Linear programing models for portfolio optimization using a benchmark

We consider the problem of constructing a perturbed portfolio by utilizing a benchmark portfolio. We propose two computationally efficient portfolio optimization models, the mean-absolute deviation risk and the Dantzig-type, which can be solved using linear programing. These portfolio models push the existing benchmark toward the efficient frontier through sparse and stable asset selection. We implement these models on two benchmarks, a market index and the equally-weighted portfolio. We carry out an extensive out-of-sample analysis with 11 empirical datasets and simulated data. The proposed portfolios outperform the benchmark portfolio in various performance measures, including the mean return and Sharpe ratio.     

Risk factor contributions in portfolio credit risk models

Determining contributions to overall portfolio risk is an important topic in risk management. For positions (instruments and sub-portfolios), this problem has been well studied, and a significant theory built, around the calculation of marginal contributions. We consider the problem of determining the contributions to portfolio risk of risk factors. This cannot be addressed through an immediate extension of techniques for position contributions, since the portfolio loss is a nonlinear function of the risk factors. We employ the Hoeffding decomposition of the portfolio loss into a sum of terms depending on the factors. This decomposition restores linearity, but includes terms arising from joint effects of groups of factors. These cross-factor terms provide information to risk managers, since they can be viewed as best hedges of the portfolio loss involving instruments of increasing complexity. We illustrate the technique on multi-factor portfolio credit risk models, where systematic factors represent industries, geographical sectors, etc.     

Evaluating portfolio Value-at-Risk using semi-parametric GARCH models

In this paper we examine the usefulness of multivariate semi-parametric GARCH models for evaluating the Value-at-Risk (VaR) of a portfolio with arbitrary weights. We specify and estimate several alternative multivariate GARCH models for daily returns on the S&P 500 and Nasdaq indexes. Examining the within-sample VaRs of a set of given portfolios shows that the semi-parametric model performs uniformly well, while parametric models in several cases have unacceptable failure rates. Interestingly, distributional assumptions appear to have a much larger impact on the performance of the VaR estimates than the particular parametric specification chosen for the GARCH equations.     

A geometric approach to portfolio optimization in models with transaction costs

We consider a continuous-time stochastic optimization problem with infinite horizon, linear dynamics, and cone constraints which includes as a particular case portfolio selection problems under transaction costs for models of stock and currency markets. Using an appropriate geometric formalism we show that the Bellman function is the unique viscosity solution of a HJB equation.Mathematics Subject Classification (1991): 60G44JEL Classification: G13, G11This research was done at Munich University of Technology supported by a Mercator Guest Professorship of the German Science Foundation (Deutsche Forschungsgemeinschaft). The authors also express their thanks to Mark Davis, Steve Shreve, and Michael Taksar for useful discussions concerning the principle of dynamic programming.     

Notes on dynamic factor pricing models

These notes discuss three aspects of dynamic factor pricing (i.e., APT) models. First, the diversifiable component of returns is unpredictable in a no-arbitrage world. Second, conditional factor loadings or betas have an unconditional factor structure when returns follow an unconditional factor structure, which provides a link between conditional and unconditional factor pricing models. Third, the estimation of dynamic factor pricing models is easily simplified in large cross sections when returns follow an unconditional factor structure. These results aid in the interpretation of existing applications and identify some of the issues in the formulation and estimation of dynamic factor pricing models.     

Risk discriminating portfolio optimization

Risk discriminating portfolio optimization provides a risk-related path to performance optimization     

Multi-period stochastic optimization models for dynamic asset allocation

Institutional investors manage their strategic asset mix over time to achieve favorable returns subject to various uncertainties, policy and legal constraints, and other requirements. One may use a multi-period portfolio optimization model in order to determine an optimal asset mix. The concept of scenarios is typically employed for modeling random parameters in a multi-period stochastic programming model, and scenarios are constructed via a tree structure. Recently, an alternative stochastic programming model with simulated paths was proposed by Hibiki [Hibiki, N., 2001b. A hybrid simulation/tree multi-period stochastic programming model for optimal asset allocation. In: Takahashi, H. (Ed.), The Japanese Association of Financial Econometrics and Engineering. JAFEE Journal 89–119 (in Japanese); Hibiki, N., 2003. A hybrid simulation/tree stochastic optimization model for dynamic asset allocation. In: Scherer, B. (Ed.), Asset and Liability Management Tools: A Handbook for Best Practice, Risk Books, pp. 269–294], and it is called a hybrid model. The advantage of the simulated path structure compared to the tree structure is to give a better accuracy to describe uncertainties of asset returns. In this paper, we compare the two types of multi-period stochastic optimization models, and clarify that the hybrid model can evaluate and control risk better than the scenario tree model using some numerical tests. According to the numerical results, an efficient frontier of the hybrid model with the fixed-proportion strategy dominates that of the scenario tree model when we evaluate them on simulated paths. Moreover, optimal solutions of the hybrid model are more appropriate than those of the scenario tree model.     

Early warning systems using dynamic factor models: An application to Asian economies

This study develops an early warning system for financial crises with a focus on small open economies. We contribute to the literature by developing macro-financial dynamic factor models that extract useful information from a rich but unbalanced mixed frequency data set that includes a range of global and domestic economic and financial indicators. The framework is applied to several Asian countries—Thailand, South Korea, Singapore, Malaysia, the Philippines and Indonesia. Logit regression models that use the extracted factors and other leading indicators have significant power in predicting systemic events. In-sample and out-of-sample test results indicate that the extracted factors help to improve the predictive power over a model that uses only sufficiently long history indicators. Importantly, models that include the dynamic factors yield consistently better out-of-sample crisis prediction results for key performance measures such as a usefulness index, the noise to signal ratio, and AUROC.     

Private equity benchmarks and portfolio optimization

Portfolio optimization using private equity is typically based on one of three indices: listed private equity, transaction-based private equity, or appraisal value-based private equity indices. However, we show that none of these indices is fully suitable for portfolio optimization. We introduce here a new benchmark index for venture capital and buyouts, which is updated monthly, adjusted for autocorrelation (de-smoothing), and available contemporaneously. We illustrate how our benchmark enables superior quantitative portfolio optimization.     

An international analysis of REITs and stock portfolio management based on dynamic conditional correlation models

This paper analyzes the volatility spillovers and asymmetry between REITs and stock prices for nine countries (Australia, Belgium, Germany, Italy, Japan, The Netherlands, Singapore, the United Kingdom, and the United States) using eight different multivariate GARCH models. We also analyze the optimal weights, hedging effectiveness, and hedge ratios for REIT-stock portfolio holdings with respect to the results. The empirical results indicate that dynamic conditional correlation (DCC) models provide a better fit than the constant conditional correlation models. The DCC with volatility spillovers and asymmetry (DCC-SA) model provides a better fit than the other multivariate GARCH models. The DCC-SA model also provides the best hedging effectiveness for all pairs of REIT-stock assets. More importantly, this result holds for all cases and for all models that we consider, which means that by taking spillover and asymmetry into consideration, hedging effectiveness can be vastly improved.     

Dynamic portfolio optimization with ambiguity aversion

This paper investigates portfolio selection in the presence of transaction costs and ambiguity about return predictability. By distinguishing between ambiguity aversion to returns and to return predictors, we derive the optimal dynamic trading rule in closed form within the framework of Gârleanu and Pedersen (2013), using the robust optimization method. We characterize its properties and the unique mechanism through which ambiguity aversion impacts the optimal robust strategy. In addition to the two trading principles documented in Gârleanu and Pedersen (2013), our model further implies that the robust strategy aims to reduce the expected loss arising from estimation errors. Ambiguity-averse investors trade toward an aim portfolio that gives less weight to highly volatile return-predicting factors, and loads less on the securities that have large and costly positions in the existing portfolio. Using data on various commodity futures, we show that the robust strategy outperforms the corresponding non-robust strategy in out-of-sample tests.     

Nonparametric correlation models for portfolio allocation

This article proposes time-varying nonparametric and semiparametric estimators of the conditional cross-correlation matrix in the context of portfolio allocation. Simulations results show that the nonparametric and semiparametric models are best in DGPs with substantial variability or structural breaks in correlations. Only when correlations are constant does the parametric DCC model deliver the best outcome. The methodologies are illustrated by evaluating two interesting portfolios. The first portfolio consists of the equity sector SPDRs and the S&P 500, while the second one contains major currencies. Results show the nonparametric model generally dominates the others when evaluating in-sample. However, the semiparametric model is best for out-of-sample analysis.     

Transaction cost optimization for online portfolio selection

To improve existing online portfolio selection strategies in the case of non-zero transaction costs, we propose a novel framework named Transaction Cost Optimization (TCO). The TCO framework incorporates the L1 norm of the difference between two consecutive allocations together with the principles of maximizing expected log return. We further solve the formulation via convex optimization, and obtain two closed-form portfolio update formulas, which follow the same principle as Proportional Portfolio Rebalancing (PPR) in industry. We empirically evaluate the proposed framework using four commonly used data-sets. Although these data-sets do not consider delisted firms and are thus subject to survival bias, empirical evaluations show that the proposed TCO framework may effectively handle reasonable transaction costs and improve existing strategies in the case of non-zero transaction costs.     

Dynamic portfolio optimization across hidden market regimes

Regime-based asset allocation has been shown to add value over rebalancing to static weights and, in particular, reduce potential drawdowns by reacting to changes in market conditions. The predominant approach in previous studies has been to specify in advance a static decision rule for changing the allocation based on the state of financial markets or the economy. In this article, model predictive control (MPC) is used to dynamically optimize a portfolio based on forecasts of the mean and variance of financial returns from a hidden Markov model with time-varying parameters. There are computational advantages to using MPC when estimates of future returns are updated every time a new observation becomes available, since the optimal control actions are reconsidered anyway. MPC outperforms a static decision rule for changing the allocation and realizes both a higher return and a significantly lower risk than a buy-and-hold investment in various major stock market indices. This is after accounting for transaction costs, with a one-day delay in the implementation of allocation changes, and with zero-interest cash as the only alternative to the stock indices. Imposing a trading penalty that reduces the number of trades is found to increase the robustness of the approach.     

Particle swarm optimization approach to portfolio construction

Particle swarm optimization (PSO) is an artificial intelligence technique that can be used to find approximate solutions to extremely difficult or impossible numeric optimization problems. Recently, PSO algorithms have been widely used in solving complex financial optimization problems. This paper presents a PSO approach to solve a portfolio construction problem, since this methodology is a population-based heuristic algorithm that is suitable for solving high-dimensional constrained optimization problems. The computational results show that PSO algorithms have advantages in optimizing the Sortino ratio, especially in speed, when the size of the portfolio is large.     

On the stability of portfolio selection models

One of the main issues in portfolio selection models consists in assessing the effect of the estimation errors of the parameters required by the models on the quality of the selected portfolios. Several studies have been devoted to this topic for the minimum variance and for several other minimum risk models. However, no sensitivity analysis seems to have been reported for the recent popular Risk Parity diversification approach, nor for other portfolio selection models requiring maximum gain–risk ratios.Based on artificial and real-world data, we provide here empirical evidence showing that the Risk Parity model is always the most stable one in all the cases analyzed with respect to the portfolio composition. Furthermore, the minimum risk models are typically more stable than the maximum gain–risk models, with the minimum variance model often being the preferable one. The Risk Parity model seems to be the most stable one also with respect to profitability when measured by the Sharpe ratio. However, the maximum gain–risk models, although quite sensitive to the input data, generally appear to attain better profitability results.     

Forward-looking portfolio selection with multivariate non-Gaussian models

In this study, we suggest a portfolio selection framework based on time series of stock log-returns, option-implied information, and multivariate non-Gaussian processes. We empirically assess a multivariate extension of the normal tempered stable (NTS) model and of the generalized hyperbolic (GH) one by implementing an estimation method that simultaneously calibrates the multivariate time series of log-returns and, for each margin, the univariate observed one-month implied volatility smile. To extract option-implied information, the connection between the historical measure P and the risk-neutral measure Q, needed to price options, is provided by the multivariate Esscher transform. The method is applied to fit a 50-dimensional series of stock returns, to evaluate widely known portfolio risk measures and to perform a forward-looking portfolio selection analysis. The proposed models are able to produce asymmetries, heavy tails, both linear and non-linear dependence and, to calibrate them, there is no need for liquid multivariate derivative quotes.     

On portfolio optimization: Imposing the right constraints

We reassess the recent finding that no established portfolio strategy outperforms the naively diversified portfolio, 1/N, by developing a constrained minimum-variance portfolio strategy on a shrinkage theory based framework. Our results show that our constrained minimum-variance portfolio yields significantly lower out-of-sample variances than many established minimum-variance portfolio strategies. Further, we observe that our portfolio strategy achieves higher Sharpe ratios than 1/N, amounting to an average Sharpe ratio increase of 32.5% across our six empirical datasets. We find that our constrained minimum-variance strategy is the only strategy that achieves the goal of improving the Sharpe ratio of 1/N consistently and significantly. At the same time, our developed portfolio strategy achieves a comparatively low turnover and exhibits no excessive short interest.     

A comparison of static and dynamic portfolio policies

Garleanu and Pedersen (2013) show that the optimal static portfolio policy in light of quadratic transaction costs is a weighted average of the existing portfolio and the target portfolio. In this paper, we demonstrate the importance of the robust target portfolio in the static portfolio policy that considers quadratic transaction costs. By using both empirical and simulated data, we find no evidence that the optimal dynamic portfolio policy proposed by Garleanu and Pedersen (2013) is superior to the static portfolio policy that trades towards the robust target portfolio. The robust target portfolio is achieved by either introducing time-varying covariances or restricting portfolio weights. Furthermore, the static portfolio with time-varying covariances and the short sale-constrained static portfolio are both very efficient in reducing portfolio turnover. The good performance of the static portfolio policy is robust to parameter uncertainty and trading parameters.