An Eye on the Residuals: MPI Presents to CFA Societies on Dynamic Factor Analysis

In conjunction with the CFA Societies of the UK, France and Germany, MPI recently presented at sold-out events in London, Paris and Frankfurt in an effort to address best practice quantitative methodologies for dynamic factor analysis and their applications for the wealth and investment management industry.

The presentation, entitled “Hidden sources of alpha: from market timing to insider trading”, asked, “What portion of a manager’s returns is due to luck or skill, leverage, excessive risks, or clever and timely market bets and use of derivatives? Is there any indication of wrongdoing? How can we find the answers, especially when a portfolio’s positions are not available or suspect?”

Looking to answer these questions with the predictive powers of MPI’s proprietary method for dynamic factor analysis, Dynamic Style Analysis (DSA) – via case studies of the Flash Crash and Galleon Group – MPI’s CEO and Director of Research, Michael Markov, counseled practitioners to “focus on the residuals” when analyzing and monitoring funds and investment portfolios.

Transcending the limitations imposed by traditional returns-based style analysis (RBSA), DSA empowers users to capture active portfolio dynamics, such as rapid strategy shifts and leverage, and utilize multiple return sources to precisely measure the predictability of an investment. DSA was patented in 2009 and is available in MPI’s flagship product, Stylus Pro.

To see the presentation, Contact Us.

  • Paolo

    Hi, i find your quant-platform very appealing but i don’t understand what kind of methodology you apply: do you use a constrained kalman filter?

    thank you in advance for your response


    • admin

      A key element of MPI’s Stylus software and analytical engine is based on our proprietary, patented technique, Dynamic Style Analysis (DSA). DSA inherits insights from returns-based style analysis (RBSA), but goes well beyond RBSA by creating a much more general setting that is free from the limitations of regression. In short, instead of trying to find the best local fit, it searches for a dynamic trading strategy that has the highest ability to trade through missing data with the least possible turnover. DSA, like Kalman filters, is dynamic in nature but it differs in two significant respects; it makes no assumptions about the distribution of errors and supports financially motivated constraints transparently.

Leave a Reply
Fields to complete: