Dear readers,
Summer time has been very active and stimulating at Alphien as we prepared ourselves for an Alphathon about to be launched in mid-September. We upgraded the Research Portal and built an Investor Portal, while at the same time the Paris team was shifting to a brand new office located at the heart of Paris’ financial district.
BUSINESS
The Alphathon goes live in the next few days. We will come back to you on the launch date with full details on the sponsor’s name and the competition specifications.
COMMUNITY
The Research Portal is getting a fresh design to accommodate for the upcoming Alphathon. Researchers will be able to browse the different tools Alphien offers through more intuitive and logical groupings.
- Research: Here we feature AlphienStudio, our rich integrated development environment (IDE) based on the widely-used RStudio interface, and Notebooks, which have become a standard way to develop models and share results in data science.
- Explore: Researchers can explore the vast and growing raw and augmented data universe available on the platform via Data Explorer and the quantitative library (Qlib) developed by Alphien, giving full transparency on computations.
- Learn & Share: To facilitate collaborations and provide in-depth explanations, we have the Wiki (thorough technical documentation), AlphaChat (collaboration and communication tool) and tutorials.
- Monitor & Compete: This section offers a way to participate in Alphathons, access dashboards (a monitor of researchers’ activities and their strategy performances) as well as access to the leaderboard where researchers can compare how they perform against their peers.
Alphien’s quantitative development team has also been working hard in order to integrate new functionalities to the platform.
A framework to allow users to group single-underlying strategies has been developed. The new process integrates nicely within Alphien’s existing backtesting engine, such that users can keep on developing single strategies using algoEngine, the backbone of Alphien’s backtesting environment, and use the function to find the optimal way to mix them together as a portfolio. This functionality will soon be released to investors as well so that they can optimise their allocations between strategies subject to their diversification needs.
We are also adding new data science capabilities to the platform. We have implemented different ways to cluster financial contracts, proposing alternatives to traditionally-used cross correlations and offering a fresh perspective to researchers when it comes to diversification. To give an overview of the technical details, researchers will be able to use different metrics to measure distances between financial contracts and to look for similarities between their time series based on these metrics. Such features can be used to select a basket of securities in order to diversify portfolios, or to identify correct pairs to design pairs-trading strategies for instance. We plan to release a research notebook about these new clustering techniques in the coming weeks and will be made publicly accessible through our blog.