Innovative and esoteric projects
We have a Google-like culture and indulge 10% of our time in R&D projects unrelated to Quant Finance. We have been studying ways to improve the intuitive cognition of high frequency data with shapes and algorithmic music derived from market data or trying to help create the best Go playing engine in the world.
-
Algorithmic Music derived from Financial Market Data - Baptiste Bohelay - Arbitragis Trading.
- We are studying ways to improve the intuitive cognition of high frequency data of an orderbook. We use a software inspired by the French research center on music and sound (IRCAM), Pure Data, to "shape" sounds of financial data. Our conclusion is that the hidden melody of markets is quite ... noisy. A future project will try and create better melodic lines based on stock prices, and try and create canon structures with highly correlated stocks. A Youtube video of sound and shape of our resarch will follow soon.
-
Best Go playing engine in the world...
- At Arbitragis Trading, we love technology, mathematics, algorithmics and challenges. Checkers have been solved, chess is now routinely won by computers, but go is not. The number of games in Go is greater than the sum of atoms in the universe, consequently a human being can still easily beat a computer. Inria and Ecole Polytechnique have designed Mogo, the best computer Go game. The savoir-faire that that we generally apply to deal with Quant Finance, is going to be applied to solve a noble cause : the Go game. Our team will try and improve the current algorithm in order to make a much faster Go engine.

