Quantified Knowledge

Trader's Mindset

Dec 29

There is something I have noticed, over time the more that one shares, that is writes, on trading (or any topic really) that the more likely one is to conflict with ones own writings.

Early on, I really liked one trading book– which I won’t mention the name of. I had made notes and written down quotations from the book which I later referenced to support parallels that I seen in my trading. When the author became aware that I had referenced his work, he was very upset and was adamant that he didn’t claim the things I referenced in his book. I was flabbergasted when the author started arguing against his own quotations– but I shouldn’t have been because any author, who writes enough, will eventually come into conflict with their own word. And, this author’s book was filled with self-conflicting information.

Now, there is another famous trading author who routinely writes material that conflicts with his opinions that he wrote in the past. This author writes material that conflicts on a regular basis. There are a few reasons this author is so self-conflicting. First, he produces a huge body of work on trading which is just bound to conflict and second most of what he writes is very generalized.

You see this is the second aspect on writing– the more generalized statements one makes then the more likely that one is likely to conflict with oneself. It is basically a known that science can only say the most about the least.

While many lessons from trading may not generalize, I do believe there are several lessons, however, from these observations.

The first is to take anything that anyone writes as tentative hypothesis, uncertain, and incomplete. For example, we often hear anecdotes on trading that sound sensible but conflict– such, “as no one ever went broke taking profits” or “the trend is your friend”. In my own trading, I have found that very specific clusters work together. For example, the trend is your friend if you have a strong bias on the day and hold until the end of the day. However, if you set tighter stops then reasonable profit targets are essential.

The second is that the narrative form of traditional blog posts, books, and other written forms most often lacks the specificity to help traders and if you really want to help traders then you need to provide a greater degree of specificity and greater total information content. Written text is poorly suited toward these objectives.

Third, if you want to talk about trading performance you have to talk the numbers. Talking about trading performance without numbers gives us no context to evaluate the worth of a method. We need to see a least a few weeks of data– even if that isn’t conclusive.

Fourth, traditional narrative forms of learning are unlikely to provide real value and that you should look toward the tools of data science, collaboration, and rich information sharing to make a difference.

Fifth, given that one will inevitably conflict with oneself, one should not attempt to be completely self-consistent. Self-consistency is only valuable as long as it is.

I claim to be offering some of the best information on trading and I believe the relatively high specificity of information in my posts supports my claim. In other words, you can objectively measure that I am providing better information then most others out there– and I know how to provide even greater value.

Of course, no one wants to talk numbers because (1) they know that the performance is not good enough to really be meaningful for most small traders or worse the methods only work about 50% of the time, (2) the author doesn’t have enough experience with the method to truly know *and is unwillingly to be ruthlessly honest like I have been*, or (3) they are concerned to share any performance due to legal or regulatory concerns— a possible solution for this is just to be clear that the performance may not be indicative of future returns.

To be clear, two things that bug me are people who commentate the markets post-fact and authors who write about trading performance without talking numbers. If you really want to share a trading method for the benefit of the trading community and you cannot provide a backtest then you should at least be able to provide a few weeks of performance data. Notice even here, I conflict with my admonitions against trusting track records. However, in this case, the reason that sharing performance is relevant is that without performance there is no context for the value of a method. Are your methods consistent north of 68% of the time? Are you able to average 30% of your max daily risk?

Yet, the biggest lesson may be an admonishment against overly narrative forms of learning– we have more computing power then ever before. Technology has already eliminated many professional trading roles– and yet most retail traders are still trying to trade without a technology first perspective.

My next posts and future technologies, to be released in 2020, will fully demonstrate and realize the potential of applying technology toward trading development and performance.

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 curtis@beyondbacktesting.com.