Within the time-frame that it takes to read this sentence, an ultra-low-latency trading algorithm can place millions of orders. If the reader is slow to process information and savors details, a few million additional orders can be placed. However, one error in the algorithm may cause a great turmoil in the market, and millions of dollars can be lost in a minute. Potential losses dictate that a better automatic trading interaction with markets is crucial, which was clearly proved by the Knight Capital fiasco in 2012. Knowing the loss-potential and having a desire to prevent these losses is one of the great motivators pushing me to advance my knowledge in the field of financial engineering. Moreover, the intensely secretive and constantly evolving nature of high frequency trading (HFT) sparks my interest and makes me want to better understand the mechanics of financial markets.
Early on in my career as an electrical engineer it became apparent that the field of electrical engineering itself is just a representation of mathematical phenomena visualized in the mind. Similar visualization for analyzing financial data structure plays a vital role. One must understand the power of alpha, and one needs to be able to explain profits and losses with a mathematical approach. The challenge is not only to build an algorithmic strategy, but to also predict the emotional state of traders influencing market flow. It is almost always uncertain whether future market volatility adequately reflects the historical data, which contrasts with the belief of using a purely rational approach.
Through my experience with the development of a trading platform in my current position as a novice financial software developer in a small HFT company, I have learned to appreciate the complexity involved in improving algorithm designs. I realized that the development of trading algorithms is not as simple as managing memory, creating parallel computing tasks, synthesizing C code or analyzing tick data and latency, but it is also related to the changes of coding habits for efficient use of a designated system. There are many factors involved in the design of an efficient algorithm, one essential factor being that the developer must consider a trading engine behavior on the other end. In addition to the anticipation of trading engine behavior, I started to think about markets’ inefficiencies. At the same time, I learned that some HFT strategies, such as latency arbitrage depend on the ability to access and execute market data faster from other competitors. Since we are closing to zero-latency, it is becoming more difficult to make systems operate faster. This challenge in hardware forces quant-traders to design smarter algorithms based on market data structure. Technologically, I enjoy the challenge of learning the best programming practices and implementing them to develop algorithms, ultimately achieving better results.
One day, I hope to find myself in an opportunity where I am not only challenged personally and intellectually, but I also hope to find a sense of fulfillment through making a tangible impact in the financial industry. My recent work at the HFT company is a stepping stone to reaching my goal, allowing me to be immersed in the creative environment, in which I have always hoped to be involved.
Through teamwork and collaborative projects, we can improve ourselves individually, while working toward creating smart trading algorithms and innovative trading platforms in the world of finance.