Difficult Question

Read this in "The Evaluation and Optimization of Trading Strategies" recently. Very thought-provoking:


Where is the application that lets the strategist design, create, and apply
her own custom genetic search algorithm to a Walk-Forward Analysis
of a trading strategy that employs an autoregressive integrated moving average
(ARIMA) forecasting model of volatility, a neural net that predicts the
magnitude and direction of tomorrow’s close change, which auto-adapts to
market conditions and has a genetic algorithm that balances one’s portfolio
automatically, strangling models trading stagnant markets with a strategy
that is losing in the last year and feeds the strategies that are profiting unusually
well and markets that are really moving?
Why is this important? Yes, such a trading platform would be quite sophisticated.
These technologies all exist now, however, and did so in the
1990s, too. The computer hardware is now up to the task. There are strategists
who can design and create applications like these and even more sophisticated
ones, at that. Yet, the trading strategy development application
that would make such a thing possible does not yet exist (at least to my
knowledge and I am always looking). One might ask, so what? That, however,
would be a rather uninformed question. For the existence of such a
trading strategy development application with the ability to integrate and
apply such technologies at a usable speed would really be a development
that would be at least somewhat in proportion to the vast explosion of
computer hardware that we have seen over the last 15 years.
You can count on the fact that if one of the large prop trading firms
desires to apply a trading approach as complex—or vastly more so—as
what has been mentioned, they have the resources to pull it all together to
create and trade such a trading model.
The point is that this capability should be available to the average
trader and investor too. The capability is there. The leading trading strategy
development software application vendors have been complacent—perhaps they just lack imagination—in the extreme. Part of the problem
here as always been that most of the developers of the trading strategy development
applications do not have a trading background; as such, they
are not as driven by trading profit as traders are to continuously seek out
cutting edge technology that can provide the trader with an edge. And the
trading community at large, primarily the individual trader and investor,
has been willing to accept this sorry state of affairs. Perhaps they too lack
imagination. . . ?


Cheers,
Dima

 

I stopped reading the thread after seeing the word "her".