2017 Business School Seminar - December

We are delighted to announce that the 11th Seminar in our business school series will be hosted by Imperial College Business School on 1st December 2017.

We are very grateful to James Sefton and his colleagues for their help in arranging the Seminar.

The conference will consist of 8 presentations of roughly 35 mins each with 10 minutes of Q&A.  There is one modification as a new paper from Tarun Ramadorai using popular machine learning techniques in default estimates is now available (please find updated programme attached).

  • The first presentation by Raman Uppal (EDHEC) uses a parametric portfolio approach to assess which set of combined factor characteristics give the best out-of-sample portfolio performance.  A real innovation in the paper is that also investigates how transaction costs affects the answer.  
  • Another treat for quantitative equity investors is Harrison Hong's presentation on a robust measure of earnings surprise. He argues that a measure based on the percentage of analyst forecasts that underestimate announced earnings outperform other predictors of abnormal returns after these announcements. 
  • Paolo Zaffaroni's paper looks at the econometrically difficult problem of estimating factor risk premium when the number of traded assets is far greater than the number of time periods. In contrast to the literature, he finds strong evidence for significant risk factor risk premia when estimates are correctly adjusted for estimation bias.

Two presentations look at the debt markets.    

  • Tarun Ramadorai estimates a default rate predictor using popular machine learning techniques on a large sample of retail mortgages.  He finds that these newer techniques far outperform more standard econometric models .   As a result there is far larger cross-characteristic (such as race and gender) disparity in default prediction rates; this may be of concern to regulators.  
  • Andrea Buraschi documents that primary dealers are better able to forecast short term rates than non-primary dealers but appear to have no advantage when forecasting longer rates. An implication of this is that long rate changes are driven more by changes in risk premia than by short term rate variation. He then explores how these subjective expectations influence the shape of the yield curve.e

The final three presentations focus on aspects of money management.   

  • Marcin Kacperczyk finds evidence that the higher the proportion of foreign investors, the more efficient the market is. 
  • Tianyu Wang - an Inquire Scholar - explains patterns in fund flows over the economic cycle. He finds evidence that managers do time the cycle effectively and tend to hold low beta portfolios in bad times.  
  • Finally James Sefton looks at the design of the glide path of target dated funds. He argues that that once you consider the outside risk of the investors, these glide paths should be more weighted to bonds than is currently the case in nearly all target dated funds.

This content is only available to INQUIRE members. Please log in to view this content.