Backtest help

There might be something with Donchian Channels. They find the lowest and highest points in the lookback period, always struck me there had to be some way to connect the changes to draw diagonal S/R lines. Perhaps from a given point, go forward and connect to the other points that push the Donchians; whichever one results in greatest area between the Donchian line the highs/lows (in between, that did not push the Don line) would be the ending point. But this is only an idea, I never actually worked it out or figured out how to do it.
 
Quote from Joe Doaks:

I asked because to me there are four different kinds of trades: a daytrade, a swing trade of days, a position trade of weeks, and an investment of months. I approach each differently from an S/R viewpoint. Trailer Trash Trading.

Understood, then in your eyes, I aim to be a swing trader, but when the time is right, a position trader.
 
Wait a minute, Jgills do you know what backtesting really means?

It means having an automated strategy run on charts from the past and see what it would do. To me it sounds like you're asking advice for discretionary trading, not automated. Do you want to "backtest" your own trading decisions or a computer's?


Quote from tradingjournals:

If you knows the results for N(0,1), one can know the results for other normal distributions."

Except when normal distributions change together with other variables and produce different curves, like per volatility, per sample period, etc. I know what you mean, but that approach is only applicapble to a limited set of circumstances. You still need 2d or 3d tables sometimes, or you need a table of N(0,1).
 
Quote from Jgills:

Understood, then in your eyes, I aim to be a swing trader, but when the time is right, a position trader.

I asked because it makes a difference. Swing trade entries want to be made on retraces of the previous day's action. Position trades want to be made at the end of panic or greed reactions. Investments want to be made on secular changes.
 
Seems like some complicated/expensive solutions here. IMO, the easiest way to practice is download some free yahoo data into excel and start by coding the most simple rule you can think of. (i.e. buy tomorrow, sell after tomorrow).

Go from there depending on what you want to test.

...as others have pointed out, if you want to test a rule against support/resistance you will need to clearly define what that is. No way around that.

GL.
 
Quote from Soon2Bgreat:

Seems like some complicated/expensive solutions here. IMO, the easiest way to practice is download some free yahoo data into excel and start by coding the most simple rule you can think of. (i.e. buy tomorrow, sell after tomorrow).

+

Amibroker is IMO the best backtesting tool in the market for the money. www.amibroker.com

You can even use the free version and modify the existing sample startegies. I think they have support/resistance level code in their free library. It downloads data from yahoo. No need to pay anything.

Before that, it would be a good idea to read a little about backtesting and what you can expect from it. Here are some links:

http://en.wikipedia.org/wiki/Backtesting

Read about postdictive errors here. This is very important:

http://www.tradingmarkets.com/.site...iases-in-Backtesting-Forex-Strategi-80442.cfm

This is a good blog on the subject. A must read:

http://www.priceactionlab.com/Blog/2010/10/proper-use-of-back-testing/
 
Back
Top