|
PureBytes Links
Trading Reference Links
|
Leslie Walko wrote:
> Both Trading Recipes and Portfolio Stream are based on the
> erroneous assumption that you can add the output of two or more
> revenue generating processes together to model a portfolio. Not
> valid! If this assumption were valid, every mutual fund would
> have outperformed all other forms of investing.
Hi,
You raise a very interesting point, for practically everyone here. Even
people who trade just one market (e.g. S&P) usually do so via several
systems. And ofcourse to those who are trading multiple markets.
I haven't used recent RINA products like PS, but I use TR.
I think in the case of MCS, you're talking about a different aspect of
the whole problem of correctly simulating the trades of a portfolio
system. What I THINK you're saying, is "Portfolio MCS" allows one to
assess the possibility of different outcomes when trading a portfolio.
Can you please clarify?
In my experience, TR correctly simulates the way you want to trade a
system on multiple markets, whereas other products didn't (as of 2001
which was the last time I looked). Like being able to equalise exposure
across different market and not e.g. just simulated 1-lots in every
market (1-contract in corn has a very different $-risk than nat.gas),
manage risk (correlated markets risk, long/short risk, total heat etc).
TR has a feature called "Worst Case Analysis" which re-runs the
portfolio using every day as STARTDATE, to simulate how history would
have been, if you had started trading on a different date. I guess one
can also run an MCS of the daily returns from TR's text output (no MCS
function included in TR, but everything can be saved in plain delimited
text files)
What exactly does "Portfolio MCS" do, which can't be done ot is done
incorrectly with TR (other than the obviously important issue that it
works directly with TS)?
Regards, M
|