|
PureBytes Links
Trading Reference Links
|
gwen ecrit:
> Further questions:
> - Does the old signal reappear if you load data up until yesterday only? -
> How does it behave with Ascii data up until yesterday, until today?
dont know and dont care - I want to have an explantation of WHAT has
happened in front of my eyes - not what else can go wrong - I want to SOLVE
a problem not create others or more ....... I have enough :-))
> Maybe it's a data problem with unshown Globex prices or something like
> that.
cant - there is only 1 data stream - and its DBC´s S&P symbol - thats ONLY
daysession
Probably have to wait for OM..........come on TELL ME !!!
rgds hans
> Gwenn
>
>
>
> | -----Ursprungliche Nachricht-----
> | Von: tradejacker@xxxxxxxxxxxx [SMTP:tradejacker@xxxxxxxxxxxx]
> | Gesendet am: Wednesday, October 20, 1999 3:03 PM
> | An: omega-list@xxxxxxxxxx
> | Betreff: Re: whats this FEATURE ?
> |
> | some questions....
> | 1. if it is a bought system, what's its name? someone on the list may
> have | had similar problems. | 2. what is the system logic? it sounds like
> it might be a recursive | calculation | causing the problem if a prior
> day's bar had dropped off when the new bar | is set to be added after
> midnight. | 3. could it be a mmb problem? | 4. what happens if you use
> intraday data (set at 405 min) for daily bars | rather than daily data? |
> 5. when all else fails, have you contacted the system developer to report
> | the problem? is the system white box, gray box, or black box? | | TJ | |
> with troubleshooting hat on | | At Wed, 20 Oct 1999 16:30:45 +0200, "hans
> esser" <he96@xxxxxxxxxxxxxx> wrote: | | >Its not a SECRET system, its a
> BOUGHT one, therefore I cant provide | >any | >code as its not allowed.
> ALL I can say: | > | >+ It went LONG during the DAY on S&P on a DAILY BAR
> chart | > | >+ it uses the MONEY MANAGEMENT STOP - didnt YOU say NEVER to
> use | >that, but to use self coded ones ? - or was that one of those damn
> bashers | >here ? | > | >+ that was set to 16 !!!.... thats 16 big points,
> my symbol universe | >for S&P is | >set to 1 - so I have no problem with
> backtesting OVER the period where | >the | >contract multiplier changed
> from 500 to 250. THIS way has worked million | >times correct before. | >
> | >+ the market went down MORE than 16 bps - the system EXITED on the |
> >money management stop ! (sounds good to me - sounds even better as I |
> >didnt do the trade <g>) |
|