Parameter uncertainty in portfolio selection: Shrinking the inverse covariance matrix

Kourtis, Apostolos; Dotsis, George; Markellos, Raphael N.

doi:10.1016/j.jbankfin.2012.05.005

Cited by 74 publications

(67 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Most likely there exists more estimators (see for example Kourtis et al (2012) and Kubokawa & Inoue (2014)) in this class that can be used in a combined rule. Ledoit & Wolf (2014) have also introduced a new class of estimators of non-linear shrinkage of the covariance matrix, that are interesting for future research.…”

Section: Resultsmentioning

confidence: 99%

“…An experiment is presented in Table 2.1 where methodology from Kourtis et al (2012) is mimicked. They remove all parametric errors by letting µ µ µ = k1 1 1 N for some constant k. By doing that, all parameters can be expressed in terms of θ 2 and k. It follows that µ g = k, σ 2 g = k 2 /θ 2 and ψ 2 = 0.…”

Section: Uncertainty In the Global Minimum Variance Portfoliomentioning

confidence: 99%

See 1 more Smart Citation

Optimal Linear Combinations of Portfolios Subject to Estimation Risk

Jonsson

2015

SSRN Journal

View full text Add to dashboard Cite

The combination of two or more portfolio rules is theoretically convex in return-risk space, which provides for a new class of portfolio rules that gives purpose to the Mean-Variance framework out-of-sample. The author investigates the performance loss from estimation risk between the unconstrained Mean-Variance portfolio and the out-of-sample Global Minimum Variance portfolio. A new two-fund rule is developed in a specific class of combined rules, between the equally weighted portfolio and a mean-variance portfolio with the covariance matrix being estimated by linear shrinkage. The study shows that this rule performs well outof-sample when covariance estimation error and bias are balanced. The rule is performing at least as good as its peer group in this class of combined rules. ACKNOWLEDGEMENTS

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Uncertainty In the Global Minimum Variance Portfoliomentioning

confidence: 99%

Optimal Linear Combinations of Portfolios Subject to Estimation Risk

Jonsson

2015

SSRN Journal

View full text Add to dashboard Cite

show abstract

“…Second, I choose the parameter that maximizes the portfolio return of the previous month. This is motivated by the positive autocorrelation of portfolio returns (see, Campbell et al, 1997, DeMiguel et al, 2009b, Kourtis et al, 2012 and may lead to further performance improvements.…”

Section: Results In An Annualmentioning

confidence: 99%

“…This literature has been recently challenged by the work of DeMiguel et al (2009a) who find that 1/N outperforms mean-variance portfolios and most of their extensions. In response to this finding, DeMiguel et al (2009b), Tu and Zhou (2011), Kirby and Ostdiek (2012) and Kourtis et al (2012) develop new sample-based portfolio strategies that offer significantly higher riskadjusted returns than 1/N . I show that the latter portfolio may still be superior in the presence of transaction costs.…”

Section: Introductionmentioning

confidence: 99%

“…This is because most of the sample-based strategies are significantly more unstable than 1/N . Notably, the approaches of DeMiguel et al (2009b), Tu and Zhou (2011) and Kourtis et al (2012) lead to negative Sharpe ratios in several cases once I account for proportional transaction costs of 100 basis points (bp). For example, in a dataset of 8 international market portfolios the strategies of Demiguel et al (2009b) and Kourtis et al (2012) …”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Stable and Efficient Portfolios

Kourtis

2013

SSRN Journal

Self Cite

View full text Add to dashboard Cite

DeMiguel et al. (2009a) show that estimation errors lead to mean-variance portfolios that underperform naive diversification (1/N ). In response to this finding, new studies have developed more efficient sample-based portfolio strategies that offer higher risk-adjusted returns than 1/N . In this paper, I show that the latter portfolio may still be superior in the presence of transaction costs. This is because sample-based strategies tend to be unstable over time. Instability leads to high transaction costs that significantly reduce portfolio returns. I resolve this issue by introducing a stability penalty to the standard mean-variance objective that controls the deviation from the investor's portfolio before rebalancing. In contrast to other techniques in the literature that account for transaction costs, the stability penalty offers intuitive analytical solutions to the portfolio choice problem. As such, it allows to treat estimation errors and transactions costs in a joint manner and it can be directly applied to any portfolio strategy. I evaluate the stability approach by applying it to nine sample-based strategies from the literature and study their performance in five data sets of monthly returns. The results confirm that it leads to stable and efficient portfolios that offer higher Sharpe ratio than 1/N and equal or lower turnover.

show abstract

Understanding the interplay between covariance forecasting factor models and risk‐based portfolio allocations in currency carry trades

Ames

Bagnarosa

Peters

et al. 2018

Journal of Forecasting

View full text Add to dashboard Cite

With the exception of naive methods for portfolio selection, such as the equal weighted approaches, all other methods of portfolio allocation are more or less sensitive to the quality of the inputs considered in constructing the models and risk measures utilized in the allocation framework. The extensively used factor model proposed initially by Sharpe in 1963 has provided a robust backdrop for development of relevant, micro, macro, and context‐specific or asset specific explanatory variables to be incorporated in a statistical manner as inputs to forecasting models that can then be used to obtain risk measures upon which portfolio allocations are based. However, like all statistical models, a set of statistical assumptions accompany this factor model regression framework, one of which has recently been highlighted as seemingly nonvalidated in financial data. This is, of course, the assumption such factor models make on homoskedasticity or weak‐sense covariance stationarity of the returns processes being modeled. Such factor models, therefore, have typically failed to cope with an important and ubiquitous feature of financial assets data, which often demonstrates heteroskedasticity of the returns variances and covariances. We propose a novel generalized multifactor forecasting structure utilizing a covariance regression model, which allows us to incorporate the required heteroskedasticity effects while also admitting potential dependence in the idiosyncratic error terms. We argue that such a modeling approach allows for more explicit relationships to be interpreted between the driving factors and the conditional responses of the portfolio returns. We then compare the forecasting performances of our model with the multifactor model and the time series dynamic conditional correlation model through a currency portfolio application.

show abstract

Parameter uncertainty in portfolio selection: Shrinking the inverse covariance matrix

Cited by 74 publications

References 39 publications

Optimal Linear Combinations of Portfolios Subject to Estimation Risk

Optimal Linear Combinations of Portfolios Subject to Estimation Risk

Stable and Efficient Portfolios

Understanding the interplay between covariance forecasting factor models and risk‐based portfolio allocations in currency carry trades

Contact Info

Product

Resources

About