RT Journal Article SR Electronic T1 Quintet Volume Projection JF The Journal of Trading FD Institutional Investor Journals SP 28 OP 43 DO 10.3905/jot.2017.12.2.028 VO 12 IS 2 A1 Vladimir Markov A1 Olga Vilenskaia A1 Vlad Rashkovich YR 2017 UL https://pm-research.com/content/12/2/28.abstract AB We present a set of models that are relevant for predicting various aspects of intraday trading volume for equities and showcase them as an ensemble that projects volume in unison. We introduce econometric methods for predicting end-of-day volume, volume u-curve, close auction volume, and special day seasonalities and emphasize a need for a unified approach in which all submodels work consistently with each other. We rely on the application of Bayesian methods, which have the advantage of providing adaptive and parameterless estimations of volume for a broad range of equities while automatically taking into account uncertainty in the model input components. We discuss the shortcomings of traditional statistical error metrics for calibrating volume prediction and introduce asymmetrical logarithmic error to overweight an overestimation risk.TOPICS: Security analysis and valuation, statistical methods