AutoFX - Mick

[autofx]

The computers are actually flying the planes much
more of the time than most passengers would ever dream.

There are situations where you DO want a human on
board who can truly FLY THE PLANE though, and I recently
heard of an incident where a commercial pilot's experience
as a glider pilot is all that saved himself, his crew, and
the passengers during a power outage.

I sincerely believe I've achieved a balance between
surviving and even thriving in choppy conditions and
doing well in strongly trending conditions. Both 2003
and 2004 hourly data are provided in my backtest kit.
Forward testing (real trading) is going well.

A few months ago, I reached the point where I can
say I'd much rather trust my computers to trade my
systems than I would human traders to execute them
for me. My human abilities are needed two or three
times a week to restore the datafeed when there
are Internet service or online broker hiccups, but
that's all I'm good for in regards to keeping the system
running.

There are no secrets to what my systems do. The tactics
are straightforward. Setting a modest 22-pip or 26-pip
profitlock stop after an hourly close 40 to 50 pips in a position's
favor is obvious, effective, easy to understand, and easy to
implement. Tightening initial stops slowly after a certain amount
of time has elapsed is simple to understand and simple to
automate. Executing a trailing stop scheme where profit-taking
stops are slowly tightened according to the market action
as large moves go in your favor is also straightforward (though
this is the most complicated part of my systems, algorithmically).

The indicators to trigger trades are almost unimportant, but
not quite. You mainly have to make sure you enter in
conditions that are statistically likely to give many shots
at very good entry points. You don't want to mindlessly
toggle, as in crossover systems. I've proved to myself
that those systems are no good.

That about covers what I wanted to say.

I hope it makes sense to someone besides me.

[MickMason]

I'm reluctant to admit it but I can see the potential in such a system, the main advantages being emotionless trading and limited screen watching. Perhaps this really is the way of the future, maybe I'm just too set in my ways, and you know what they say about an old dog and new tricks!

It's not so long ago that personal computers were rare and only used for basic functions like games and accounting, and the internet wasn't even thought about....now look, it's amazing how many things are automated and decided by computers, why not trading I guess.

It's sad in one way, the required skills and pride have been replaced by a machine......jeez, I sound old!

Mick

[autofx]

I think the trading profession is going to be slow to embrace
automation. Lots of money involved, lots of turf to protect.
Stodgy mindsets in the financial professions. We'll see. It's
an interesting time to be alive.

[united46]

Slow to embrace?

What about all program trading in the stockmarkets as well as all the hedge funds that use arbitrage? That is mostly automated.

Also, I read an article the other day that showed how volatility has dropped dramatically over the last few years as program trading has increased. Making daytrading harder and harder as range has been cut. Maybe the PPT referred to in other threads?

Anyway, will automation kill the goose that lays the golden egg?

United

[autofx]

I don't think so, because plenty of people will still trade
for the thrill.

Commercial airplanes do a lot of things automatically,
but there are light planes, hang gliders, parasails, ultralights,
etc., because people love the rush of flying. Trading is
as much of a thrill as flying for some.

Maybe the analogy doesn't apply but I don't believe we're
in danger of robotized trading taking over the markets
and destroying volatility.

[TOPGUN]

I am torn between wanting to believe ( I guess in all of us is the hope that something will work) and the proof I have actually seen.

A member of MT some time ago....cant remember who.....bought 10 yrs of tick data at some crazy cost from a bank to get pure (as pure as you can) data as that is what is important.....

He then tested something like 50 systems that had been programmed on 1 - 2 yrs data and were giving the developers the belief that they had finally found something that worked.....not over the top systems but just 20-50 pips per week systems.

The results showed that not 1 !!!! of these systems worked over the 10 yrs and that in something like 95% of the systems they were only profitable over the yrs they had been programmed for....i.e if the developer had 12 months data it worked on that 12 months.

why....dont know....but I guess it has something to do with, optimizing, curve fitting and the fact that 1-2 yrs is not as long as we think in FX and that the mkt has subtle and not so subtle changes that basically end up ruining what looked like a winning system

[autofx]

Just because this one group of people couldn't do it
doesn't mean it can't be done.

The Russians couldn't send a guy to the moon, remember.
It didn't mean it was impossible, because someone else
pulled it off.

Let me ask you this: How long a period of profitable
automatic trading would it take to convince you?

[TOPGUN]

who knows....thats the point

But if on 10 yrs of data it was profitable every yr (neglecting the fact that back testing is not 100% accurate) I think i would feel very unlucky if I traded it and it didnt make money.