Start timing of EA operation

This may be a rather troublesome issue for those who use EAs,

I think it might be quite a dilemma. (^^;)

Here,

this “start timing of operation”

could change the thinking significantly between a “low-win-rate EA” and a “high-win-rate EA.”
I believe so.

What I mean is,
first of all, for a low-win-rate EA (i.e., a low PF EA),
in my view,
it is better to “operate after a certain amount of drawdown.”
I feel this is appropriate.

To explain further,
a low-win-rate EA
is an EA that aims for a rising equity curve while repeating wins and losses to a similar extent,
so there will be a number of losing trades as well.
(In some cases, there can be a streak of losses, too.)

Therefore,
when starting operation after a streak of wins,
there is a higher chance that losing trades will occur,
somewhat more often than not.

Strictly speaking, it might not always be the case,
but from my experience, such patterns appear relatively often.
(⇒ I think many of you have experienced this as well.)

For this reason,
we think that “low-win-rate EAs should start operating after the drawdown”

,

and next, regarding “high-win-rate EAs.”

In my EAs,
Revolution and the ED version, Scal_USDJPY, Koshigure (Powder Snow), etc., fall into this category.

If, for these EAs, as described above, we also wait for a drawdown,
what would happen?

For example, Revolution,
this EA has remained undefeated in forward testing for nearly a year.

In other words, since the start of sales,

if you had “waited for a drawdown”
you would have missed out on about a year’s worth of profits.

(Of course, if you waited to assess forward performance, that can be meaningful…)

By the way, the revenue in that forward one-year is

“about 1400 pips.”

How one should interpret this is the question.

In other words, for high-win-rate EAs,

if you wait for a drawdown, you may well miss out on profits that you could have earned (e.g., 1400 pips per year).

That is how I think.

In short,

regarding “high-win-rate EAs,”

no matter what, the key is to start operating as early as possible.

I feel this is the right approach.

For example, at this very moment,

if you refrain from starting operation to observe,

and then look back at the same time next year, and still see that you had undefeated performance for a year,

that possibility is quite realistic, I think.

(Of course not 100%, but considering past results, it is plausible.)

In a sense, this is an obvious matter,

for a “highly advantageous EA,” whatever period you take, the likelihood of positive earnings is higher, and

the earnings from this type of EA correspond simply to the length of time the EA runs.

In other words, there is almost no need to wait for a drawdown,

if the current performance is good, just start operating,

and then scale the lot sizes according to the revenue,

and, as needed, stop the EA based on key indicators,

these are the measures I think are appropriate.

~Postscript~

※The above is merely “my personal viewpoint.”

Details such as the start timing of EA operation and lot size

will be left to the judgment of each user. Thank you. m(_ _)m

Neko Hakase