Again BS. I have done so many backtests and not found any wrong results. Also on the Amibroker yahoo list there are so many experienced people (more experienced than you will ever be) having done dozens of tests. Until now not one serious problem of the backtester has been mentioned there. Amibroker is known for being one of the fastest and most reliable backtesters around. And as seen here your whole claims have been put to dust by visual proof. On top of that I don't take anyone serious who calls five small simple lines of code "being complicated". Laughable. You have something to hide and that's fishy.
I said, I have no time for taking screnshots now and doing long posts. You have to wait. You will see. Nothing fishy from me. You will see somehting fishy from the software soon. You have to have patience. It is a human quality you lack.
Then also provide your badly coded script. Not just manipulated screen shots. Do the same that you demand from others What a pathetic excuse. It takes me a few seconds to do a screenshot and max 5 minutes to make a video. Posting the code is just another seconds. Sooo ridiculous.
Code: SetBacktestMode( backtestRegularRawMulti ); Buy = 1; Sell = Short = Cover = 0; StopLevel = Param("N-bars", 1, 1, 100, 1 ); ApplyStop( stopTypeNBar, stopModeBars, StopLevel, 1, False, 0 ); BuyPrice = C; SellPrice = O; SetOption("InitialEquity", 100000); SetOption("FuturesMode", False ); SetOption("UsePrevBarEquityForPosSizing", True); SetOption("ActivateStopsImmediately", False); SetOption("Allowsamebarexit", True); SetOption("AllowPositionShrinking", True); SetOption("MaxOpenPositions", 1); SetPositionSize( 1, spsShares ); //Reports SetOption("PortfolioReportMode", 0);// 0-Trade list, 1- Detailed Log, 2- Summary, 3- No output SetOption("GenerateReport", 1); // force generation of full report RoundLotSize = 1; You are too nervous and impatient. Just add the option SetBacktestMode( backtestRegularRawMulti ); on top of your code. This is the option I had in the past and it should not have affected the entries and exits because the system does not generate overllaping signal anyway. But when that option is present, the backtest is screwed up as I showed in my screenshot before. Therefore, the option/function interferes with the backtest in an unpredictable way. Satisfied now? You took my time from other important things.
Well, finally after so many pages of chit chat. Jesus. Let's see whether T.J. is gonna reply to this. I won't comment to it. You took it yourself since you could have mentioned that in the first PM!
This is from year 2010 in regards to backtest raw modes and ApplyStop n-bar in comparison to SetBacktestMode( backtestRegular ) and ApplyStop n-bar http://finance.groups.yahoo.com/group/amibroker/message/149665
from the previous link As for different backtest modes (default backtest mode is backtestregular) http://www.amibroker.com/guide/h_portfolio.html
I don't care what he says. The fact of the matter is that leaving a seemingly unrelated function there messes up the backtest. I hope you and he do not expect every user of Amibroker to hung around in newsgroups. How do I know that other functions do not also create conflicts? This is the problem. Do you think I will spend time to find out? No, I will not. I have written by own backtesting program I am sure what it does. Edit: by the way I appreciate the work Tomasz has done and his program is much better as compared to other sh*ty ones that sell at 10X the cost. He is a good guy and a good software developer. But I think he must go ahead and do a function compatibility test. This is lacking from the manual. I think this discussion was maybe of more value to him.
I won't worry too much about what funnyguy says. He knows a few buzzwords and like to throw them around to impress people. He thinks 64 bit software should be faster than a 32 bit software. He thinks his backtest is the road to the holy grail. He must be waiting for people to beg him for his code.
Oh sbokov over there, I didn't say that 64-bit software should be faster. I wrote that there are softwares that run faster than their 32-bit version of same vendors. And FYI, I've already posted the code. So better buy some glasses. My backtest? It's not my backtest. Bill wanted me to do the job since he wasn't able to provide one himself. Troll somewhere else because here you failed badly.