[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Risk and TS Maximum Drawdown or Updated maximum Drawdown



PureBytes Links

Trading Reference Links

 Hi Aaron,
Hi to all,
Thank you for your reply,
I've found that to do it  is not as simple as taking TS perf summary 
maxDD and update it with actual price.
Because when Tradestation calculate and point the maxIDD it does not pay 
attention to the relative level of prices and do not take care of % values.
Consequence is that  it will most of the time indicate a wrong maximum 
drawdown level and date. So we cannot simply convert TS MaxIDD into 
updated value. We do have to Calculate it from scatch. And the same for 
nearly all fields of perf summary if we want to see what could happend 
now if we apply the system in real trading.

Let's take an example for Drawdown : say that the actual price is now  
3500. Baktest a system from 1990 to 2003 , big pointvalue = 25
There are # periods of drawdown
1/    One in 1991 for  $10.000   at   price level 2000  ( = 440 points 22%.
2/    One in 1997 for  $13 000   at   price level 3000 ( = 520 points 20%)
3/    One in  2000 for $ 21 000 at price level  7000   ( = 840 pts - 12%)

In this backtest TS will reatin MAXIDD of $ 21 000 on year 2000  and we 
may focus on that period to analyse it further.

But if actual price is 3500 and we updated the DD calculations 
according to the relative price levels, we obtain  updated MaxDD of :
1/    3400 *22% *25 = $ 18 700
2/    3400 *20% *25 = $17 000
3/    3400 *12% *25 = $ 10 200

And so the system DD in actual price is not a  of $ 21 000 Max DD in 
year 2000 , but a $ 18 700 Max DD in 1991with -22%
Same logic apply with Net profit , average trade win/loss ..... if we 
take care of % relative moves rather than of  number of points.
Looking at TS perf summary report conduct us to wrong values and 
possible wrong backtest conclusions.
And have huge possible effects  on  account size required, Risk tio 
reward ratio, Net profit , average trade and so on...
That's why i am looking for somebody who would have already done an 
updated performance summary in a function calculation or signal to 
include in a system.

If  i am wrong sowhere please point it to me .

May be i am very knew to this but i wonder how experienced successfull 
traders are doing with this problem and how wrong are those who uses 
only Tradestation performance suummary report to made their backtesting 
decisions.
I will be delighted if somebody with system testing good experience tell 
me how he deals with this subject.

AS i am not at ease with programming I am looking for a piece of code to 
re-calculate all fields of performance summary in a percent updated values.

Thank's for your attention.

Best regards
Philippe



Aaron Schindler a écrit:

Hi Philippe,
 
Well, if you want to stick with one contract, then you can also scale 
your drawdown by dividing by the price to get a percentage drawdown.  
This will allow you to compare drawdowns regardless of whether the 
price is 7000 or 3000.  Otherwise, as you say, a 700 point drawdown is 
a lot more severe at 3000 than at 7000.  But a 10% drawdown will be 
just as bad in all cases.
 
Good luck,
Aaron

    ----- Original Message -----
    From: Synergy <mailto:Synergy@xxxxxxxxxxxxxxxx>
    To: Aaron Schindler <mailto:aaron@xxxxxxxxxxxxxxxxxxxx> ;
    omega-list@xxxxxxxxxx <mailto:omega-list@xxxxxxxxxx>
    Sent: Thursday, November 06, 2003 3:59 PM
    Subject: Re: Risk and TS Maximum Drawdown or Updated maximum Drawdown

    Thank you  Aaron,
    But this do not answer my question or I did not understand well
    your answer:
    Backtesting a system begin for me with One contract, without any
    position sizing model.
    I do not want to choose and put position sizing on a system before
    having evaluated it correctly among others systems for one contract.
    And to do it I need to have a good backtesting system report with
    at least an updated maximum drawdown for ONE single contract.
    I do not see how to do it differently.
    So is the reason of my question for Updated Max DD
    Or am I wrong somewhere ?
    Thank's
    Regards
    Philippe




    Aaron Schindler a écrit:

Well, I can help you with one of your questions:

 

What about a 15 000$ DD if This DD may have been reached when the market
was at a price of 7.000$ whereas price is now at the 3 000 level. Or the
inverse?
   


As another poster recently suggested -- take a position size (number of
contracts or shares) inversely proportional to the current price level.
That way it will take the same % change in the underlying price to give an
equal dollar drawdown and your drawdowns will be independent of the price
level.

Aaron



----- Original Message ----- 
From: "Synergy" <Synergy@xxxxxxxxxxxxxxxx>
To: <omega-list@xxxxxxxxxx>
Sent: Thursday, November 06, 2003 2:24 PM
Subject: Risk and TS Maximum Drawdown or Updated maximum Drawdown


 

Hello to all,
I would like to know the opinion of the experienced system traders on
the following question: :
Did you found it is pertinent or not pertinent to take care of Updated
Maximum Drawdown compared with TS Maximum Drawdown ?
What do you do with the Max DD shown by TS System report as it is not
actual price updated ?
What about the risk to reward ratio of the system ? what about the
Minimum capital required to trade the system? &
I wonder why Tradestation does not shows both Max Instantaneous IDD and
MAX Updated IDD ?
Dont you calculate the present equivalent value of the max DD encounter
by the system when you backtest ?
What would happened to the one who begin to trade a system with 10 000 $
Max DD if this DD was hit when prices was around 2000 and are now around
8000 ??
I do not see any recommendations on this nowhere.
Why ?
Does somebody have coded a function to include on a system to calculate
the max DD on a updated bar to bar basis ?
Or could somebody do it for me ?
Thanks to share your experience
Regards
Philippe