi everyone,

I am after some specific,

I wish to back test my trading.
I do not wish to use an automated system, mine is wholly manual.

What I do need though is a piece of software that I can allow me to go back in history and see the market evolve in real time (tick by tick)

I am currently looking into Neoticker.

Any suggestions would be much appreciated.

Cheers, Shane.

[darmasdt]

you can try metatrader

[Medran]

hi Cummish

> What I do need though is a piece of software that I can allow me to go back in history and see the market evolve in real time (tick by tick).

It all sounds so easy... this backtesting holy grail... many perform simplitic back-tests but the reality of a truly objective test is that it is hideously complex.

I have written a bespoke application to back test and along the way have discovered how complex a seemingly simple strategy is to back test.

Take the simplest strategy in the universe... the good old moving average cross. I've programmed a back-test to definitively prove if you can actually make money with this... and so it begins:-

EMA or SMA, periods for each MA... what to average (open, close, hi, low, (hi+low)/2? buy at market close last tick... or same with a percentage for noise? Program in the spread for your dealer. Buy/sell when the cross is at a great angle (approching 90 degrees)... or maybe when the MA's touch but don't cross? Buy/Sell only if the price is above/below fast MA?

Then shall we go for 1 min, 5 min, 15 min, 30 min, 60 min, 1 day, 1 week, 1 month?

You can produce a program that loops through all of these permutations for the last five years (though the fastest computers availalable today will take many days to complete) and answers the question of EMA/SMA?, which periods?, best strategy for minimal drawdown?, best strategy for best win/lose ratio?, best strategy for minimal risk? ... and so it goes on... even for the simple MA cross this would be difficult if not impossible even for programmers with years of experience.

But remember the above example is for the simplitic MA cross, what about if you want to back-test Elliot waves... think about how many control parameters you will be into here.

Backtesting is doable (just) but I'm not sure anybody has actually managed to program a totally optimised model to definitively answer the question "if I follow this (mechanical) strategy to the letter will I make money"?

But it does keep life interesting!

Have fun