Interview with Deep Blue Capital

Today at SECTOR we had the pleasure and opportunity to interview an ex-Kraket member to talk about his almost 5 year long experience working at Deep Blue Capital.

Could you introduce yourself?

My name is Luuk. I obtained my Econometrics bachelor and master degree at the Vrije Universiteit, where I was also a member of Kraket. Shortly after finishing my master degree, I joined Deep Blue Capital, where I have now been working for nearly 5 years. 

Could you give a summary of what Deep Blue Capital does?

Deep Blue Capital is a highly automated proprietary trading firm specialized in algorithmic trading that only trades with its own capital (no clients). We trade on equity and future markets across the world. With a quantitative and computational approach, we build our algorithms around the principle of statistical arbitrage.

What is your position within the company?

We don’t consider ranks and titles very important, but officially my position would be called ‘quantitative trading developer’. In practice it comes with plenty of different responsibilities. My main responsibilities are doing research into existing and new trading algorithms, the coding and implementation of algorithms, as well as the monitoring of all our actual trading activities and the financial markets.

How did you first hear of Deep Blue Capital and when did you decide to join?

I think I read about it on the Kraket website somewhere during my bachelor's degree. As it appeared to offer the combination of financial markets, programming and mathematics I kept it in mind ever since and decided to apply after finishing the master program.

What was it like starting working there?

It was a period with many new encounters. In the first months you will have to get familiar with all the internal systems we use. This means you will need to learn the basic idea behind the trading strategies and get to know how to guide the algorithms through the trading sessions. You will get responsibilities quickly, for example having to deal with our Asian trading on your own within a couple of months (yes, at night). Aside from that we have all kinds of infrastructure like SQL databases, Git repositories, Python libraries, Linux systems etc. you will be introduced to.

How is the culture at your company?

As we have no clients, we have an informal culture in our office. You will not see anybody wearing a suit. Since we have roughly 25 employees and a fairly flat organizational structure you will get to know all of your colleagues. Most of your colleagues will have a similar background. On the trading floor people tend to be focused on the trading or their projects. Outside of the trading floor we sometimes have other events, like (board) game night, poker tournaments or just having some drinks.

Would you like to share some of your fondest memories of working at Deep Blue Capital?

They are not necessarily the fondest memories, but the work-related memories that come to mind are the start of the covid crisis, as well as the Russian invasion of Ukraine, since our algorithms come into untested territories. Outside of working hours I think of several events, like Christmas dinners or our 10-year anniversary weekend.  

What does a workday usually look like for you?

A typical day could be me coming in at 08:15, well on time before the European markets open. After taking a quick look at how our Asia desk is performing, I read up on all overnight news and prepare several different algorithms for trading. At 09:00 we start trading and I take a look at our trades and positions. Throughout the day I monitor all European markets, but there is plenty of time to work on other projects as trading cools down. In general, the idea is to intervene as little as possible with our algorithmic trading, but it could be that some news comes out for which we want to make some adjustments. At 15:30 we start trading US stocks. After Europe closes at 17:30, I take a look at our exposures before heading home.

What are some projects you are working on right now?

I’m currently working on a project where I’m adjusting the way one of our algorithms computes a theoretical price for a stock. This process mostly consists of several steps. It starts with an idea. Often this idea is tested in a simplified market python environment. In case it looks promising we move on to implementing the change in the actual code, which is written in the language called Ada. Further testing can be done using our simulated bourse environment, before it’s taken into production. Of course, good results in simulation don’t guarantee good results in reality. As we are a small company there's plenty of other aspects where you might spend time on. This could be related to our risk management, where you might have to dig deeper into certain stocks, sectors or any news which impacts the financial markets.