|
PureBytes Links
Trading Reference Links
|
Brian Massey wrote:
>
> Many textbooks advocate never risking more than around 2% of your equity on
> any one trade. This means that with, say a $20K account in the futures
> markets, you can only risk $400-$600 on one trade. Let's say you're
> trading Bonds, then you can only buy 1 contract and risk roughly 16 ticks.
> This is a very conservative way to trade.
>
> This also means that you have $17K unused funds. To maximize the utility
> of all the money invested with our broker, but still adhere to the 2% rule,
> 8-10 markets must be traded simultaneously. This obviously isn't practical
> yet what the 2% rules implies. Some would call it overtrading.
>
>
If you are going to use a fixed portfolio (the commodities you run your
system on and trade), then you also must allow for the drawdowns that
particular portfolio will endure. Many suggest that a port. should be
picked so that, if you were in every position in the port at the same
time, you only have committed 25 - 50% of your funds. This is
consistent with some backtesting results I have seen for some
commercially available systems. Most of these systems allowed losses of
up to $2000 so except for the largest accounts, the results would be for
risking more than 2%. If you can risk less, then a larger portfolio.
But continuing with the published results: The systems have very large
drawdowns for a particular commodity (5-15x margin!) Small portfolios
may have drawdowns of 3x total margin so that you should only commit
25%. Large portfolios (20 or 30 comm.) have 1x drawdowns so you could
commit 50%. But you have to keep in mind this is backtesting which is
likely to be more favorable than the future because it's optimized more
or less (see below).
When i first started trading 2 years ago, using a commercially available
technical system whose signals were sent by fax, I committed to a port.
representing about 70% of my funds in margin. This was a big error.
The account drew down 75% before i abandoned the system. Then that was
when the wins came along in the particular commodities I was trading.
The real time track record of this system makes some gross
approximations, but it was otherwise profitable over the course of the
whole year, but I was not still with the system, and in fact could not
have continued with my original portfolio if I had wanted too.
In a second phase of trading I used my own methods but again picked
about 70 or 80% committment. I decided to use an evolving portfolio
where as the account grew or shrunk I would add or subtract certain
commodities. Well, the acct. shrunk. And again it seemed as I dropped
a given commodity, that was when a win considerably bigger than my
losses came along. While this might be an approach where you stay in
the game, it makes for big equity swings and the dips are going to last
a long time because you will be trading few commodities after the
drawdown.
In all cases I was risking about 5% per trade. This might be high but
I had a fairly small account. I suspect the bigger mistake was
following too many commodities, rather than the bigger risk (5% vs. 2%).
The conclusion may be that you only should select a portfolio such that
the total margin in only about 30% of your account balance if you are
trading a moderate number of commodities. (If trading one commodity
only, then the committment probably should be even less.) But it gets
worse.....
John Hill of Futures Truth said in an article that most systems drawdown
realtime about twice what they do in testing. He then suggested a
cutoff point of 50% more than that for stopping a system. Taking those
points in account, you would only trade a portfolio with margin of 12.5%
of your funds for a system whose testing allowed 30% committment! That
may be overly conservative as it allows for a drawdown of 3x the worst
backtested value.
If your system has lower drawdowns than the ones mentioned above then
you could modify this rule. Or if you were willing/able to add funds to
get through drawdowns.
But I think the general point is that you have to commit much less of
your funds than the average trader would probably pick if you are
trading a mechanical system. I know when I started with the commercial
fax service, I did not understand the size of drawdowns that systems can
produce. I had only their brochure that said the worst overall drawdown
in 30+ years of realtime results was 40%) But that was for every
commodity!! Not all vendors show detailed results for the backtesting.
And even those that do, how did they pick their moderate size
portfolios? If they picked among many choices based on the lowest
drawdown, then this is another point of optimization that has entered.
Conrad Bowers
|