Yield Curve Forecast Combinations Based on Bond Portfolio Performance

Journal of Forecasting, Forthcoming

33 Pages Posted: 29 Mar 2017

See all articles by João Caldeira

João Caldeira

Universidade Federal do Rio Grande do Sul (UFRGS)

Guilherme V. Moura

Universidade Federal de Santa Catarina (UFSC) - Department of Economics

Andre A. P. Santos

Universidade Federal de Santa Catarina (UFSC) - Department of Economics

Date Written: March 27, 2017

Abstract

We propose an economically motivated forecast combination strategy in which model weights are related to portfolio returns obtained by a given forecast model. An empirical application based on an optimal mean-variance bond portfolio problem is used to highlight the advantages of the proposed approach with respect to combination methods based on statistical measures of forecast accuracy. We compute average net excess returns, standard deviation, and the Sharpe ratio of bond portfolios obtained with 9 alternative yield curve specifications, as well as with 12 different forecast combination strategies. Return-based forecast combination schemes clearly outperformed approaches based on statistical measures of forecast accuracy in terms of economic criteria. Moreover, return-based approaches that dynamically selects only the model with highest weight each period and discard all other models delivered even better results, evidencing not only the advantages of trimming forecast combinations but also the ability of the proposed approach to detect best performing models. To analyze the robustness of our results, different levels of risk aversion and a different data set are considered.

Keywords: Forecast Combinations, Portfolio Optimization, Yield Curve, Bond Returns

JEL Classification: C53, E47

Suggested Citation

Caldeira, João and Moura, Guilherme Valle and A. P. Santos, Andre, Yield Curve Forecast Combinations Based on Bond Portfolio Performance (March 27, 2017). Journal of Forecasting, Forthcoming. Available at SSRN: https://ssrn.com/abstract=2941616

João Caldeira

Universidade Federal do Rio Grande do Sul (UFRGS) ( email )

Av. Carlos Gomes 1111
Porto Alegre, Rio Grande do Sul 90480-004
Brazil

Guilherme Valle Moura (Contact Author)

Universidade Federal de Santa Catarina (UFSC) - Department of Economics ( email )

PO Box 476
Florianopolis, SC 88010-970
Brazil

Andre A. P. Santos

Universidade Federal de Santa Catarina (UFSC) - Department of Economics ( email )

PO Box 476
Florianopolis, SC 88010-970
Brazil

HOME PAGE: http://sites.google.com/site/andreportela

Register to save articles to
your library

Register

Paper statistics

Downloads
75
Abstract Views
289
rank
315,248
PlumX Metrics