WorldQuant’s Triple-Axis Plan for Alpha Diversification: From video games to hedge funds
Mr. Tulchinsky is the Founder, Chairman, and CEO of WorldQuant. Formerly a statistical arbitrage portfolio manager, venture capitalist, and scientist, Mr. Igor Tulchinsky was once a video game programmer creating rudimentary games reminiscent of Pacman.
Mr. Tulchinsky, who completed his MA in Computer Science is no more than 9 months at the University of Texas, is known for his mathematical and systematic approach to investing at WorldQuant.
WorldQuant preaches a Triple-Axis Plan (TAP) for systematic research. The three axes for a TAP are the following:
- Ideas and the dataset axis;
- The regions and the universe axis and
- The performance parameter axis.
The Ideas and The Dataset Axis
The first axis is the ideas and the dataset axis. There are plenty of ideas to create alphas, ranging from mean-reversion, momentum, and trend-following to value, growth and min-vol.
Igor Tulchinsky et al. launched “The 101 Alphas Project” with a paper entitled “101 Formulaic Alphas”, which spells out pseudo-code for 101 “alphas”, volume and price based factors that can be computed across a large universe of stocks.
101 Formulaic Alphas by Zura Kakushadze :: SSRN
We present explicit formulas - that are also computer code - for 101 real-life quantitative trading alphas. Their…
There are also many datasets available. The most common ones being core financial data such as the price volume data set, the company fundamentals and the analyst recommendations. More recently, the emergence of Alternative Data has opened up new possibilities with the availability of non-traditional datasets generated by individuals, businesses, and sensors that can be used as a source of ideas and alpha.
Wall Street’s War on Alternative Data: Market Trends, Innovation and Sales Insights
The Market, The Players, and The Rules. What you need to know about today’s emergence of Alternative Data on Wall…
The Regions and The Universe Axis
The second axis is the regions and the universe. This axis defines the strategies’ tradable universe. The most commonly used regions are the US and Europe and Asia is gaining a lot of traction recently as well. In this axis, we should define a subset of stocks. Criteria may include:
- Country and Regions
- Industry and Sectors
- Regulatory and Compliance rules
- Mandates (e.g. risk exposures and allocations)
Any group of selections depends on the idea that the researcher is trying to execute or criteria imposed by mandates, regulatory or compliance rules.
For country/regions filtering, a common practice is to use an underlying benchmark index that provides a standard and transparent definition of a liquid and tradable universe of securities from a target country/region. For industry/sectors, a common approach is to use benchmarks such as GICS or ICB, which provide widely recognizable industry classification systems.
Quantopian provides a free option to get you started with access to a stable, robust stock universe as the foundation for your algorithm.
Quantopian is a free online platform for education and creation of investment algorithms. Selected algorithms get…
The Performance Parameter Axis
The third axis is the performance parameter. The performance parameters are being seen as the ways of analyzing the performance of the Alphas, but can also be interpreted as a source of alpha ideas as well. It could be the Sharp of the Alphas, the returns of the Alphas, or drawdowns. The definition of the performance parameters part of the target objective function is key when sourcing alpha ideas.
Axioma is one example of a commonly used best-of-breed solution for portfolio analytics and portfolio optimization. Its Portfolio Optimizer models a wide range of hedge fund strategies, including long/short equity, market neutral, event-driven, global macro and fund of funds with great flexibility.
Axioma | Products / Axioma Portfolio / Axioma Portfolio Optimizer / Hedge Fund Manager
Model a wide range of hedge fund strategies, including long/short equity, market neutral, event-driven, global macro…
WorldQuant’s Triple-Axis Plan provides a simple-to-use yet robust approach for systematic investing. While machines are powerful like never before, humans still need to define the right objective function and data.