CFD Backtesting / wealth-lab?

Discussion in 'Automated Trading' started by akardar1, Feb 14, 2009.

  1. akardar1

    akardar1

    --------------------------------------------------------------------------------

    Hello,

    Do we have any members who have backtested strategies for the CFD market?

    i.e - lets assume an individual wanted to to gain exposure to mid/long term
    trendfollowing across a basket of diversified products: indices, metals, agri, interest rates and FX. He cant enter the futures market due to smaller capital, and therefore
    looks at the CFD makert as way to trade the strategy through "downscaling".
    Instead of trading 2 crude futures with £1 million, i ll trade 8 crude CFDs with £100k for instance.

    Do we have any members that have backtested strategies using proffesional backtesting software such as Wealth-Lab / Trading Blox etc..?
    I am assuming the process would involve uploading a database of CFD margins
    and trading costs etc .. for each asset, though a CFD broker's dealing guide etc.

    Does any1 have experience in this field?

    Any thoughts appreciated

    Dre
     
  2. I don't think anyone has every done backtesting with CFDs, because of the lack of historical data.
     
  3. akardar1

    akardar1

    Sure, but one could:

    ...Do the backtesting on historical futures data to establish that the stratery has some positive results?
    But using CFD margins requirements and CFD tick sizes/values
    instead of the futures contract specifications.

    Does this sound viable ?

    Once this is done, signals would be given from futures end of day data, but executed through CFDs..
     
  4. pneuma

    pneuma

    I imagine that you could, though i haven't with those particular instruments except FX (using Spot not futures).

    As i am not familiar with your situation, are you suggesting that your margin and trading costs will change with time? I would have thought that these things would have been fairly stable?

    I think you should look at writing a simulscript to help with your backtesting results.

    pneuma
     
  5. Only in WL, but you'll most likely have the time problem against you, because you'll have to be doing at least 10 trades a month to keep it in the US. If you think you can do it quickly and have at least 25k....it's possible.
     
  6. If you need to trade oil using CFD's, I have a simple message for you...

    give your money to charity or visit a casino.

    At least you'll have a bit of fun that way.

    Why do you think they don't charge you commish in many cases?
    Because they hedge your dumb trades and pass all costs to you via the spread.

    How long would you last in business buying at retail and selling at wholesale prices?

    Pretty sad people are foolish enough to even contemplate this shit.
     
  7. One of the major difference I find with CFDs and actual future contract is the spread, usually with an activate contract like oil futures spread are within 1 tick. Depending on the CFD brokerage you use, there can usually be 5 to 10 tick spread.

    I guess FX traders who have system that take into the account of spread is more suited for CFDs.
     
  8. Most FX "brokers" are bucketshops.
    They have many ways of skinning the average retail 'trader'.



     
  9. akardar1

    akardar1

    Sure, I understand that retail spreads are wider,
    however, if you catch a trend and are in it for the longer term,
    its not a few ticks @ entry that can completely diminish your returns.... I am not talking about intraday trading.... provided your backtested results incorporate the extra spreads... with an extra margin of safety, and still shows positive expectancy.. there is no reason why the CFD market cannot be used an substitute to the futures markets despite higher costs.
    Just my opinion.

    P.S: I already give something to charity and i dont do casinos, iam not a gambler - thanks for input though