Trading Plan and Backtesting, which one first?

Discussion in 'Professional Trading' started by wuming79, Feb 25, 2010.

  1. wuming79


    Hi Guys, I'm separate my question here from the Business plan thread. Hope to see some views on this if it is my own thread.

    I'm always wonder which one comes first. Trading Plan or backtesting? If we write a trade plan based on a strategy that is not proven profitable yet, do we still write that plan? If we wrote the plan before back test, when do we start to alter the plan if we feel a particular plan is not up to expectation? Anybody has this dilemma before?
  2. 1) Do backtesting and "research" first in order to get an idea of what you feel comfortable with. Then it can be easier to develop a trading plan that is better focused and less likely to have to be altered.
    2) Have the expectation that you will need to learn "everything about everything" and then you'll have to unlearn certain things that have no value. :cool:
  3. Backtesting, I think, then your Trading plan.

    Like what was mentioned above, research and backtesting first before setting out a plan. That way, you'll be able to create a trading plan, suitable to you and your personality.
  4. I disagree, you can spend your life backtesting. Its dangerous if you forget that you backtest because trading is your business, you dont start a business because you backtest. It may sound trivial, but its a function of how you spend your time and efforts. Backtesting can be a real lag on setting up a proper trading business.

    You need to structure everything, especially the consumption of your time backtesting as a business..

    Of course you need a basic background to understand that backtesting is necessary and time intensive, but theres a greater need to learn that trading is a business first, before losing yourself in a myriad of trading language, environments, and strategies.
  5. Redneck



    1.) Learn what you need to – to understand how the market works and what you’ll need to do to extract money from same (like NAZ said at some point you’ll need to unlearn some of crap because it will be useless – btw none of us can tell you what will be, and what will not be useful/ useless - we are not you)

    2.) Develop a thesis on how to extract money from the market based on above – this is your trading plan

    3.) Write it down in painstaking detail, and in an exact step by step procedure

    4.) Be sure to add in a healthy dose of risk management – if you don’t fully understand risk management – learn about it first – it will literally save your butt ..

    5.) Back test your plan, or forward test it on Sim
    WARNING – there is a very fine line between back/ forward testing, and curve fitting your plan… vs. back/ forward testing, then modifying your plan…

    If you cure fit it – you’re dead.

    If on the other hand you identify gaps, omissions, errors and modify accordingly – fine… Just be mindful nothing is 100% with regard to trading - neither will your plan to extract money – which is why you must learn to be a trader, and utilize #4 above

    6.) Forward test on sim (optional of course but cost effective)

    7.) Like it = trade it live – verbatim…, Don’t like it = revise your plan till your satisfied – again remember it/ you will not be 100% successful with every trade which is why #4 above is vital…

    8.) Never deviate from your plan – that is a recipe for disaster, never curve fit – again that is another recipe for disaster

    9.) Wash, rinse, repeat

    Easy as pie

  6. This is really good advice. One of the smartest guys I've ever met said, "measure what is important, because that's what gets optimized". He was talking about the situation where corporate management measures each employee's hours closely and they end up with a lot of people really good at punching the clock. Usually just a little effort on management's behalf to figure out how to measure real productivity for their department breaks the spell and everybody gets a lot happier and more productive.

    As an independent trader, you are also the manager, and your productivity is your dollars pulled out of the market. Eye on the ball!
  7. wuming79


    Hi, ok, let's say I wrote a plan now but the strategy is not being tested yet. It's sort of proven that a group of people are making money with it but when I try it, it's not as easy as it seems. Something seems to be missing. I want to change something in the plan to find out the results but getting the data can takes weeks or months if I do swing trades. When does one start to alter the plan? Or do I really stick to the plan and trade it until I get maybe 100 trades before I start analysing and do alterations?

    Regarding backtesting or forward testing. I think backtesting at most can test particular indicators or patterns right? How does one take into consideration of the general making movement at the time you put up a trade? I think that is not achievable unless you forward test it. but forward testing a strategy that is not proven might drill your mind to act in the wrong way or instill bad habits in yourself. I'm currently doing some forward testing and I found out I had a lot of bad habits to change. I'm worried after I change the plan, I add more bad analysis into it. Is there a number of trades to clock so that we can really say "this plan is not working" and we can change it before we instill the bad habits in ourselves?