Sonia Schulenburg, founder and chief executive of Level E Capital, says the firm, which has just launched the Maya Market Neutral Fund, draws on the use of artificial intelligence techniques to model price behaviour, find and exploit market inefficiencies, offering investors highly liquid strategies and complete transparency.
GFM: What is the history and background of your company, principals and funds?
SS: I founded Level E in 2005 as a research boutique specialising in the development of state of the art artificial intelligence trading technologies. Over the years we have built a strong team and been fortunate to receive financial backing privately and through competitive research grants from the government.
In 2009 we launched Level E Capital to focus on the provision of fund management services. The Level E Capital SICAV was established in early 2010, and our first fund, the Maya Fund, was launch in April of that year.
The Maya Fund received institutional investment from one of the UK’s leading investment groups. The Maya Market Neutral Fund was launched in January this year, seeded with capital from an established fund of funds. This, along with our US-based managed account, has given exposure to a broader investor base.
I hold a BSc in computer engineering from the Instituto Tecnológico Autónomo de México, a PhD in artificial intelligence from Edinburgh University, a postgraduate degree in corporate strategy and finance from Edinburgh-Napier University, a professional certificate in accounting from the University of California San Diego, where I also studied bioengineering. I like to learn, or I like universities!
I also worked for an investment bank in Mexico City, where I began developing my interest in modelling the behaviour of financial market participants, focusing on their decision-making process, and integrating it into a fully systematic trading technology covering the entire investment process.
GFM: What is the structure of your funds?
SS: The Maya Market Neutral Fund is a sub-fund of Level E Capital SICAV, which is licensed and regulated as a Professional Investor Fund by the Malta Financial Services Authority. The fund offers daily dealing and liquidity with no lock-ups. We are focused on being entirely transparent and offering liquid and scalable strategies.
GFM: Who are your main service providers?
SS: Bank of America-Merrill Lynch is the prime broker and custodian, Deloitte & Touche provides auditing and tax advice, and Valletta Fund Services offers fund administration, transfer agency and accounting services.
GFM: What is your distribution strategy and targeted client base?
SS: Our target client base is professional investors, initially within the family office, fund of funds and institutional marketplace. Our approach is entirely scalable and therefore the size of investment does not alter the strategy. Our focus primarily is within the UK and Europe and we offer euro, sterling and US dollar share classes.
GFM: What impact has the recent global financial crisis and economic downturn had on your business?
SS: Because we offer low volatility and correlation to markets and other hedge fund styles, the recent crisis has opened up a range of totally unanticipated new opportunities for us. Recent market conditions have certainly motivated investors to focus in great detail on operational risk and due diligence, which I believe is one of our strengths. As an emerging manager we recognised this need from an early stage, and have sought to employ best practices operationally across our business.
GFM: Please describe your investment process.
SS: The use of artificial intelligence techniques lies at the heart of our investment process. Advances in theory, in combination with computing power, mean that real world problems can be addressed through the use of AI. This has been particularly evident in fields including medicine (diagnostic and predictive), the military (novel aircraft manoeuvres, security), engineering, fraud detection, pattern recognition and control systems.
Through the use of quantitative algorithms, we employ the latest AI techniques, specifically ‘machine learning’ in order to model price behaviour, find and exploit market inefficiencies.
Our investment process is driven by a fully automated trading platform that is dynamic – it self-evolves through time. Through the gathering, evaluation and storage of stock price and volume data at a tick level on a daily basis, our models are able to identify ‘signatures’ – patterns – in sequences of individual, pairs and clusters of stock price data. Our system can then optimise which models to allocate to, given current and predictive market conditions.
What makes our approach different is that rather than just produce buy and sell commands based on identifying a pre-programmed situation, our systems can alter the decision-making process through time. A dynamic framework is in place for the system to learn continuously and adapt to changes in market behaviour or structure. Importantly, the Level E team has enabled the system to test and train the algorithms more regularly and using smaller time series data intervals, what we refer to as continuous learning.
GFM: How do you generate ideas for your funds?
SS: All the Maya trading signals are generated through analysis of thousands of potentially predictive signals, gained through analysis of price/tick and volume data. Our proprietary systems clean, store and interpret this data for use across our models.
Our work is focused on research and system design. We research three areas: science, markets and technology. I manage the research and technical projects at Level E, working very closely with our chief technology officer and chief architect, and with our research analysts. It is through this continual investment in science and research that we can optimise our systems and model.
GFM: What is your approach to managing risk?
SS: One of the most challenging tasks in the analysis of financial markets is to measure and manage risk sensibly. The abrupt moves in market volatility and changes in market structure created large-scale losses during the sub-prime crisis in mid-2007, and such periods have persisted one way or another to date.
Traditional risk systems are typically built on academic assumptions that the risk factors (such as VaR) are normally distributed. However from practice we know that financial risk does not follow normal distributions! In fact, these tend to be leptokurtic distributions with a high peak and fat tails.
This makes risk management a complex area, and the industry has started to develop the next generation of risk systems. We went through several phases of R&D to create our risk management system, AI-Risk. It has been built in-house, as has all our infrastructure, and it is fully integrated with every step of the investment process.
The robot trader (the order management system) and AI-Risk applications read and react instantly to market moves (within milliseconds or microseconds). The system is capable both of reacting very quickly to an event, and acting in anticipation of an event we think is coming. The main function of the decision-making module is to predict under strict risk constraints, not just react.
Diversification is also an important factor. We diversify across stocks (such as across geographical regions), models and market features (learning methodologies within AI), and across execution styles.
GFM: How has your fund performed?
SS: We are looking for the Maya Market Neutral Fund, launched in mid-January, to deliver approximately 10 per cent per annum with low levels of volatility (below 5) and low correlation (below 0.5) to major indices.
The Maya Fund combines three distinctive strategies: market neutral, directional long/short and directional long only. In 2011 net returns exceeded the performance of the FTSE100 Index and major hedge fund indices such as DJ AllHedge, HFRX and the Barclay Hedge Fund Index.
GFM: What do investors currently expect from managers, and how do you deal with those expectations?
SS: Investors are increasingly looking for managers who generate returns with low volatility and correlation to mainstream indices, as well as full transparency, real-time reporting, high liquidity (e.g. daily pricing), and absolute or equity-type returns over rolling 12-month periods.
We believe we can meet this combination of expectations, adding diversity to an investor’s portfolio and helping them to better deal with highly dynamic market conditions.
GFM: What differentiates you from other managers in your sector?
SS: We are research-based, fully transparent and target lower-risk returns. The term ‘black box’ is often used, but we are certainly not one. We offer investors highly liquid strategies and complete transparency, full performance attribution in real time via mobile phone, iPad, web app, and daily one-click reporting via our Dynamic Factsheet Reporting system.
GFM: How do you view the environment for fund raising over the coming 12 months?
SS: We have had encouraging conversations with a range of investors, and we are optimistic about raising assets. In this environment any product has to be extremely competitive, whether through strength of record and return profile, or through the investment premise that the funds are built around.