Money management with multiple markets

[Futures Guy]

Hi,

I have a question about how to allocate funds if you are trading
several markets. The way I see it, there are two basic approaches:

1. Commingled Funds. Risk a certain percentage of all available funds
on each market.

2. Segregated Funds. Treat returns from each market as if they were in
a separate account. For each market, risk a certain percentage of
funds available for that market.

For example, say there is a very profitable trade in market A. Then,
under the first approach, on my next trade in market B, you risk a lot
more in dollar terms. However, under second approach, your dollar risk
for market B does not change. Only the dollar risk for market A
changes. (I hope this is clear.)

My question is: which of the two approaches is better? Are there any
books / articles on this? Which of the two approaches do you use? Why?

IMO: If you are using a trading system that performs equally well in
all markets, then you should go with Comingled Funds. If you think
your system works better in some markets and worse in others, then you
should use Segregated Funds.

The question then becomes: If you are using a technical system that
doesn't care what the market is, is it plausible that, long term, it
performs better in some markets than in others.

Any ideas?

TIA

[Futures Guy]

(Chet Hayes) wrote in message >...

> "For example, say there is a very profitable trade in market A. Then,
> under the first approach, on my next trade in market B, you risk a lot
> more in dollar terms."
>
> In my opinion, any money mgt strategy where you risk a lot more after
> a very profitable trade is doomed to failure. Money mgt should be
> based on risking a small percentage of total capital on any given

That was just an example meant to describe the two different
approaches to money management. One approach is to risk a (small)
percentage of *all* the money available. Another approach is the
pretend that you have several different accounts, and that you trade
one market per each account.

My question was, Which of these two approaches is better? I understand
that in either case you should be risking a small percent of what you
have.

> " The question then becomes: If you are using a technical system that
> > doesn't care what the market is, is it plausible that, long term, it
> > performs better in some markets than in others."
>
> I'm not sure exactly what you mean by a technical system that doesn't
> care what the market is. There certainly are technical systems that
> work well in certain markets and poorly in others. That's why many of
> the systems you see being marketed are targeted to specific markets
> like bonds, stock indices, etc. Just about any system will have
> varying results depending on the markets traded.

Right. You are talking about "systems that we see being marketed". The
question is, if a system, in the **past**, performed well in one
market and not in another, do you think it will perform well in the
future, in any market? For any system, it's always possible to find a
stretch of time in the past when it worked well. It doesn't mean that
this is a good system.

[Chet Hayes]

"My question was, Which of these two approaches is better? "

If you have two accounts trading different markets and/or systems, I
would treat them seperately from a money mgt standpoint. By doing
that, if one account strategy is doing well why the other is losing,
you will be increasing the amount per trade in the winning account,
decreasing it in the losing account. That will lessen the chance of
wiping out.

"The question is, if a system, in the **past**, performed well in one
> market and not in another, do you think it will perform well in the
> future, in any market? "

If the system was tested in a particular market over a reasonable
timeframe, including varying market conditions, then that is the best
indication you can have that it will continue to work in that market.


(Futures Guy) wrote in message >...
> (Chet Hayes) wrote in message >...
>
> > "For example, say there is a very profitable trade in market A. Then,
> > under the first approach, on my next trade in market B, you risk a lot
> > more in dollar terms."
> >
> > In my opinion, any money mgt strategy where you risk a lot more after
> > a very profitable trade is doomed to failure. Money mgt should be
> > based on risking a small percentage of total capital on any given
>
> That was just an example meant to describe the two different
> approaches to money management. One approach is to risk a (small)
> percentage of *all* the money available. Another approach is the
> pretend that you have several different accounts, and that you trade
> one market per each account.
>
> My question was, Which of these two approaches is better? I understand
> that in either case you should be risking a small percent of what you
> have.
>
> > " The question then becomes: If you are using a technical system that
> > > doesn't care what the market is, is it plausible that, long term, it
> > > performs better in some markets than in others."
> >
> > I'm not sure exactly what you mean by a technical system that doesn't
> > care what the market is. There certainly are technical systems that
> > work well in certain markets and poorly in others. That's why many of
> > the systems you see being marketed are targeted to specific markets
> > like bonds, stock indices, etc. Just about any system will have
> > varying results depending on the markets traded.
>
> Right. You are talking about "systems that we see being marketed". The
> question is, if a system, in the **past**, performed well in one
> market and not in another, do you think it will perform well in the
> future, in any market? For any system, it's always possible to find a
> stretch of time in the past when it worked well. It doesn't mean that
> this is a good system.