Experience-sharing: Going through some tedious work trying to get my head around various currency futures contracts, correlation with spot products and making sure I am using all the data correctly in back-testing and optimizations.
Comments on TradeStation’s ‘Portfolio Maestro’ product add-in.
TradeStation – ‘Portfolio Maestro’
Subscribed to TradeStation Maestro product last week. Maestro allows users to back-test multiple strategies over multiple markets/instruments simultaneously. It also allows optimization in the same manner. This is a big time saving compared to trying to back-test and optimize the same trading strategy over say 20 markets. Thus it is particularly helpful to traders looking to test strategies on a large number of instruments – say individual equities – of which there may be hundreds or thousands.
The product is less relevant if working on optimizing a trading strategy for a single market such as one index, one currency, one fixed income product etc.
Maestro costs US$60/month but this fee is waived if doing ten round trips with futures contracts.
Maestro and FX
Maestro does not work on FX instruments if looking at intraday timeframes. Let me clarify what I mean by ‘FX’ – am specifically referring to spot FX. Maestro works fine on intraday timeframes for all futures contracts, including currency futures. This becomes helpful if able to work with continuous futures contracts.
There are a bunch of useful currency futures contracts traded on CME with decent volumes such as NE1 (New Zealand dollar), JY (Japanese Yen), SF (Swiss Franc), CD (Canadian Dollar), DX (US Dollar Index), EC (Euro), MP1 (Mexican Peso) and BP (the British Pound).
However, other than those, I could not find currency futures with any decent volumes for other crossrates such as AUDNZD, EURJPY or GBPCHF.
Ideally I want to run my testings over 30-40 currency pairs – I will need to eventually concentrate on the spot FX market (as opposed to the futures market). Hence I will need to look at a solution such as MultiCharts to increase the testing sample sizes and to implement the automation of strategies.
Additionally, trading in the spot market will allow me to position size more accurately as opposed to trading in the currency futures. In the latter, I would have to trade in whole/standard 100k lots as opposed to mini-lots (10k) in the spot market.
Rollover calculations for continuous futures contracts
One issue I had encountered with futures contracts is rollover calculation methodology distorting the pricing of the contracts. Thus when I looked at the continuous CL contract (the crude oil future), the monthly chart shows CL topping out at $200/barrel some years ago. Price never traded at that level. The all-time peak for CL was around $160. A colleague, whom I was introduced to recently (thanks to FXTraderPaul), showed me how I can adjust the rollover calculations. This was a nice easy-win. Using his guidance, I was able to tell TradeStation to calculate the rollover in such a way that the all-time CL peak does indeed show around $160, even when using the continuous CL contract. Awesome!
I ran backtests over the same time periods with both the currency futures and the corresponding FX spot instrument to make sure that there was a strong performance correlation between the products. This was tedious work and required some investigative reconciliation work – thankfully I have an accounting/number-crunching background – begin/end dates, bar closing times, equity curve shapes, number of trades, slippage/commissions and strategy parameter settings to name a few.
This breakthrough allowed me to go ahead with the testing in Maestro for 9 of the FX markets.
Regardless I will need to move over to MultiCharts in the very near future and/or use both TradeStation and MultiCharts simultaneously.
I have not come across discussion of the above topics much at all. Either I don’t know where to look for these types of topics, or I am going to deep, or other people can’t be bothered.
Anyways, cracking on. Happy trading.