Building forex trading systems is the reason that this company
exists. Nearly every one of our clients aspires to be a fully automated
trader.
The percentage of clients that succeed over the long run,
however, is incredibly small. I don't think that is a big secret. If
that is news to you, allow me to be clear. The chances of developing a
successful expert advisor over the long term are minimal.
The
inevitable next question from new clients is an attempt to pick my brain
for some features common to successful systems traders. I decided to
outline the traits common to this rare breed and some of the challenges
that they encounter.
Traits of a successful EA developer
Expert
advisor developers that succeed are either incredibly bright
(engineers, Ph.D. holders, etc) or dumber than a sack of rocks. No
middle ground exists. A friend/client told me about an experiment that
struck me as telling. If you put a mouse in a cage where cheese appears
40% of the time on one side of the box and 60% of the time on the other,
all mice eventually pick the side that "wins" 60% of the time. The mice
stop guessing and go with what works more often than not.
That's
my theory for why the "intellectually challenged" crowd seems
disproportionately likely to profit from trading. They see right through
the noise to pick out a simple observation that wins more than it
loses. Holy grail systems certainly do not come from this group. That's
not honestly not the goal. It's about eating cheese more often than
going hungry.
The brainy crowd benefits from a heavy dependence on
analysis. Ideas usually stem from pet theories, many of which look like
the crackpot variety on first glance. The theory shifts and twists as
market tests reveal strengths and weaknesses. Logical conclusions
instill a robustness and variety into the system. Interestingly, the
approach to testing and verifying the systems follows the opposite path
of most developers. The ideas tend to get simpler with time rather than
more complex.
Emotion does not play a large role in the rationale
for developing an expert advisor. Again, the reasons are usually
logical. Examples like "I've been trading this system for 20 years and I
need to automate it" or "I came up with a way to accurately predict
market direction."
It may sound a bit zen, but the systems that
make money are the ones that do not set their primary goals as making
money. Making money is a totally open ended goal. The potential
variables know no limit. The lack of constraint encourages would be EA
traders to shoot off on tangents. Unless the wild ride accidentally
leads to the pot of gold at the end of the rainbow, they will fail. The
limitless bounds make it impossible to measure any progress.
In
that same vein, the journey is as important as the destination. Many
traders pay lip service to limiting drawdown. Few are able to
effectively rein in drawdown, especially within the context of an expert
advisor. To a certain extent, limiting drawdown is not possible. My
experience with money management makes it clear how much wandering a
return can make, even with a heavy advantage.
The easiest way to
limit drawdowns is to fight against severe losses. The goal is to obsess
over risk and reduce it to the lowest possible point. I want to
emphasize this point carefully. Risk always exists. The market
chaotically switches from deep slumber to extreme violence. No amount of
system engineering eliminates or alters the structure of the market.
The most effective way to handle risk that I see is not to prevent it
from happening (not possible), but rather how to react when risk
eventually flares up.
Approaching system design as a process
rather than a destination also encourages global thinking. My old boss
would describe it as the 40,000 foot perspective. When a forex trader,
for example, emphasizes a system's percent accuracy, it typically comes
at the expense of exiting at a better point in the market. A need to win
slightly more often actually drags the system away from its optimal
performance. A process oriented design watches how changes in the entry
method affects the
exit efficiency. Clever
money management removes the emphasis from entry and exit methods.
Designing
these systems takes hundreds of man hours or results from more than a
decade of experience. Emotion will certainly enter the equation at some
point. The real question is not how to remove it. It is how to
appropriately channel it.
Designing an expert advisor quickly
takes on the characteristic of an obsessive hobby. If the system's
success is measured in quantities of dollars, the emotional roller
coaster rises in tune with the account balance. A more appropriate use
for the emotion is to take heart in the system working correctly, not
profitably. Ideally, correctly also means profitably. They are not the
same thing, however. Drawdown and unlucky losing streaks are an inherent
part of trading. Gaining satisfaction from the trade behaving as
designed rather than as desired makes your emotional well being far less
dependent on market performance.
I have been running this company
for nearly five years and literally spoken with thousands of traders
over that time. In spite of that overwhelming number, I still have yet
to speak with anyone that traded a commercial EA successfully for more
than a few months.
The successful traders that I know personally
all developed their systems and strategies on their own. Maybe that
reflects on the abysmal quality of expert advisors for sale on the
internet. Maybe it reflects on not-so-expert trading coaches, or perhaps
genuine trading coaches and hapless traders unable to follow sound
advice. I suspect it's a little of everything. Regardless of the
problems, the take away point is that the person doing the succeeding is
also the person doing the developing. They take control of the
experience from the beginning and lead it into a successful outcome.
Systems trading is not a process where you buy your way to success or
follow like a sheep to green pasture. You must earn it.
Challenges of using a winning expert advisor
The
path to achieving a profitable strategy on paper is a long, hard road.
The switch from theoretical turns to live trading is a challenge in its
own right. Backtests spit out instantaneous results. Traders see
something like a 20% annual return, then mentally prepare themselves for
the steady accumulation of funds.
That kind of trajectory puts a
number in the trader's mind that is terribly misleading in real life.
Steady annual growth of 20% suggests a constant return of 1.67% (20/12).
That number 1.67% is a hiccup in a trading account. Many forex traders
risk more than that amount on a single transaction. Factoring in the
inevitable drawdown or periods of loss, the market makes it impossible
to distinguish whether or not the expert advisor is behaving correctly
or if it is falling apart.
Uncertainty in the face of risk makes
any normal person panic. At the very least, it instills a degree of
anxiety over the future. The moment that the machine stops churning out
profits is the moment where doubt enters the picture. Doubt feeds on
itself, which leads to changing the trading plan in midstream.
I
can tell you from experience that the transition from research to
execution is huge. A learning curve exists, like with everything. I
learn to cope by taking the confidence that I gain through the same
debugging techniques
used in testing. The same techniques work for comparing live results to
backtested results. Does the backtest over the same time period as my
real trading match the actual performance? If so, then all is well. If
not, it at least gives me as the developer an opportunity to think about
specific problems. Execution or
forex broker manipulation
can cause problems. More often that not, however, is some flaw in the
strategy design or code. Developing a methodical, reason based process
helps calm nerves and to stay focused. The alternative is a general,
undefined worry.
