Info!! A Systematic Approach To Discretionary Trading

There's a lot to live on said for wandering.  Pick a promising area, wander about, too you lot tin discovery quite a bit.  Discovery is all well-nigh opened upwardly eyes too opened upwardly minds.  

Here's how I've been wandering through markets lately:

Speaking amongst trading too investment professionals, specially inwards the quant world, I've been struck yesteryear the fact that the models they employ to direct their decisions are non the sort of models nosotros typically read well-nigh inwards trading texts.  There are no technical indicators or nautical chart patterns inwards their inputs.  Nor are at that topographic point whatever inputs pertaining to society earnings, economical growth, upcoming fundamental banking concern meetings, or geopolitical events.  Rather, returns from markets are broken downwards into basic "factors", such equally momentum/trend, volatility, value, too carry, amongst models designed to capture these factor-generated returns.  

A discretionary trader powerfulness justify trading a tendency next method or a mean-reversion/reversal method based on his or her "personality".  This is nonsensical to the coin managers I verbalize with.  It's similar maxim that I'm correct handed, hence I'll exclusively accept correct turns inwards my car.  If returns come upwardly from a diversity of factors, the best surgical procedure tin live on achieved yesteryear trading signals derived from each of these factors.  This volition diversify returns too arrive at amend risk-adjusted results.

A comparing of returns from quant property managers vs. discretionary trading firms finds that returns indeed have got been amend from the erstwhile group.  That is non exactly because these funds are quantitative--I tin hollo back of quant funds that have got lost coin lately.  Instead, the superior surgical procedure comes from generating returns from multiple factors across multiple fourth dimension frames.  Many goodness models producing independent, positive returns--not necessarily eye-popping ones--can combine to shape a robust P/L stream.

So I began my wandering.  Of all the marketplace position information sets I track, I identified the ones that: a) produced the most reliable too valid trading signals; too b) had real depression correlation to ane another.  Influenza A virus subtype H5N1 full of half-dozen variables popped up.  To my surprise, ane was based upon volatility; 2 were derived from cycle-based forecasting methods; 2 were based on momentum (trend); too ane was based upon value (mean-reversion).  I built half-dozen uncomplicated forecasting models based on the half-dozen variables too hence combined the model outputs into a unmarried "committee of experts" signal.  (See this article for an overview of creating ensembles of forecasting models).

I was surprised yesteryear the aeroplane to which the trading betoken from combining the private forecasting models handily compaction whatever of the private models.  Still to a greater extent than surprising from my perspective was that the combined trading model seemed to "know" when to merchandise similar a tendency trader, when to merchandise similar a reversal trader, too when to brand coin from shifts inwards volatility.  The model seemed to operate yesteryear navigating the ebb too catamenia of factors.   

Most surprising of all, however, was that when I trained the models to forecast shorter-term cost change, for certain forecasting models dropped out, some received extra weighting, too some less.  This raised the possibility of using shorter-term models inwards a Bayesian fashion to navigate longer-term signals.  That is, you lot would role short-term forecasts too returns to alter your assessment of the odds of a longer-term forecasted motion playing out.

I'm non at all convinced that this way nosotros should all toss our experience aside too larn programmers, statisticians, too systematic traders.  Could it live on that the discretionary "tape-reading" science of the short-term trader tin aid navigate shorter-term forecasts, exactly equally those shorter-term forecasts tin aid us participate inwards longer-term forecasts?  In other words, such trading would live on neither wholly discretionary nor wholly systematic.  It would live on discretion--with route maps.

But isn't that the way nosotros move when nosotros drive cross country?  We don't exactly rely on experience too intuition; nosotros await at maps too nosotros rely on GPS signals.  On the curt fourth dimension frame, however, nosotros *do* role our experience to navigate lane changes, pick out optimal places to halt too rest, too accommodate our speeds to route conditions.  The skilled driver has experience too route feel--and maps too GPS signals from trusted information sources.  Perhaps the skilled trader is non hence different, navigating 2nd to 2nd cost action, fifty-fifty spell benefiting from the route maps of forecasting models.

Further Reading:  Factors too Short-Term Market Returns
.