This is the first installment in a series of interviews with financial authors where we explore the business of trading.
Ernest Chan is a former trader for investment banks (Morgan Stanley, Credit Suisse, Maple) and hedge funds (Mapleridge, Millennium Partners, MANE) and he currently manages money for QTS Capital Management. He received a Ph.D. in physics from Cornell University and was a member of IBM’s Human Language Technologies group before joining the financial industry.
FIXGlobal: What are the three most important skills for traders to have? What was on this list when you started in the industry, but has since been removed?
- The ability to see through the randomness and complexity of the markets and to be able to distill the simplest recurring patterns or principles, much like a scientist analysing reams of data and coming up with a simple rule to explain them all.
- An appreciation of the laws of probability and statistics and not to succumb to the many behavioural biases such as realizing losses too quickly (‘risk aversion’) or putting too much weight on a model’s recent performance (‘representativeness bias’). Many of these biases limit short-term losses whilst also limiting long-term growth in wealth. Basically it comes down to an ability to endure pain in a rational way.
- Programming skills – as quantitative traders, we express our ideas in the form of a program. You can, of course, hire a programmer to code ideas for you, but it is very hard to ensure that they capture your ideas completely faithfully. You also need to be able to test modifications to your ideas quickly. Even if you have the resources to purchase high-end trading platforms that allow you to graphically specify a strategy, they are never as flexible as a general programming language.
I used to think that mathematical skills were important to quantitative trading. I have since changed that view as many successful algorithmic traders use a minimal amount of mathematics.
FG: What significant advantages do quantitative traders have over traditional fundamental traders? What are some of their disadvantages?
EC: Quantitative traders rely more on statistics so they can be more objective in their reward versus risk assessment of the trading opportunity. They are therefore less subject to behavioral biases. However, as no model is able to capture all the relevant aspects of the current market environment, fundamental traders can apply such knowledge to avoid applying an outdated model to a new environment.
FG: Given your experience in teaching others to trade algorithmically, which markets seem to display the greatest aptitude for algo trading?
EC: I believe all markets present opportunities for algo trading, and none are more favorable than others. The defining characteristic is liquidity. Algo trading can thrive wherever there is sufficient liquidity.
FG: Is it feasible to competitively differentiate an algo amongst brokers’ and vendors’ offerings, or is the amount of effort required not matched by increased performance?
EC: Brokers’ and vendors’ offerings can have a substantial impact on strategies that trade at higher frequencies. Higher frequency trading depends very much on data and order latency, and so both the efficiency of our trading platform and the latency of the broker’s internal network matters a great deal. In fragmented FX markets, different brokers offer quite different quotes and so profitability of the same strategy can vary considerably.