ECC Information Management and Advisory Services

January 26, 2025, Hong Kong

With the first training program of Dr. Chou’s Boot Camp in Quant Trading finished in Hong Kong, I started to feel that, in many aspect, I was the trainee instead of a trainer. This is mostly on my thinking of how to design and deliver a practical finance training, like this quant trading program.

Although I have always been conducting academic research in finance (esp. on market microstructure), throughout my professional career, I have been a practitioner. I know what are important skills and knowledge areas for a practitioner, and such belief is the foundation and motivation for designing and delivering a training program like this one. However, like one has to constantly improve a trading strategy in order to achieve long-term positive investment results from it, any training program needs to be constantly reviewed and revised in order to ensure a smooth, effective and constructive experience for all program participants.

With details of the training program still fresh in my mind, I consider the following three areas that this program did well, and another three areas that future program can improve on.

First of all, the program seems to do well on the “practicality” of the training content. This training program stems from my understanding of the lack of practical discussions and implementation in the Market Microstructure course that I offer at Hong Kong University of Science and Technology (I offer a short version of it at the University of Chicago as well given time constraints). Although trying to strike a balance between theory and application of the subject, the HKUST course (and the U of Chicago one) lacks content that gives students a hands-on experience on execution algo development and market making strategy development. The intense python coding content in this training program allows trainees to gain insights into how an execution algo and a market making strategy really work in the real world. As a result, a trainee can essentially start trading such strategies based on what they learn from this training program, albeit they still need to adapt the code base from this training program into their own trading system (if it already exists).

Secondly, the combination of quantitative modeling practices and the utilization of such models in real life trading strategies is a good practice in designing and delivering this training program. In addition to learning how to conduct related quantitative modeling works (both the research part and the coding part), trainees gains first hand experience on how to implement such models in real life trading strategies. Furthermore, they can compare the trading results when they turn on the models with the cases that such models are turned off. Not only such back-testing result analysis allows trainees to compare different models, but also can they further understand the logics of the strategies, at least in the way that they are implemented and discussed in the training program.

Thirdly, the flow and the logistics of the training program did work rather well. This includes the venue set-up (early check-ins and dry runs always help), arrangements of lunch breaks for all trainees, sharing of contents via Canvas, and email communications with regard to registration for all/individual topics, etc. The training venue is of top quality in terms of classroom set-up, classroom technologies, etc.

Now come the areas that future training programs can help improve on:

First, the timing of the first training program was arranged to be too close to the Chinese New Year. As a result, some potential participants have left Hong Kong. Several them asked for on-line sessions, and some others asked if the program can be held in their own cities.

Secondly, at least for the first training program of Quantitative Trading, the content can be somewhat “squeezed“ into a program that does not require 27 hours. My feeling is that 24 hours may be enough, which means 8 sessions (each session with 3 hours) for the whole program. Therefore, a future program on this topic may be delivered twice, each of which can be delivered on Thursday, Friday evenings from 7pm to 10pm, and another two session can be delivered on Saturdays. Or, four sessions on Saturday and Monday; as a result, the program can be done in two weekends.

Thirdly, the use of canvas can be further enhanced. Some students complained that they were not able to access files on two occasions. A more formal canvas website (instead of the currently free one) may be able to help, which can also add its own Zoom links. Also, it may be desirable that trainees can use certain website to learn coding, instead of the current specifically downloaded python files. Downloading codes onto trainees’ personal computer can be sufficient, but to some students using a designated web portal to access the code and conduct learning can be easier to handle.

To summarize, I believe the most important practice in professional and practical finance training is “practicality” of the training content, and a good training program should focus on the applications of the training content. Once a good training content is created, the delivery of it – no matter via python code downloading or via a web portal (such as a designated website or a Github site) – should allow trainees to gain hands-on experiences of the training subject.