|
PureBytes Links
Trading Reference Links
|
Which Contract to Analyze?
A Friend just asked me what type of time series should he
consider to analyze and test his trading systems for US T-
Bonds.
He sent me some description of the type of calculated contracts
that he can organize from his CSI data.
Since he is a position trader and analyzes daily bars, I
recommended him to use the so-called “artificial High-Low”
format because settlement prices are real prices which have a
clear impact on tomorrows opening prices.
In my opinion settlement prices are not real prices for day
traders only.
On the type of calculated prices, I recommended that he used
price data that are the closest possible to what he can get in
actual trading so that his simulations could be as accurate as
possible to reality.
As Jack Schwager said in his interview to Trader Talk in Omega
Research Magazine (summer 1998, p. 24) if it doesn’t match
what’s happening in your account, it is not the correct time
series to be using.
The perpetual contract is the great choice, but it is not actually
what it is really traded.
In the case of T-Bonds due to the giant size of that market there
is good liquidity in the 2nd nearest contract. Besides, the 2nd
nearest does not the huge rollover gaps and the other problems
extant in the 1st nearest.
Therefore, for the case of T-Bonds I suggested to him that he
should use the 2nd nearest contract.
I would like to hear your opinions. Is my suggestion a correct
one?
Is the artificial high-low/2nd nearest a good time series to test a
trading system and position trade T-Bonds?
Thank you all in advance for your time!
Alfredo
|