Abstract
We compare various methodologies to estimate the covariance matrix in a fixed-income portfolio. Adopting a statistical approach for the robust estimation of the covariance matrix, we compared the Shrinkage (SH), the Nonlinear Shrinkage (NSH), the Minimum Covariance Determinant (MCD) and the Minimum Regularised Covariance Determinant (MRCD) estimators against the sample covariance matrix, here employed as a benchmark. The comparison was run in an application aimed at individuating the principal components of the US term structure curve. The contribution of the work mainly resides in the fact that we give a freshly new application of the MRCD and the NS robust covariance estimators within the fixed-income framework. Results confirm that, likewise financial portfolios, also fixed-income portfolios can benefit of using robust statistical methodologies for the estimation of the covariance matrix.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
As explained in Boudt et al. (2017), S* is a transformation of the sample covariance matrix SU(H), obtained as: \( S^{*} = \Lambda ^{1/2} Q^{{\prime }} S_{U} \left( H \right)Q\Lambda ^{1/2} \).
- 2.
Principal Component Analysis (Hotelling 1933) is a dimension reduction technique that works on a covariance (or correlation) matrix identifying the volatility factors that drive the time series under investigation. The PCA relies on the spectral decomposition of the covariance matrix \( \Sigma \):
$$ \Sigma = {\text{G}}\Omega {\text{G}}^{{\prime }} $$where G is the square matrix of the eigenvalues of \( \Sigma \), and \( \Omega \) is a diagonal matrix filled with the eigenvalues of the covariance matrix. The principal components are given by the normalized eigenvectors, ranked in descendant order according to the size of related eigenvalues. This because the total variance is equal to the sum of all the eigenvalues, so that the size of a single eigenvalues is the percentage of total variance explained. As a limited number of eigenvalues is usually enough to explain at least the 99% of total variation, the reduction of the covariance matrix can be performed by retaining only the eigenvalues that explain a certain threshold of the variance, eliminating the others.
- 3.
The sample and MCD methodologies are estimated using MATLAB embedded functions. The Shrinkage code is from Ledoit and Wolf (2004b), while the Nonlinear Shrinkage is estimated with the nlshrink R package, https://rdrr.io/cran/nlshrink/man/nlshrink.html, and the MRCD is from the KU Leuven webpage, https://wis.kuleuven.be/stat/robust/Programs/MRCD.
References
Agulló, J., Croux, C., & Van Aelst, S. (2008). The multivariate least-trimmed squares estimator. Journal of Multivariate Analysis, 99(3), 311–338.
Anderberg, M. R. (1973). Cluster analysis for applications. Monographs and textbooks on probability and mathematical statistics.
Anderson, T. W. (1963). Asymptotic theory for principal component analysis. The Annals of Mathematical Statistics, 34(1), 122–148.
Bengtsson, C., & Holst, J. (2002). On portfolio selection: Improved covariance matrix estimation for Swedish asset returns. Univ.
Black, F., & Litterman, R. (1992). Global portfolio optimization. Financial Analysts Journal, 48(5), 28–43.
Boudt, K., Rousseeuw, P., Vanduffel, S., & Verdonck, T. (2017). The minimum regularized covariance determinant estimator.
Candela, J. A. (1996). Exact iterative computation of the multivariate minimum volume ellipsoid estimator with a branch and bound algorithm. In COMPSTAT (pp. 175–180). Springer.
Chen, Y., Wiesel, A., Eldar, Y. C., & Hero, A. O. (2010). Shrinkage algorithms for MMSE covariance estimation. IEEE Transactions on Signal Processing, 58(10), 5016–5029.
Croux, C., & Haesbroeck, G. (1999). Influence function and efficiency of the minimum covariance determinant scatter matrix estimator. Journal of Multivariate Analysis, 71(2), 161–190.
Fan, J., Peng, H., Huang, T., & Ren, Y. (2005). Theory and methods-rejoinder-semilinear high-dimensional model for normalization of microarray data: A theoretical analysis and partial consistency. Journal of the American Statistical Association, 100(471), 808–813.
Fisher, T. J., & Sun, X. (2011). Improved Stein-type shrinkage estimators for the high-dimensional multivariate normal covariance matrix. Computational Statistics & Data Analysis, 55(5), 1909–1918.
Frahm, G., & Jaekel, U. (2005). Random matrix theory and robust covariance matrix estimation for financial data. arXiv preprint physics/0503007.
Hawkins, D. M., & McLachlan, G. J. (1997). High-breakdown linear discriminant analysis. Journal of the American Statistical Association, 92(437), 136–143.
Hotelling, H. (1933). Analysis of a complex of statistical variables into principal components. Journal of Educational Psychology, 24(6), 417.
Hubert, M., Rousseeuw, P. J., & Vanden Branden, K. (2005). ROBPCA: A new approach to robust principal component analysis. Technometrics, 47(1), 64–79.
Ledoit, O., & Péché, S. (2011). Eigenvectors of some large sample covariance matrix ensembles. Probability Theory and Related Fields, 151(1), 233–264.
Ledoit, O., & Wolf, M. (2003). Improved estimation of the covariance matrix of stock returns with an application to portfolio selection. Journal of Empirical Finance, 10(5), 603–621.
Ledoit, O., & Wolf, M. (2004a). A well-conditioned estimator for large-dimensional covariance matrices. Journal of Multivariate Analysis, 88(2), 365–411.
Ledoit, O., & Wolf, M. (2004b). Honey, I shrunk the sample covariance matrix. The Journal of Portfolio Management, 30(4), 110–119.
Ledoit, O., & Wolf, M. (2012). Nonlinear shrinkage estimation of large-dimensional covariance matrices. The Annals of Statistics, 40(2), 1024–1060.
Lintner, J. (1965). The valuation of risk assets and the selection of risky investments in stock portfolios and capital budgets. The Review of Economics and Statistics, 13–37.
Litterman, R. B., & Scheinkman, J. (1991). Common factors affecting bond returns. The Journal of Fixed Income, 1(1), 54–61.
Markowitz, H. (1952). Portfolio selection. The Journal of Finance, 7(1), 77–91.
Martellini, L., Priaulet, P., & Priaulet, S. (2003). Fixed-income securities: Valuation, risk management and portfolio strategies (Vol. 237). New York: Wiley.
Mehta, M. L. (1990). ‘Random matrices.’ Academic Press, 2nd edn.
Michaud, R. O. (1989). The Markowitz optimization enigma: Is ‘optimized’ optimal? Financial Analysts Journal, 45(1), 31–42.
Michaud, R. O., & Michaud, R. O. (1998). Efficient asset management: A practical guide to stock portfolio optimization and asset allocation. Oxford University Press.
Jagannathan, R., & Ma, T. (2003). Risk reduction in large portfolios: Why imposing the wrong constraints helps. The Journal of Finance, 58(4), 1651–1683.
Jorion, P. (1986). Bayes-Stein estimation for portfolio analysis. Journal of Financial and Quantitative Analysis, 21(3), 279–292.
Pafka, S., Potters, M., & Kondor, I. (2004). Exponential weighting and random-matrix-theory-based filtering of financial covariance matrices for portfolio optimization. arXiv preprint cond-mat/0402573.
Pantaleo, E., Tumminello, M., Lillo, F., & Mantegna, R. N. (2011). When do improved covariance matrix estimators enhance portfolio optimization? An empirical comparative study of nine estimators. Quantitative Finance, 11(7), 1067–1080.
Pison, G., Rousseeuw, P. J., Filzmoser, P., & Croux, C. (2003). Robust factor analysis. Journal of Multivariate Analysis, 84(1), 145–172.
Ross, S. A. (1976). The arbitrage theory of capital asset pricing. Journal of Economic Theory, 13(3), 341–360.
Rousseeuw, P. J. (1984). Least median of squares regression. Journal of the American Statistical Association, 79(388), 871–880.
Rousseeuw, P., & Yohai, V. (1984). Robust regression by means of S-estimators. In Robust and nonlinear time series analysis (pp. 256–272). Springer, New York.
Rousseeuw, P. J., & Driessen, K. V. (1999). A fast algorithm for the minimum covariance determinant estimator. Technometrics, 41(3), 212–223.
Rousseeuw, P. J., Van Aelst, S., Van Driessen, K., & Gulló, J. A. (2004). Robust multivariate regression. Technometrics, 46(3), 293–305.
Schäfer, J., & Strimmer, K. (2005). A shrinkage approach to large-scale covariance matrix estimation and implications for functional genomics. Statistical Applications in Genetics and Molecular Biology, 4(1).
Schyns, M. (2008). Robust portfolio selection. JSM Proceedings, Statistical Computing Section.
Sharpe, W. F. (1964). Capital asset prices: A theory of market equilibrium under conditions of risk. The Journal of Finance, 19(3), 425–442.
Stein, C. (1956, January). Inadmissibility of the usual estimator for the mean of a multivariate normal distribution. In Proceedings of the Third Berkeley Symposium on Mathematical Statistics and Probability (Vol. 1, No. 399, pp. 197–206).
Tola, V., Lillo, F., Gallegati, M., & Mantegna, R. N. (2008). Cluster analysis for portfolio optimization. Journal of Economic Dynamics and Control, 32(1), 235–258.
Wolf, M. (2004). Resampling vs. shrinkage for benchmarked managers.
Woodruff, D. L., & Rocke, D. M. (1994). Computable robust estimation of multivariate location and shape in high dimension using compound estimators. Journal of the American Statistical Association, 89(427), 888–896.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG, part of Springer Nature
About this chapter
Cite this chapter
Neffelli, M., Resta, M. (2018). A Comparison of Estimation Techniques for the Covariance Matrix in a Fixed-Income Framework. In: Mili, M., Samaniego Medina, R., di Pietro, F. (eds) New Methods in Fixed Income Modeling. Contributions to Management Science. Springer, Cham. https://doi.org/10.1007/978-3-319-95285-7_6
Download citation
DOI: https://doi.org/10.1007/978-3-319-95285-7_6
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-95284-0
Online ISBN: 978-3-319-95285-7
eBook Packages: Economics and FinanceEconomics and Finance (R0)