Systematic Trading and Quantitative Hedge Funds
Mon, 24 Jan 2011 00:40:00 GMT
An Interview with Dr Sonia Schulenburg
In this interview for the High Frequency Trading Review, Mike O’Hara talks to Dr Sonia Schulenburg, CEO of Edinburgh-based trading technology firm Level E Limited, and the guiding light behind MAYA, a complete financial modeling and investment solution that analyses high-frequency financial data to provide intelligent and robust investment strategies. Level E Capital runs the MAYA Fund, an absolute return fund that employs a purely systematic and fully automated decision-making process based on cutting-edge advances in Artificial Intelligence.
High Frequency Trading Review: Sonia, welcome to the High Frequency Trading Review. Can I start by asking you about your own background?
Sonia Schulenburg: This is actually pretty difficult! I guess one could say that my academic background is mixed with science and business.
My main degrees are a BEng in Computer Engineering from the Instituto Tecnológico Autónomo de México, (ITAM) and a PhD in Artificial Intelligence from the University of Edinburgh.
To help me on practical and business matters, I have also completed a Professional Certificate in Accounting from the University of California San Diego (UCSD), a Postgraduate Diploma in Corporate Strategy and Finance from Edinburgh-Napier University and an Enterprise Fellowship from the Royal Society of Edinburgh.
During the past decade I have also taken a significant number of workshops and courses related to entrepreneurship, startups, business plans, marketing, commercialisation, legal, intellectual property rights investment, and stuff like that.
HFTR: That’s an impressive academic record! Can you tell us about how you got started in the industry in Mexico and what brought you to Edinburgh?
SS: It all started while I was working in a leading investment bank in Mexico City back in 1995. I quickly realised that there would be significant pressure over the coming years, as I saw a new range of inexorable trends would emerge and the impact these would have in the modern financial world – one I was certainly very anxious to live in. In my view these were things like intelligent decision-making in a multi-class asset setting, full transparency, integration, speed and so on. I became obsessed with the idea of modelling behaviours of real traders, so I started working days and nights, including weekends, until I actually visualised a slightly more defined process. I arranged a meeting with the CEO of the bank and armed with a document I had written (which was over 100 pages long!), I proposed to him to support me financially in this futuristic quest, which involved a PhD in Artificial Intelligence from one the top Universities in the world specialised in AI. If you are about to, please do not constrain yourself to a good laugh. Many people have, but believe me, this is a true story! Not surprisingly, the CEO patiently explained to me, in a very nice manner, that they couldn’t support this idea as it involved longer time and investment than for example a Masters, which is what was being offered to a few exceptional people who had many years working in the bank.
So without financial support from the bank, I applied for a scholarship offered by the Consejo Nacional para la Ciencia y la Tecnología (CONACYT), which covered tuition, some extras and maintenance costs during the 5-year period. In addition to this, I also applied and obtained financial support in the form of a long-term loan in US dollars from Fundación Mexicana para la Educación, la Tecnología y la Ciencia (FUNED), and some impressive academic achievements from an early age gave me the financial backing I needed to pursue the next phase of my plan: devising and developing an intelligent trading system.
At the time, back in ’95, Mexico was suffering one of its worst economic recessions since the Great Depression of the 1930s and a significant devaluation of its currency, el peso, which went up from around 3.3 pesos for 1 U.S. dollar to a peak of 8 pesos per dollar by the end of 1995.
In the middle of this crisis, I decided to resign and leave an excellent and very well paid job. I embarked myself upon a one-way trip to a place called ‘Edinburgh’, with two suitcases and about GBP £10,000 for the rest of my PhD. I didn’t know where Edinburgh was in the map, and to be quite honest with you, it didn’t matter. What mattered to me was that it was a City with two major strengths: investment management and artificial intelligence – these were my carrots.
The rest is history. Maybe we can write a book one day. It has taken these last 14 years of my life to make this vision a reality. These are the origins of MAYA.
HFTR: What exactly is MAYA and who is it aimed at?
SS: MAYA is the trading technology we have developed at Level E.
Over the past few years, the combination of domain knowledge in the area of Artificial Intelligence and technical expertise in leading financial services technologies has motivated the Company to invest in developing its own product suite, which includes MAYA.
Our long-term objective – my personal mission and dream for over a decade now – has been to integrate in an automatic way all the steps of an investment management process, including the decision-making process. This is an extremely challenging task on its own.
The technology is entirely self-sufficient and has all been developed in-house. We do not have dependencies with any software vendor in terms of licensing. We develop all of our systems in a Linux platform and using open source software.
MAYA does not require human traders and teams making the day-to-day decisions of the fund, and roles involving mundane functions such as order placement, order management, back-office operations, compliance or reporting.
It accommodates investment management processes of different types, from traditional long-only styles to highly quantitative approaches. So it is aimed at a broad group of end-users, ranging from investment banks to hedge funds and private investors.
HFTR: So how does it work?
SS: It all starts with data…
If you are not a registered user, click here to register for free