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

Re[2]: Bund-futures tradingsystem



PureBytes Links

Trading Reference Links

Hello  >:-),

> Whoa!! Hold on a minute. What makes you think that TS 4.0 is mathematically
> acurrate?  :-)

I never thought anything but - TS 4 is more mathematically correct than
TS  5 that is for sure.  I have tested a few indicators on a C program
against  TS  4  and then against TS 5 and its good, bad - worst is the
rating  -  with TS 5 being the worst.  So I'm not defending TS 4 - but
if  you  had  to  choose  between one arm getting cut off or both arms
getting  cut  off what would you choose?  Especially when others have
clearly  demonstrated  what  it looks like having two arms cut off, yes
they are here every day displaying their stumps.

Mark


> TS 4.0 is still inaccurate due to its use of single precision, as is TS2ki.
> For example, you have to change the scaling on the Dow and NDX indicies, as
> well as their respective futures contract, just to make sure your data is
> not messed up now that the some of the data being sent out is over 6 digits.
> With the big drop in the NDX and @cco in April, if you don't change your
> scaling, the data will then be skewed due to TS's use of single precision.
> Gary Fritz went over this in a lot more details than I can remember.

> All I know is that both cannot be use for systematic trading, regardless of
> what the Omega advertisements keep saying. Anykind of intensive mathematical
> calculations in TS will result in errors, unless it is sent out to a DLL.

> As for me, I'm trying to use UMDS with Matlab as a stopgap until something
> better comes out. With Omega going off into the brokerage business, I'm sure
> we can count out a TS 6.0.



--
Best regards,
  Mark Brown   mailto:markbrown@xxxxxxxxxxxxx
  Y = Offset + Amplitude * sin(Frequency * X)