As an open-source community, we can stretch to the limits of where our members take us. That’s why we’re launching our newest blog series, Powered By LEAN, that tells the stories of the hedge funds that are using QuantConnect and running on LEAN, our proprietary trading engine. LEAN offers many benefits to funds, including it’s open-source architecture built on world-class technology, built-in risk models and slippage systems, and simply the fact that it’s developed by quants.

In our first installment of Powered By LEAN, we feature Chris Kramvis, Chief Investment Officer of fully-algorithmic hedge fund DropShot Capital.

Broad: Who is DropShot Capital and what is your role at the company?

Kramvis: DropShot Capital is a fully-algorithmic hedge fund. With a wealth of experience in quantitative finance and data science, we bring several years of experience deploying algorithms; building the data pipeline, exploring parameters which produce the best fitting models and testing their results in production. We are equally at home in the trading and tech spaces and have worked with big data stacks found in the internet space. I serve as the chief investment officer and Uppili Krishnamachari is our chief information officer. As a boutique fund, we wear many hats and aim to bring our extensive and diverse experience to the design of our investment experiments.  

B: How did you get your start in the industry and what led you to DropShot Capital?

K: I’ve always been interested in trading and got my start in the prop shops of Chicago. Working as an options quant, I developed models used for volatility arbitrage trading. From there, I transitioned into coding, testing and deploying HFT strategies in futures and equity markets. During my time at an internet startup, where I was running the data science team, I continued to trade longer-term strategies in my own accounts via QuantConnect’s LEAN platform. After gauging the needs of some initial investors, it seemed that exposure to AI and machine learning was in high demand. Given the interest, we decided to start DropShot Capital and offer the strategies we built on LEAN to other investors.

DropShot’s primary strategy is a machine learning approach to a relative value trade. Our approach uses both supervised and unsupervised learning when determining what products to trade and how to trade them. The current version of the algorithm is deployed in the ETF space, but it has also shown a lot of promise in equities as well. The strategy trades long only and can go defensive versus trying to short as a hedge, and can provide alpha even when the S&P 500 has a down year.

B: How did you come across QuantConnect and how does LEAN solve your previous pain points?

K: I discovered QuantConnect in 2014, after reading about the company in a LinkedIn group that focused on algorithmic trading. After signing on, I connected more in-depth with Jared about the platform, as he was quite excited by the possibilities it could bring to the individual and professional trader. As the feature list grows and LEAN’s functionality becomes more robust, we have continued to foster a great relationship with the LEAN team and QuantConnect’s community.

LEAN solves the major pain point of having to write the basic scaffolding to test algorithms. It’s a great way to go from idea to backtest in just a couple of hours. This allows traders to cut right to the alpha generation step without spending countless hours trying to purchase, clean and line up data sources from a variety of financial products and vendors. These steps, while very important, don’t add any alpha to strategies and consume most of a quant researcher’s time. LEAN’s technology stack is very impressive compared to the competition in that backtests are already distributed and run at CPU throughput speeds. We have also had no issues with our brokerage deployments and have seen 100% uptime of their colocated live trading environment every day.

B: What sets LEAN apart from other solutions?

K: LEAN’s brokerage integration and management is a feature we value very highly. It simplifies the process of sending orders directly to Interactive Brokers and other API brokerage services. LEAN allows us to worry about our strategy and focus on alpha generation. In our view, this is a great differentiator between LEAN and other technologies which either do not have brokerage integration or have poor execution. We have been able to trust in the LEAN infrastructure so that we can focus our energy in more productive areas, such as strategy development, news analysis and identifying new trading opportunities.

B: How do you see the open-source nature of LEAN as a benefit to hedge funds? How can LEAN impact the future of hedge funds?

K: Democratizing financial technology is good for everyone as funds can focus on alpha and become more efficient. We believe that all funds should take the systematic approach to investing that LEAN enables. Support for many coding languages also greatly facilitates contributions from the open-source community. Additionally, being able to study the underlying source code is a key benefit to crisp and concise algorithm development. We find the documentation to be very well organized and easy to use and we’re eager to continue to benefit from and contribute to the open source community. Having a shared resource is of great value to individual traders and smaller funds that would have to ordinarily re-create LEAN to start trading and compete against mega funds with big teams of developers and higher AUM. Having LEAN at our fingertips provides us with a mature technology in which to run our fund.

The industry is seeing a lot of the big players taking more and more of the assets. We believe technologies like LEAN will be necessary to enable both individual and smaller institutional investors to overcome the technological challenges of setting up a trading infrastructure in a cost-effective manner. This, coupled with better open source packages on the AI, machine learning and programming side, should allow for a lower barrier to entry and quicker time to market. The focus for smaller funds will always be on novel signal generation that really outperforms the hedge fund indices. By being able to go straight into algo development and deployment, small firms can begin building their track record right away. We also hope that the trend towards more API accessibility for all services, including data sources, continues to grow so more people can have access to traditionally hard-to-access resources. Many of the traditional long and short fundamental funds are now pushing towards automating portfolio management decisions as discretionary stock picking continues to be met with varying degrees of success. We think, as these firms have to re-tool and re-staff, there is an opportunity for those with the know-how to come in while the transformation is taking place.