Peter, I think it would be best if you post the code for us.
Preston --- In [email protected], Peter Fertig <pfertig@...> wrote: > > Hello, > > I have a problem with the system tester. I want to test a smoothed indicator. > The logic of the system is to buy if this indicator is higher than at the > previous bar. Accordingly, a sell is allowed if the indicator is lower than > at the previous bar. The buy order should be executed at the next bar but > only if the high of the current bar is exceeded. Similar, a sell order should > be executed only if the low of the next bar is lower than the low of the > current bar. The system would be always in the market. Thus, orders should be > entered on a stop and I have put in the Limit or Stop Price field High for > the orders to buy and Low for the orders to sell. > > The problem is now the accuracy of the system test reports. I have tested the > system also with TradeStation and with Excel. In the case that the order > could be executed at the next bar after the indicator turned direction, the > results of the system tester in MetaStock are the same as those from > TradeStation or Excel. However, if the condition for the order is still in > place but the order could be executed only a few bars later, the results > diverge significantly. The system tester uses not the high or low price of > the current bar, but of some bars ago. Even the number of bars ago is not > constant but differs from trade to trade. > > The manual does not provide any help. Are those incorrect order executions a > bug of the system tester or do I have to put in another value in the Limit or > Stop Price field for the various order types? Any hint is highly appreciated. > > Best regards > > Peter >
