A Dynamic Semiparametric Characteristics-based Model for Optimal Portfolio Selection

Abstract

This paper develops a two-step semiparametric methodology for portfolio weight selection for characteristics-based factor-tilt and factor-timing investment strategies. We build upon the expected utility maximization framework of Brandt (1999) and Aït-sahalia and Brandt (2001). We assume that assets' returns obey a characteristics-based factor model with time-varying factor risk premia as in Li and Linton (2020). We prove under our return-generating assumptions that in a market with a large number of assets, an approximately optimal portfolio can be established using a two-step procedure. The first step finds optimal factor-mimicking subportfolios using a quadratic objective function over linear combinations of characteristics-based factor loadings. The second step dynamically combines these factor-mimicking sub-portfolios based on a time-varying signal, using the investor's expected utility as the objective function. We develop and implement a two-stage semiparametric estimator. We apply it to CRSP (Center for Research in Security Prices) and FRED (Federal Reserve Economic Data) data and find excellent in-sample and out-sample performance that are consistent with investors' risk aversion levels.

A Novel Weighted Hybrid Recommendation System using Sharpe Ratio for a Profitable Diversified Investment Portfolio

“…G. Connor et al [16] and J. Chen et al [17] presented semi-parametric models for the selection of portfolio which is optimal. S. E. Satchell and O. J. Williams [18] have lamented the lack of skills and difficulties in predicting the future of financial markets.…”

Semiparametric Option-Implied Information and Median-Variance Approach: A Game-Changer in Integrating Sustainable Practices in Portfolio Optimization

“…While many studies have explored the use of historical return data to approximate optimal portfolios [2][3][4][5], these approaches often exhibit poor out-of-sample performance and are susceptible to estimation biases [2,4,6,7]. Even forwardlooking approaches struggle to outperform simple heuristic strategies like the naï ve 1/N portfolio, especially in portfolios with a large number of assets [8]. This study is motivated by the persistent underperformance of portfolios constructed using traditional methods, highlighting the need for alternative approaches that can mitigate these challenges.…”

Characteristics-driven returns in equilibrium

“…3 It is interesting to note that a recent trend in the literature expects to get the best of both worlds by resorting to characteristics within factor models, their alphas and their loadings: see Cederburg and O'Doherty (2015), Cosemans et al (2016), Dittmar and Lundblad (2017), Kelly et al (2019), Connor et al (2021), Ge et al (2021, Kim et al (2021a,b) and Windmüller (2021).…”