Machines + Humans
Humans and machines have fundamentally different strengths and weaknesses in trading and decision-making. In recent months, markets have become more volatile and uncertain, and our ability to trust machines into making complex and thoughtful decisions has also become more difficult. At QB, we love algorithmic trading, but we know markets cannot operate without humans at the helm. Financial markets would be so different without algorithms. While they have been indispensable in equity markets, they are still underutilized in fixed income. Algorithms overall have made markets more efficient. Their ability to analyze and process data while simultaneously making decisions makes them an invaluable tool. Yet, when financial markets become extremely unstable and volatile, who should be executing?
At Quantitative Brokers, we are experts at designing the best execution algorithms which help cut transaction costs, and boost productivity and efficiency. Our suite of algorithms are constantly improved, and robustly built to make intelligent decisions under any market circumstance. Each strategy is built for a different purpose, using hundreds of real-time data points to confirm the market's stability ahead of placing client orders. We have thought of it all, numerous preventative measures during execution in the event of a flash crash movement or unexpected volatility. The intelligence behind the algorithms helps them analyze broader market microstructure conditions in real-time, assisting traders to capture regime cycles while also providing a 10,000-foot view of the markets via our transaction cost analytics (TCA).
In our latest whitepaper, we analyze 10-years of data from CME’s E-Mini and 10-Year Futures to understand the fraction of the markets operated by humans or machines. Since algorithmic trading is untraceable in data feeds, we used order lot sizes as the critical differentiator because machines tend to break up trades in irregular sizes, while humans tend to trade in round lots. (Yes, we are predictable). We used this data to make a ratio (score) and compare machine and human trading distribution. The overall downward trend on E-Mini futures shows the adoption of machines trading in addition to other changes in the microstructure; the red line shows volatility (as CBOE VIX Futures) for the corresponding months. The data indicates that humans adopt technology when there are uncertain moments in the market. Another explanation is that the quote sizes are relatively shallow during market turmoil, and employing machines to “watch and wait” until the markets correct is a better strategy. We have highlighted this phenomenon in several papers related to regimes and execution quality.
This relationship between humans and machines trading and market volatility could differ for other instruments and market microstructures, and we plan to explore it in the future. For now, even if it is not a perfect measure, it’s an inferred estimate that the balance between humans and machines is the best combination when markets are uncertain and volatile. Algorithms, in many cases, trade better than humans because they are unlikely to react based on emotions or biases. They approach regime changes differently and reduce human errors, which can cause flash crashes that are quickly presumed to be created by algorithms.
The QB algorithm suite is built differently. We have tailored them to be your best partner in any market environment. They reduce market impact during execution, navigate irregular market movements cautiously, and hide your footprint in the markets. We are constantly improving and applying new microstructure techniques, recalculating our robust models, and adding to our framework with hundreds of proprietary signals. There will always be a role for humans in trading. We can make subjective decisions based on experience and intuition. Our ideal clients want to be bionic traders combining human intellect and QB’s intelligent algorithms to achieve the best execution.
Quantitative Brokers
April 6th, 2022
New York, NY