people with nt,
i have been running a lot of optimizations in nt and have also been trying to trade automated on simulator over the last months. this has led me to discover some issues with the platform over this period that i will report in this forum as soon as i'm able to.
one of the biggest issues has to do with slippage. i have already created two threads to share my findings in these fora:
the conclusions from all the optimization processes i have run are:
- the nt platform already has all the information it needs to provide a column in every performance report for the observed, real world slippage when trading automated with real funds. it knows which signal generated any particular entry - exit order and what the price was when that signal was generated. the platform also knows the price that an order was executed - filled, so it can easily calculate the observed slippage by subtracting the price at which a signal was generated minus the price at which an order was executed.
- when one is setting the value for slippage in a strategy analyzer window one must define it as a number of ticks. however, slippage is measured not in ticks but in points in the optimization results. and even when one can change the unit of measurement for the results between currency, points, ticks, etc, the values for slippage and commissions are not changed in accordance. it should be the easiest thing for the people with nt to make adjustments to the platform so that the values for both slippage and commissions are reported in the same units as the profit and loss results in a strategy analyzer window.
also related to this point is the fact that nt's help documentation doesn't mention the units that must be used in strategy analyzer and chart windows. if one intends to use non minute bars then size must be defined in ticks. slippage is also defined in ticks. commissions must be defined in currency in a very different part of the nt platform. it would be helpful if this was explained in nt's documentation, i had to figure this out all by myself.
- when running backtests or optimizations in equities the value for slippage makes no sense whatsoever. it seems like the platform does apply slippage correctly when it makes the calculations required in a backtest or optimization but the format in which the platform reports slippage when working with equities does not correspond to currency, points nor ticks. it makes no sense. nt should look at this and fix it as soon as possible.
- and the most important request i have for nt right now concerning slippage is for the nt platform to have two different methods to apply slippage to backtests and optimization processes. at the moment the platform takes the value defined by the user but only applies as much slippage to any entry or exit as allowed by the historical data. there is a lot of merit to this method, but when one is using non minute bars then it is not so brilliant. in my case, i want slippage to be applied uniformly to all entries and exits. this is because i know that slippage is very significant and it always goes entirely against retail traders, and also because i use optimizations in equities to approximate how positions in options (calls and puts) would have performed historically. from extensive automated trading of options i have done on simulator i know that there is a very significant under-performance to positions in options when compared to equivalent positions in equities and i use the value for slippage to incorporate this under-performance into my evaluations. i use quite high values for slippage to this end and if nt doesn't apply these values uniformly then the results i get are not very useful. nt's strategy analyzer windows are already full with options and inputs, i would like nt to create a drop-down menu with two alternatives right after the box where one defines the value for slippage with two options: 1) apply slippage uniformly to all executions, and the current setting 2) never let slippage be larger than the high - low of the following bar.
very well, i will report on the other issues i have found as soon as i have a chance. thanks, regards.
Comment