Switching strategies: best backtesting sw?

Discussion in 'Trading Software' started by cybercat56, May 28, 2003.

  1. Good morning,

    I am from Italy and have landed onto this nice forum
    Please, consider that I am a Tradestation User and I trade the Italian Stock Index Future (Fib) with a computerized intraday trading system. TS is a nice sw but can not do everything (for example can not test switching strategies among different securities).

    Lets assume that:

    - I had a basket of 50 stocks/indexes
    - I would like to test this strategy:

    Buy the 5 stocks/indices that:

    - are over their 25 days EMA and

    - whose 50 days EMA is upward sloping and

    - whose 21 days ROC is the highest among the 50
    stocks/indexes and

    - keep the position in each of the 5 selected securities
    until their ROC rankings remain among the first 10
    unless stop loss is executed and

    - if one security quits the selected portfolio, insert a new
    one

    - back test this strategy and give the equity curve of this
    switching strategy

    Could I do that with Wealth Lab? Could I do with other programs?

    I have to be sure that I have explained the real meaning of my question. The reason is that I am looking for an alternative to TradeStation (which, in any case, is good for some specific works) and, in addition to the cost, a wrong choice about my next program would be a huge waste of time. While I am writing, just now TradeStation is working in real time for me with an intraday system which has a very good risk control. I do not need to follow the market every minute. But I would like to move also to daily trading with a different philosophy.

    With Trade Station I can size the position but I can not switch among different securities, selecting the most promising (according to a backtested strategy). My interest is about monitoring simultaneously various (country and sector) ETFs to decide not only when buy one of them but WHICH ONE buy among the many available. Why? In statistics you call this principle "degrees of freedom", the more the better because you have more choices. Moreover, if you look at the charts of some real estate and gold ETFs, you understand that even in a bear market you can be long of something on condition that you monitor multiple uncorrelated markets.

    In 1998 I started an academical study on sector rotation strategies using Fidelity Select Sector Funds. I bought a US sw called Monocle II, I made a lot of analysis and at the end I printed a study on the Review of the Italian Financial Analysts (I am a member and editor of the quarterly magazine), now reprinted in a book (chapter freely available in Italian on your request). Monocle II was excellent for preliminary analysis but not for sophisticated strategies. It allowed the selection of the best funds on indicators based on relative strength (and related backtesting strategies) but did not allow multiple conditions (filters). So it could happen that it bought a fund in a bear trend just because it had bounced more strongly then others. The use of moving average as filters could have reduced this problems.


    Thanks for the reply,

    paolo_sassetti@yahoo.it
     
  2. maxpi

    maxpi

    I think you can use Radar Screen in TS2000 and monitor multiple securities and multiple strategies and autotrade with a third party interface to brokerage. I am working on that exact problem currently and so far I have discovered I have to buy (and master!!) Hashnums for $200.

    Max
     
  3. jimcrist

    jimcrist

    It would be nice if someone did a comparative review of Wealth-Lab versus Amibroker versus MetaStock versus TradeStation versus TC-Companion.

    Are there any comparative reviews out there?? For that matter are there any individual reviews of any of these programs??