Productive vs Non Productive Discretionary

Trader's Mindset

Sep 26

I wanted to share some observations I’ve made regarding when discretionary input can help or hurt the performance of a systematic strategy and in specifically what ways it is likely to do so.

First, fear based decision making is most-likely to hurt the performance of your systems. A typical fear based decision is when something hits the news and a trade triggers and you think “this time is different”.

On the other hand, if a trader is able to overcome various forms of biases, and process the market neutrally and from an expert understanding of the other traders in the market and the factors likely to drive the market in the near term then that is where the greatest value comes in.

Non Productive
Fear based or greed based
Catastrophic reasoning
Anchored in the past
Hopeful thinking
Feeling:Emotionally charged (amateur emotion)
Thinking: Driven by limited observations with limited insight into the expectation

Insight driven
Processing driven, i.e. from processing the market in real-time
Feeling: Expert intuition (expert feeling)
Thinking: Driven by statistical research and active study (expectation based)

Something I have seen is that even for expert level discretionary traders it is easy to shift from the productive to the non-productive processing modes if not careful. For example, a trader may identify the factors that are driving the market and predict it accurately but then succumb to fear based logic. One possible short-circuit for this is to force oneself to be continuously active.

A example of how this can work, a trader has a demonstrated system that buys weakness. The trader identifies near-term factors in the market that will signify continuation instead of reversion. The trader accurately is able to capture some of that momentum. Often in cases like these, after the insight is realized then switching to the system, provided it is still nominal, is going to be the optimal course of action.

If a discretionary trader can find ways to stay engaged with the market then they are more likely to be able to identify those predictive factors that are driving the other short term traders and the markets.

Elite level discretionary trading is likely the result of market cognition which consist of implicit learning combined with rational or speculative understanding of the other traders or factors driving the market. Market cognition is what really drives markets. Market cognition is strongly on the “right hand” envelope of trading because it is an active unfolding and dynamic process. Because of these factors, discretionary traders will often do best when they trade more frequently. The limiting factor and what must be carefully accounted for is that the spreads in futures, commission, and fees if not completely controlled for will make it impossible to profit over costs.

But, I mention it because most good programmed systems are very selective with high profit factor. It is rather difficult to improve such systems because (1) they are already strong and (2) trade too infrequently for the implicit learning aspects of trading to be maximized. This makes such systems poor candidates for general improvement. However,  and here is a possible hint for those who are interested to try it: you would need an active program to train yourself on specifically for those conditions. You would need to decrease the selectivity of the system to generate more signals, reduce to marginal costs, and use your discretionary insight and training to make up the “difference”.

And, of course, that’s a huge amount of work for something that you already will know how to make work by controlling the variable for the profit factor and just running the system. The alternative probably makes more sense which is simply to let the system be and obtain the benefits of diversification.

About the Author

The author is passionate about markets. He has developed top ranked futures strategies. His core focus is (1) applying machine learning and developing systematic strategies, and (2) solving the toughest problems of discretionary trading by applying quantitative tools, machine learning, and performance discipline. You can contact the author at