Below is an abstract from the paper published by the prestigious Finance Research Letters journal in June 2023: ¨Kurtosis-based vs volatility-based asset allocation strategies: Do they share the same properties? A first empirical investigation – ScienceDirect¨
M. D. Braga, C. R. Nava & M. G. Zoia
About the authors
Maria Debora Braga is Full Professor of Financial Markets and Institutions at SDA Bocconi University in Milan. M.D.Braga is member of the EFPA Italy Standards and Qualifications Committee and executive member of the same committee at EFPA Europe.
Consuelo R. Nava: Department of Economics and Statistics “Cognetti de Martiis”, University of Turin, Turin, Italy
Maria G. Zoia: Department of Economic Policy, Università Cattolica del Sacro Cuore, Milan, Italy
Abstract
Abstract
Using a sample of international equity markets over the period 2001–2020, this paper aims to empirically investigate the implications in terms of asset allocation and the key properties of kurtosis-based strategies compared to the more traditional volatility-based strategies for financial portfolios construction. Furthermore, the contribution demonstrates that the portfolio recommended by the novel Kurtosis-based Risk Parity strategy introduced by Braga et al. (2023) admits to being interpreted as an intermediate portfolio between the Minimum-Kurtosis portfolio and the Equally Weighted portfolio in terms of the fourth root of the portfolio fourth moment.
Introduction
It is well known that Risk Parity strategy, formally elaborated first by Maillard et al. (2010) and Roncalli (2013), has gained popularity since 2008 as an asset allocation approach capable of identifying portfolio exposures to asset classes in such a way that the contribution to portfolio volatility is the same for all portfolio components. Recently, Braga et al. (2023) have significantly innovated the risk parity framework by replacing the most classical reference risk measure with the portfolio fourth-order moment which represents an approximation of portfolio kurtosis. The new Kurtosis-based Risk Parity (KRP, in short, or Portfolio Fourth Moment Risk Parity) achieves the goal of the homogeneous distribution among asset classes of the fourth root of the portfolio-fourth moment since it admits the application of Euler’s theorem. The advent of the traditional Risk Parity (SRP strategy) led to investigate its properties compared to those of other asset allocation strategies which, likewise, do not make use of expected return estimates as inputs. They include the Equally Weighted (EW) strategy and the Minimum Variance Strategy (MV strategy) aimed, respectively, at recommending an equal exposure to each asset class belonging to a predefined investment universe and at selecting the portfolio entailing the lowest possible variance (volatility). Maillard et al. (2010) and Demey et al. (2010) have shown the existence of a natural order for the portfolio volatility characterizing these different approaches: the MV strategy is the least volatile, the EW is the most volatile and in an intermediate position lies the SRP strategy. They also show an “equalization effect” of a specific feature for asset classes with a non-zero weight: equal weights in the EW portfolio, equal marginal risks in the MV portfolio and equal risk contributions in the SRP portfolio.
As far as known, this is the first paper that empirically investigates the validity or not of the same natural sequence of portfolios and of the same “equalization effect” for the kurtosis-based asset allocation strategies. In the following: Section 2 presents the methodological aspects, Section 3 develops the empirical investigation and Section 4 concludes.
Section snippets
Empirical investigation
In this section the portfolio construction process is performed considering both approaches based on homogeneous risk distribution (the novel KRP strategy and the traditional SRP strategy) and approaches based on risk minimization (the MK and MV strategy). The EW strategy is added since the simple “1/N” allocation is commonly suggested as a reference portfolio (Michaud, 1989, DeMiguel et al., 2009).
For the empirical investigation, the euro denominated total returns, both on a monthly and weekly
Conclusion
Starting from the well-known result provided by Maillard et al. (2010) who show that portfolio volatility is sorted in ascending order moving from the MV strategy to the SRP and ending with the EW, this paper is the first to integrate the existing literature by extending, mainly from an empirical perspective, a similar “natural” sequence to kurtosis-based risk strategies represented by the novel KRP strategy developed by Braga et al. (2023) and the MK strategy where the fourth root of the
CRediT authorship contribution statement
Maria Debora Braga: Conceptualization, Methodology, Validation, Resources, Writing. Consuelo Rubina Nava: Software, Data curation, Visualization. Maria Grazia Zoia: Conceptualization, Methodology, Writing.
To access the full paper: