That's correct. But using rotational trading will give me a completely different approach. I want to Backtest trading rules on portfolio level. See if these rules give me in a real world environment with let's say 20 stocks in portfolio a profitable system yes or no. For me the only way to find out if these rules are profitable is using the Explore function and exporting the result to a spreadsheet for further analysis. Only then you will get all the signals. And in most cases the result I get is different than the one from the Backtester. Simply because of the portfolio constraint. The Backtester is only using a small part of all created signals. And I have the feeling that the occurrence of these signals have a random character. You just do not know which signal will be used by the Backtester. But your right, with rotational trading you will get a systematic rebalancing of the portfolio ...
Now my problem is that I just do not know if I am missing something in the Backtester approach or that I am just trying to do something that by definition is impossible ... Regards, Ton. ----- Original Message ----- From: vlanschot To: [email protected] Sent: Wednesday, August 29, 2007 11:35 AM Subject: [amibroker] Re: How do I backtest placing a restricted number of limit orders each night? Hi Ton, Perhaps looking at the rotational trading version can help to clarify things (I hope). In line with your argument, new signals have no effect UNLESS some condition is met which rebalances the portfolio. This is easiest perceived in rotational mode, since it forces the portfolio to "rebalance" at each bar. I hope we can agree that one has to have some conviction as to the explanatory power for excess returns of the "factors" (or think "indicators") which are used to define the condition. In other words, any score on a factor (momentum, valuation, etc.) implies its relative expected return, i.e. a higher score is preferable. If we agree on that then, based on your condition, unless an existing holding meets this condition (i.e. has the minimum score) it is replaced by another security (assuming at least one meets this condition). In case of the condition being met, any new signals are legitimately "superfluous" in that they are not better signals. Otherwise our previous agreement falls apart. The only way, in my view, in which new (or rather confirming) signals are put into practise in the portfolio is by adding/deducting to the weights of existing holdings (particularly if you're judged against a benchmark) which is where scalein/out comes in. Now, first, any rotational system can (often more flexibly) be implemented via ordinary BSSC-rules. Second, I do agree that there are limitations to backtesters, even AB's CBT. The main one is "custom cash management": the inability to allocate cash from individual sells (which should be completed first) to individual buys. An extension of this is the inability to use cash from shorts to enter additional longs, i.e. create 130/30 portfolios (although TJ has promised to look into this functionality). May be too much OT, but hope it helps. PS --- In [email protected], "Ton Sieverding" <[EMAIL PROTECTED]> wrote: > > Thanks Mike. I know all this. Please read my answers to Ed and you will find the real problem I have with the Backtester and whatever Backtester. Because it has nothing to do with the AB Backtester. It's just the portfolio constraint that every investor in the real world has creating mentioned problem. I just don't know how to solve it ... > > Regards, Ton. > > ----- Original Message ----- > From: sfclimbers > To: [email protected] > Sent: Tuesday, August 28, 2007 9:19 PM > Subject: [amibroker] Re: How do I backtest placing a restricted number of limit orders each night? > > > Ton, > > Once your portfolio is full, yes, PositionScore will have no effect > until a slot becomes available after a Sell. > > However, a PositionScore is only good for the life of the bar (single > day when using EOD data). So if it can not be acted upon *in that > bar*, then it is worthless from that point on. The markets will have > changed by the next bar and the score will no longer apply. > > Once one or more slots become open (after a Sell), then the *current* > PositionScore(s) will be considered, and the best will be used to > fill the open slot(s). So no, the process is not random. The > *current* PositionScore is used to fill any open slots of a > portfolio. "Expired" PositionScore(s) are of no use. > > If you have a restriction in your strategy that prevent entering > multiple positions for a single symbol (i.e. prvents "scale-in", > which is the default case), then entering a position for that symbol > will be rejected, even if it has the highest PositionScore. But that > is based on your strategy, not luck. > > The next highest will be evaluated until one is found that can be > traded in accordance with the rules of your strategy. The process is > predictable. > > Hope that helps, > > Mike > > --- In [email protected], "Ton Sieverding" > <ton.sieverding@> wrote: > > > > As far as I understand Ed and assuming EOD trading, PositionScore > is selecting the best signals coming from the same day. So when at > Day1 there are 10 different signals where only one is needed then > PositionScore is selecting the 'best' signal. But what if the next > day all stocks in portfolio are filled and the system generates > another 10 signals? They are lost ... until the system gives a SELL. > Therefore the next BUY is based upon the next SELL. Put the 500 > stocks of the SP500 in a WatchList, take whatever AFL rules for the > BUY and the SELL and a portfolio with say max. 10 stocks. Do a > Backtest and what you see is that only a small part of the signals > were used to fill the portfolio. Simply because you have this > portfolio constraint. If all signals would give you the same > Winner/Looser characteristics than there should be no problem. But > that's not true. Therefore the portfolio filling proces for me has a > random character and the result is based upon luck ... Unless I am > missing something ... And that's my question. > > > > Regards, Ton. > > > > > > ----- Original Message ----- > > From: ed2000nl > > To: [email protected] > > Sent: Monday, August 27, 2007 12:58 PM > > Subject: [amibroker] Re: How do I backtest placing a restricted > number of limit orders each night? > > > > > > hi Ton, > > > > I'm not sure if I understand what you mean. There are often more > > signals then you can use but the backtester is instructed to pick > the > > best signals using PositionScore. I can exactly perform my > backtest in > > the real world, excluding the shorts I am not allowed to enter by > my > > broker. The signals the backtester chooses are not pure luck but > > chosen using positionscore. But I guess I do not understand your > question, > > > > rgds, Ed > > > > --- In [email protected], "Ton Sieverding" > > <ton.sieverding@> wrote: > > > > > > Morning Ed, > > > > > > My problem when using the Backtester and in general a Backtester > > based upon portfolio result is the fact that in the real world an > > investor will have a portfolio with > > > let's day 20 stocks. Therefore when the portfolio has been > filled, > > all other BUY signals > > > will be lost until you've a SELL signal. For this reason when > doing > > a Backtest > > > I always do an Explore analysis of all signals. In general what > I > > get is something like > > > 200 Transactions from the Backtester and 1.000 Transactions > from the > > Explore analysis. > > > What makes things worse, I often get a RAR from the backtest of > > let's say 25% with > > > 75% of the signals being winners. When looking to the Explore > > analysis of all the > > > signals I only get something like 35% of winners. Therefore the > > result coming from > > > the Backtester must be pure luck. The backtester 'randomly' > chooses > > the signals to > > > fill the portfolio. I have no idea how to solve this problem ... > > > > > > Regards, Ton. > > > > > > > > > > > > ----- Original Message ----- > > > From: Edward Pottasch > > > To: [email protected] > > > Sent: Sunday, August 26, 2007 8:45 PM > > > Subject: Re: [amibroker] How do I backtest placing a restricted > > number of limit orders each night? > > > > > > > > > > > > hi, > > > > > > the way you set it up it shoudl not be possible. However, what > can > > happen is that the backtester finds exits for the next day and > > immediatelly fills them with new positions. So you need to make > sure > > that you first exit your positions and tell the backtester to > enter > > only on the next bar. This is usually the problem. There are > several > > ways to achieve this. Maybe you will get a more satisfactory > result > > when you set settradedelays(1,1,1,1). > > > > > > I use setttradedelays(0,0,0,0) but I make sure that the trade is > > entered 1 bar after the signal (same with the exits), > > > > > > Ed > > > > > > > > > > > > > > > ----- Original Message ----- > > > From: Michael White > > > To: [email protected] > > > Sent: Friday, August 24, 2007 11:37 AM > > > Subject: [amibroker] How do I backtest placing a restricted > > number of limit orders each night? > > > > > > > > > Can anyone help me model the following scenario? > > > > > > - Assume a portfolio is allowed to consist of some fixed number > > > of "slots" with equity equally divided among them (e.g. 10 > slots at > > > 10% of equity). > > > - Check for setup criteria at close of each day. > > > - Place next day limit buy orders for as many unfilled slots as > are > > > currently available (e.g. if already have 2 fills after day 1, > then > > > there are only 10 - 2 = 8 slots remaining for day 2, etc.). > > > - Buy orders are prioritized by a calculated value. > > > > > > My problem is that if I receive a setup for more symbols than I > > have > > > available slots (e.g. receive 20 setups but only have 8 > available > > > slots), my script will try to fill all 8 slots from the 20 > > > candidates, and the portfolio manager will correctly prevent me > > from > > > having more positions than allowed (e.g. no more than 10). > > > > > > However, in reality, I will only have placed as many limit > > orders as > > > I have available slots (e.g. 8 limit orders when 8 available > slots, > > > not limit orders for all 20 candidates, since I only have funds > to > > > cover placing 8 orders). > > > > > > What is happening is that my script is filling orders that I > would > > > not have placed! I need a way to indicate that despite 20 > setups, > > > only 8 limit orders were placed. > > > > > > Following is some script snippets. > > > > > > /* > > > * Assume an initial purse and brokerage fees ($0.01/share) > > > */ > > > SetOption("InitialEquity", 50000); > > > SetOption("CommissionMode", 3); > > > SetOption("CommissionAmount", 0.01); > > > > > > /* > > > * Carry fixed number of positions, dividing 100% of Equity > between > > > * them (based on previous bar's closing). > > > */ > > > PositionSize = -100/10; // Each position is 10% of equity > > > > > > SetOption("MaxOpenPositions", 10); // No more than 10 positions > > > SetOption("UsePrevBarEquityForPosSizing", True); > > > > > > /* > > > * We recognize the sale signal at the close of a bar and > execute the > > > * sale at the open of the next one, delay sale by 1 day. > > > */ > > > SetTradeDelays(0, 1, 0, 0); > > > > > > /* > > > * Trigger a Buy signal when previous bar meets the setup > > > * requirements AND this bar's Low has dropped to less than a > fixed > > > * percentage below the previous bar's close. This emulates > having > > > * placed a limit order the night before after having seen the > signal > > > * on that day's close. > > > */ > > > setup = ... // Some position entry logic. > > > PositionScore = ... // Some prioritization logic. > > > > > > BuyPrice = Ref(Close, -1) * 0.95; > > > Buy = Ref(setup, -1) AND Low <= BuyPrice; // Problem here!!! > > > > > > Sell = ... // Some sell logic. > > > > > > As indicated in my earlier comments. The problem is that in > > reality I > > > will not actually have placed orders for all candidates, but > rather > > > only for as many as there are available slots (e.g. 8). However, > > the > > > script will attempt to fill the available slots based on all > > > candidates (e.g. 20). > > > > > > How can I restrict the Buy assignment to only apply to the top X > > of Y > > > candidates based on priority (e.g. top 8 of 20 in example > above). > > > > > > Thanks in advance. > > > > > >
