 Research
 Open Access
 Published:
Research on regularized mean–variance portfolio selection strategy with modified Roy safetyfirst principle
SpringerPlus volume 5, Article number: 919 (2016)
Abstract
We propose a consolidated risk measure based on variance and the safetyfirst principle in a meanrisk portfolio optimization framework. The safetyfirst principle to financial portfolio selection strategy is modified and improved. Our proposed models are subjected to norm regularization to seek nearoptimal stable and sparse portfolios. We compare the cumulative wealth of our preferred proposed model to a benchmark, S&P 500 index for the same period. Our proposed portfolio strategies have better outofsample performance than the selected alternative portfolio rules in literature and control the downside risk of the portfolio returns.
Background
The benefits of portfolio optimization are well known to investors and other stakeholders. Strategies in portfolio selection have been well documented and unquestionably explored in finance and other related areas from a theoretical and much more practical perspective. Undoubtedly, a widely used approach is the mean–variance model by Markowitz (1952). He demonstrates how investors can employ riskreturn tradeoff for wealth allocation using utility functions. Investors decide where along an efficient frontier a suitable balance between risk and return exist. Markowitz’s work has become the bedrock of portfolio selection analysis, thereby earning the name “Modern Portfolio Theory”. In a related study, Roy (1952) proposed a safetyfirst principle, which minimizes the shortfall probability in portfolio selection. For any investor whose general interest is downside risk measure of investment, Roy’s (1952) safetyfirst principle is much more suitable Chiu et al. (2012).
Under the safetyfirst principle, instead of employing utility functions, which is hard or even impossible to determine, a known amount of the principal is preserved. Thus, an investor predefines a minimum threshold level of return and opts for a portfolio of assets that attains this preservation of principal. Roy’s principle has over the years been studied. For instance, Bawa (1978) used the principle to maximize income, given the probability that a predefined threshold is greater than income. Haque et al. (2007) also applied safetyfirst through extreme value theory to portfolios of Mexican and U.S equities. Rachev (2001) showed how the safetyfirst approach can be more efficient than the stable Paretian approach in portfolio theory.
Levy and Sarnat (1972) and Nawrocki (1999) show how safetyfirst principle and mean–variance approach are akin. Levy and Sarnat (1972) found out that in the distinctive case where target rate of return is equal to riskfree return, both Roy’s safetyfirst principle and Markowitz’s mean–variance approach lead to the same optimal portfolio selection strategy. However, both models have several restrictions which include but are not limited to the following: Firstly, both Markowitz and Roy employ symmetry about the mean variance as a risk measure. Variance takes into account the situation in which return exceeds mean value and this does not impact risk. Secondly, transaction costs are absent in safetyfirst principle and mean–variance approach. According to DeMiguel et al. (2014), their loss in the mean–variance approach with the absence of transaction costs was 49.33 %. Borkovec et al. (2010) pointed out that 40 % of financial market participants attribute the fundamental loss in abnormal return to transaction costs. Lastly, both model frameworks are subjected to parameter uncertainty as returns of assets are considered as deterministic parameters being depicted by a single point estimate which leads to estimation risk.
Variance, as a risk measure, penalizes both underperformance and overperformance equally (Markowitz 1968). However, investors are only worried about underperformance. Roman et al. (2007) obtained a better portfolio by employing three indexes i.e. mean–varianceCVaR in comparison to mean–variance and meanCVaR. The authors demonstrated that the meanCVaR portfolio policy results in large variance, which leads to a small Sharpe ratio. Again, the CVaR of the portfolio generating from mean–variance model is large. To eliminate these inconsistencies between the strategies of mean–variance and meandownside risk models, Roman et al. (2007) proposed to merge CVaR and variance in a multiobjective portfolio selection strategy. Inspired by Roman et al. (2007)’s approach of combining a downside risk measure, CVaR, and variance, we consider a portfolio optimization model with multiple risk measures. Recognizing a need to modify and improve Roy’s safety first principle, we consider merging our improved Roysafety first approach with variance as the portfolio risk measure.
The construction of a portfolio of investments is a significant problem faced by investors and institutions. A decision ought to be made to allocate weights to each investment with the intention of striking an appropriate balance between returns and risk. In reality, setting up a new portfolio or rebalancing an existing one requires costs to be incurred and must be inclusive in any realistic analysis. We incorporate proportional transaction costs (Kellerer et al. 2000; Muthuraman and Kumar 2006) which are induced by liquidity costs, tax and brokerage fees (Dumas and Luciano 1991; Kellerer et al. 2000; Lobo et al. 2007) into our portfolio selection model.
In the mean–variance and Roy safetyfirst models, stock returns are considered deterministic and taken as a single point estimate which results in estimation risk or overfitting (Bawa et al. 1979; Merton 1980). A small variation in the input parameters of the standard Markowitz mean–variance approach and Roy safetyfirst principle usually lead to changes in the structure of the resulting portfolios (Brandt 2009). To reduce the undesired impact of estimation risk or overfitting, models in the context of robustness (Goldfarb and Iyengar 2003; Tütüncü and Koenig 2004), stochastic programming (Rockafellar and Uryasev 2000), factor models (Green and Hollifield 1992; Nagai 2003; Schultz and Tiedemann 2003) and shrinkage estimators (Jorion 1986; Ledoit and Wolf 2004) have been explored. Another method analogous to our work is the modification of portfolio weights by adding regularizers or additional constraints to the portfolio strategy (Jagannathan and Ma 2003; DeMiguel et al. 2009; Brodie et al. 2009). Correcting undesired portfolio weights are as a result of large estimation risk of unknown parameters leading to the achievement of their desired forms and characteristics. Jagannathan and Ma (2003) employed a short sale constraint in the minimum variance framework and found out the nonnegativity constraint performs well as those computed with shrinkage estimators and factor models. In DeMiguel et al. (2009), authors added a convex norm ball constraint to the portfolio weight and observed that the norm ball constrained portfolios had better out of sample performance than the portfolio strategies of the naive 1/N diversification (Jagannathan and Ma 2003). The \(\textit{l}_2\)norm constrained portfolio in general attained higher Sharpe Ratio than \(\textit{l}_1\)norm constrained portfolios. We study the weight constrained portfolios by specifying the general norm as squared \(\textit{l}_2\)norm ball.
The optimal portfolio of classical Roy safetyfirst principle and Markowitz’s model in the presence or absence of short sale constraint or stability constraints hold a large number of assets and especially small weights assigned to their proportions. However, holding a large number of assets leads to the investor incurring high transaction costs. Due to this factor and other market and economic frictions, investors often hold only a small number of stocks in their portfolio. This is usually known as a sparse portfolio (a portfolio with few nonzero weights). With sparsity, one presets limit of assets (stocks) with nonzero entries of portfolio allocations.
A method of constructing sparse portfolio termed hard threshold strategy was proposed by BrittenJones (1999) to statistically test each portfolio weight with a null hypothesis that weight is zero. Using Ftest and ttest, BrittenJones (1999) proposed that if the portfolio weights are statistically not different from zero; assigning zero reduces portfolio risk. Sparse portfolio weights can be derived by cardinality constrained portfolio optimization (CCPO) problem (Chang et al. 2000; Maringer and Kellerer 2003; RuizTorrubiano and Suárez 2015). CCPO takes into account all portfolios of given number of assets and chooses an optimal portfolio. With the inclusion of cardinality constraint, the portfolio selection strategy becomes NPhard (MoralEscudero et al. 2006), and standard quadratic program solvers can no longer be adopted to tackle the problem. One resorts to several other methods or relaxations in search of nearoptimal solutions at a moderate computational cost. Farrell and Reinhart (1997) suggested classification of assets based on geographical aspect, size, sector, etc. and made a selection of N assets from each of these classes. In Pai and Michel (2009), the cardinality constraints were handled via clustering algorithm to reduce the size of the portfolio. More recently, RuizTorrubiano and Suárez (2015) used a memetic approach that combines a genetic algorithm (GA) with an extended set encoding and quadratic programming (QP) in a mean–variance framework to deal with the cardinality constraint. An alternative relaxation method for which we employ in this paper is by constraining the \(l_1\)norm (Tibshirani 1996) which is also in connection with the upper bound of the estimation risk as shown in Fan et al. (2012). The convexity nature of this type of regularization makes it more tractable. The use of norm penalty helps investors limit transaction costs and exposure to risky stocks.
As Roy safetyfirst principle minimizes the chances that the portfolio’s return will fall below the minimum acceptable return, introducing it in the mean–variance model helps control the downside risk of the portfolio return. The modification and improvement of Roy safetyfirst principle and merging it with variance as a consolidated risk measure in a riskreturn framework represents the main novelty of this research paper. In achieving our aim, we minimize our modified and improved Roy safetyfirst principle and impose the lower constraint on the mean return of the portfolio as in Markowitz’s mean–variance model. We also investigate the portfolio strategy with variance and Roy’s safety first principle as a consolidated risk measure in a meanrisk framework. To address the problem of estimation risk, we constrain the portfolio weights with squared \(l_2\)norm and proceed to achieve sparsity via \(l_1\)norm heuristic. We will explore the impact of transaction costs on portfolio selection strategies.
The paper is organized as follows. The next section presents our proposed portfolio selection strategies. It’s subsections study Roy safetyfirst principle and its modification, portfolio revision and stable and sparse portfolio. We perform numerical tests and present computational results of our proposed methods in the subsequent section. Concluding remarks are provided in the last section.
Proposed portfolio selection strategies
Investors allocate proportions of their capital among the assets they invest in. For the purpose of this study, these proportions are allocated to stocks. We denote by N the number of risky assets, R is the required return level, and \(x^0\) is the initial risky assets before rebalancing: \(x^0_k\) is the proportion of capital initially allocated to asset \(k, k=1,2,3,...,N\). Let \(x,x^b\) and \(x^s\) be N dimensional vectors of controllable variables: \(x_k\) is the portfolio invested in risky asset k after rebalancing, \(x^b_k\) is the purchases (proportion used) of risky asset k and \(x^s_k\) is the sales (proportion obtained) of risky asset k. The transaction costs incurred when buying risky assets is \(c^b\) and that of selling risky assets is \(c^s\). The financial portfolio is described by a Ndimensional vector of random returns r. The portfolio total random return is \(R_p=f(x,r)=\sum ^N_{k=1}x_k r_k\), portfolio expected return vector \(\mu _p=\mathbb {E}_x f(x,r)=\sum ^N_{k=1}\mathbb {E}_x r_k\), \(\sigma _p^2=\mathbb {E}_x(R_p\mathbb {E}_x(R_p))^2\) as variance of portfolio return and Q as the variancecovariance matrix of the portfolio return.
Roy safetyfirst principle and it’s modification
Most investors’ aim is to maximize returns and minimize risk. The Roy safetyfirst principle advocates avoiding extreme losses through the minimization of disaster probability. To optimally construct a portfolio strategy, Roy’s safetyfirst principle defines a threshold or a minimum acceptable return R, below which the portfolio wealth is considered to be a disaster. The best portfolio is one that minimizes the chances that the portfolio’s return, \(R_p\), will fall below a minimum acceptable return, R. In essence, an investor selects his portfolio by solving this optimization problem:
where e is a vector with ones as entries and \(\mathbb {P}\) is a probability measure. Roy employed Bienayme–Tchebycheff's inequality as the investor is likely not to know the actual probability function and obtained an approximation
Thus, the optimization problem is reformulated as
The modification to the Roy’s approach we adopt is by using a coherent downside risk measure known as Conditional ValueatRisk or Expected shortfall as it has a set of desirable properties for a risk measure (Platen and Heath 2006) leading to more accurate estimates of probability. For a detailed study on desirable properties of an ideal risk measure in portfolio theory, we refer the reader to Rachev et al. (2008).
A wellknown downside risk measure known as ValueatRisk focuses on the percentiles of loss distributions and measures the predicted maximum loss at a given probability level. Mathematically it is formulated as \(\alpha\)quantile \(VaR_\alpha (X)= min \{z\mid (F_X(z)\ge \alpha \}\), where X is a loss random variable and \(\alpha \in (0,1)\) is the given probability level. Values for \(\alpha\) often used are 90 %, 95 % and 99 %. Considering ValueatRisk (VaR) has undesirable properties such as nonsubadditive and nonsmooth etc., Rockafellar and Uryasev (2000) introduced a coherent downside risk measure termed Conditional Valueat Risk (CVaR) and for \(\alpha \in (0,1)\) represented it as
where
Equivalently, for \(x \in X \subseteq \mathbb {R}^N\) and random vector \(r\in \mathbb {R}^N\) which represents the actual portfolio return has a continuous density function p(r)
First, we will determine the semideviation of the random return f(x, r) from the \(\alpha\)quantile \(VaR_\alpha (x)\). With respect to Bienayme–Tchebycheff's inequality, the following estimate is valid for \(VaR_{\alpha )}(x)>R\):
Let us consider an \(\alpha\)quantile
and a measure of risk
which is termed expected shortfall from the \(\alpha\)quantile \(VaR_\alpha (x)\) value. The estimate (3) can be written as
Therefore (1) can be reformulated by considering the approximation of the right hand side of (4) and obtain the following
Telser (1955) considered a portfolio strategy by maximizing portfolio returns under the the constraint of Roy safetyfirst principle. He solved the optimization problem:
Inspired by Telser’s approach, we constrain (5) with the minimum mean return vector \(\mu ^Tx \ge L\) from below where L is the lower bound of \(u^Tx\), where \(L>R\). Thus we obtain the optimization problem :
In another approach, we consider variance and the modified Roy’s safety firstprinciple as a consolidated risk measure in a meanrisk framework. To this end, we propose the optimization problem:
The rest of the modifications is geared towards investigating realistic constraints such as transaction costs, sparsity, and stability to \(P_0\) and \(P_1\) on the financial market.
Portfolio revision
We consider an extension of problems (7) and (8) in which transaction costs are incurred to rebalance or revise the initial portfolio \(x^0\), into an efficient portfolio x. A portfolio of investments may require rebalancing on periodical basis because of updated risk, and return information is generated over time. We make the following assumptions on the transaction cost function c.
Assumption 1
The transaction cost function satisfies the following:

(i)
c(x) is a convex function of x

(ii)
c(0) = 0

(iii)
c(x) \(\ge 0, \quad \forall x\)
To achieve portfolio \(x_k\) from the previous or initial portfolio \(x_k^0\), we make a payment of transaction costs \(c(xx^0)\). We incorporate proportional transaction costs (Kellerer et al. 2000; Muthuraman and Kumar 2006; Mitchell and Braun 2013) which are induced by liquidity costs, tax, brokerage fees (Dumas and Luciano 1991; Kellerer et al. 2000; Lobo et al. 2007) into our model. Therefore, proportional transaction cost follows this structure:
where
where cost of buying is \(c_k^b \>0\) and cost of selling is \(c_k^s>0\).
The \(P_1\) model with proportional transaction costs \((P_{1t})\) is the optimization problem
The model \(P_{1t}\) minimizes the upper bound estimate (5) w.r.t x and superimposes the lower constraint \(L \le \mu 'x\sum _{k=1}^N (c^bx_k^b+c^s x_k^s)\) on the average return after the deduction of transaction costs.
Explaining the constraints with respect to transaction costs, the above optimization problem is subjected to a set of linear constraints. Constraint (10) requires the net return of the portfolio after the deduction of transaction costs to be greater or equal to a threshold level L. Constraint (11) is the budget constraint: the capital available to cover transaction costs and shares of stocks. Constraint (12) shows that \(x_k\) represents the portfolio position to be chosen explicitly through sold shares \(x^s_k\) and purchased shares \(x^b_k\) that are rebalanced adjustments to the initial position \(x^0_k\) of stock k. Constraint (13) and constraint (15) are the complementary constraint and nonnegative constraint respectively. They prevent any possibility of concurrent purchases and sales (Dybvig 2005). Note that \(P_0\) model with proportional cost \((P_{0t})\) is defined similarly but without the variance term in the objective function.
Stable and sparse portfolio
In a portfolio selection strategy where the dimensionality of the set of candidate assets is high, sparsity is desired. When the number of assets is large, a nonregularized numerical approach will intensify the effects of estimation risk, leading to an unstable and unreliable estimate of the vector x. Typically, portfolio managers want to set up portfolios with suitable balance between risk and return by investing in a small number of assets, thereby limiting their transaction, management, and monitoring costs.
To obtain meaningful and sparse (zero components) results, a regularization procedure is usually adopted. A standard approach is to augment the objective function of interest with a \(\textit{l}_0\)norm penalty or adding a cardinality constraint \(\Vert x \Vert _0 \le N'\) to optimization problems \(P_{0t}\) and \(P_{1t}\), where \(\Vert x \Vert _0\) is the number of the nonzero entries of x and \(N'\) is the upper bound limitation of assets to be managed in the portfolio. However, with the inclusion of cardinality constraint, the portfolio selection strategy becomes NPhard (MoralEscudero et al. 2006). We therefore impose its equivalent \(\textit{l}_1\)norm penalty as employed by Brodie et al. (2009) among others. The \(\textit{l}_1\)norm is a convex function of x, and such convex relaxation makes portfolio selection strategy more tractable.
To this end, we suggest to evaluate the portfolio weights by
where the \(\textit{l}_1\)norm of a vector \(x \in \mathbb {R}^N\) is defined by \(\Vert x\Vert _1{:}{=}\sum _{k=1}^N \mid x_k \mid\) and \(\tau _1\) is an adjustable parameter that controls the sparsity of the portfolios. Similarly, \(S_{0t}\) can be formulated without the variance term in the objective function.
Optimization models \(S_{0t}\) and \(S_{1t}\) have estimation risk or overfitting problem. The \(l_1\)norm penalty expedites sparsity of x and leads to a subset of assets receiving zero weights. Such sparsity may result in underdiversification and extreme weights of the portfolio. On the contrary, convex norm ball does not produce sparsity but it can efficiently regularize size of portfolio weight vector. Thus, the norm ball constraint used by DeMiguel et al. (2009) can function as a solution to alleviate the problems of underdiversification and extreme weights of the portfolio aside estimation risk. Following DeMiguel et al.’s (2009) work and specifying the general squared \(l_2\)norm under no short sale constraint, we propose to formulate the portfolio weights by
and
where \(\tau _1\) and \(\tau _2\) are tuning parameters controlling sparsity and stability respectively, with \(\Vert x \Vert _2^2 =x'x\) as the squared \(l_2\)norm of a vector. We estimate tuning parameters \(\tau _1\) and \(\tau _2\) by a method of crossvalidation. We perform crossvalidation for various possible values of the parameters and select the parameter value that produces the minimum crossvalidation average error. A combination of the \(l_2\)norm penalty and \(l_1\)norm penalty is referred to as elastic net (Zou and Hastie 2005).
Necessary and sufficient conditions for optimal problems
In this section we would like to identify the necessary and sufficient conditions for optimality of problems (17) and (18). We investigate the KarushKuhnTucker (KKT) conditions for these problems under the assumption of normality and study whether the constrained problems have optimal solutions.
KKT conditions for optimal problem
The KKT conditions provide necessary conditions for a point to be optimal point for a constrained nonlinear optimal problem. The system
has a unique solution
i.e.
From the budget constraint \(\sum _{k=1}^N x_k +c^b \sum _{k=1}^N x_k^b+c^s \sum _{k=1}^N x_k^s=1\), we get that \(c^b \sum _{k=1}^N x_k^b+c^s \sum _{k=1}^N x_k^s=1\sum _{k=1}^N x_k\). Implementing the first constraint,\(\mu ^Tx\sum _{k=1}^N (c^bx_k^b+c^s x_k^s)\ge L\) we get that \(\mu ^Tx(1\sum _{k=1}^N x_k)\ge L\). Thus, RSMt can be represented as
Similarly, RSMVt can be represented with addition of \(x^TQx\) to the objective function.
Let x be a regular point for the problem RSMt. Then the point x is a local minimum of f subject to constraints (20) if there exists Lagrange multipliers \(\lambda _1, \lambda _2\) and \(\lambda _3\) for the Lagrangian function \(L=f_1(x)+\lambda _1 g_1(x)+ \lambda _2 h_1(x)+\lambda _3 g_2(x)\) such that the following are true.

1.
\(\frac{\partial L}{\partial x}={(1\alpha )CVaR'_{\alpha }(x)}{VaR_{\alpha }(x)R}\frac{(1\alpha )CVaR_{\alpha }(x) VaR'_{\alpha }(x)}{(VaR_{\alpha }(x)R)^2}+\tau _1 c^1 +2\tau _2x\lambda _1(\mu +I^{N \times 1})+ \lambda _2 c^2\lambda _3 VaR'_{\alpha }(x)=0\)

2.
\(\lambda _1(L\mu ^Tx(1\sum _{k=1}^N x_k)=0\)

3.
\(\lambda _3(RVaR_{\alpha }(x))=0\)

4.
\(\lambda _1,\lambda _3 \ge 0\)

5.
\(\mu ^Tx(1\sum _{k=1}^N x_k)\ge L\)

6.
\(c^b \sum _{k=1}^N max\{x_kx_k^0,0\} +c^s \sum _{k=1}^N max\{x_k^0x_k,0\}=1\sum _{k=1}^N x_k\)

7.
\(VaR_{\alpha }(x)> R\)
where
Remark 1
Since the function \(h_1\) in (20) is linear and the functions \(g_1\) and \(g_2\) are convex, then the feasible region \(\Omega = \{x : h_1,g_1,\) and \(g_2\}\) is a convex set. On the other hand, \(f_1\) is a convex function subject to the variable x. We see that any local minimum for problem (20) is a global minimum too and KKT conditions are also sufficient.
Similarly, the above KKT conditions holds for RSMVt for when \(f_1(x)\) in (20) is \(f_2(x)=x^TQx+\frac{(1\alpha )CVaR_{\alpha }(x)}{VaR_{\alpha }(x)R}+\tau _1 \Vert x\Vert _1+\tau _2 \Vert x \Vert _2^2\) and when the following are true

1.
\(\frac{\partial L}{\partial x}=2Qx+{(1\alpha )CVaR'_{\alpha }(x)}{VaR_{\alpha }(x)R}\frac{(1\alpha )CVaR_{\alpha }(x) VaR'_{\alpha }(x)}{(VaR_{\alpha }(x)R)^2}+\tau _1 c^1 +2\tau _2x\lambda _1(\mu +I^{N \times 1})+ \lambda _2 c^2\lambda _3 VaR'_{\alpha }(x)=0\)

2.
\(\lambda _1(L\mu ^Tx(1\sum _{k=1}^N x_k)=0\)

3.
\(\lambda _3(RVaR_{\alpha }(x))=0\)

4.
\(\lambda _1,\lambda _3 \ge 0\)

5.
\(\mu ^Tx(1\sum _{k=1}^N x_k)\ge L\)

6.
\(c^b \sum _{k=1}^N max\{x_kx_k^0,0\} +c^s \sum _{k=1}^N max\{x_k^0x_k,0\}=1\sum _{k=1}^N x_k\)

7.
\(VaR_{\alpha }(x)> R\)
where
Remark 2
Since the feasible region \(\Omega = \{x : h_1,g_1,\) and \(g_2\}\) in (20) and the objective function \(f_2(x)\) are convex, then the feasible region is a convex set. We can see that the KKT conditions are also sufficient and any local minimum for problem (20) with objective function \(f_2(x)\) is a global minimum as well.
Empirical application
Data and models
In this section, we use optimization models RSMt (17) and RSMVt (18) to construct optimal portfolios and evaluate outofsample performance using stocks traded on New York Stock Exchange (NYSE). Historical daily returns of 500 randomly selected stocks over the period January 2003 to December 2015 were extracted from Yahoo Finance. The riskfree rate is proxied by the 6month US Treasury Bill rate. The selection criterion of the random sampling of stocks is based on stocks being traded throughout the evaluation period. We also use S&P 500 index daily stock price data to test the robustness of our results over the same time period even though 21 % of randomly selected stocks are components of S&P 500. We chose a short term period as distribution of stock prices tend to change shape over time. To solve the portfolio strategies, we consider the lower bound of mean return as threshold return level. We compute initial positions \(x^0_k,k=1,\ldots ,N\) for constructing portfolios for January 2006 to December 2008 and January 2009 to January 2015 by solving classical Roy safetyfirst optimization problem:
and by setting \(x_0={\bar{x}}^*\), with \({\bar{x}}^*\) denoting the optimal solution of the above model.
Two portfolio strategies proposed in this research work, RSMt and RSMVt are compared against these existing ones in literature: (1) sample minimum variance without short sale constraint portfolio (minVu) (2) sample minimum variance short sale constrained portfolio (minVc) (3) sample mean–variance approach (MV) (4) naive equallyweighted (1/N) portfolio (5) \(l_1\)penalized mean–variance model (\(l_1\)MV) (Brodie et al. 2009) (6) linear combination of sample tangency portfolio, sample minimum variance portfolio and 1/N portfolio (TMN) (Tu and Zhou 2011) (7) minimum variance portfolio resulting from using a diagonal covariance matrix (VD) (Kirby and Ostdiek 2012), refer to Table 1.
We employ a 6month rolling estimation window for parameter estimations and construct portfolios on a subsample periodical basis. To obtain the portfolio for January 2006 to December 2008, we use January 2003 to December 2005 data to construct the initial positions via the classical Roy safetyfirst optimization problem. For January 2009 to December 2011, we set the portfolio weights from January 2006 to December 2008 as initial positions. We then set the portfolio weights from January 2009 to December 2011 as initial positions to construct portfolios for January 2012 to December 2015. To reflect the true risk of the portfolio, we use January 2006 to December 2008 data as initial positions via the classical Roy safetyfirst optimization problem to construct portfolios for January 2009 to December 2015. Portfolio revision is made on a monthly basis since in reality they are related to low transaction, management and monitoring costs.
Evaluation criteria
With regards to the return that a portfolio can achieve, we calculate and present the annualized outofsample Sharpe ratio with or without transaction cost. Sharpe ratio is defined as the ratio of the expected excess return to standard deviation of portfolio return (Sharpe 1966). The expected excess return is the difference between the return of the portfolio and the return obtained through a riskfree security. Mathematically, we can define Sharpe ratio as \(SRatio = \frac{\rho \rho _f e}{\sigma }\), where \(\rho _f\) is the return from the riskfree security (6month US Treasury Bill rate) and \(\sigma\) is the standard deviation of portfolio return. Regarding portfolio risk, we consider the risk reduction which is defined as the ratio of portfolio risk measure from the portfolio strategies (17) and (18) to that from (7) and (8) respectively. The other portfolio selection strategies in literature’s risk reduction are estimated as the ratio of portfolio variance to that from Markowitz’s mean–variance framework.
Sparsity may result in underdiversification and extreme weights of the portfolio. We therefore present the average number of a subset of assets with nonzero weights. We consider another performance metric known as portfolio turnover. A portfolio turnover measures the frequency with which assets in this case stocks are bought and sold. We employ the measure used by DeMiguel et al. (2009) and Kourtis et al. (2012) by defining the turnover rate of the portfolio between t to \(t+1\) as
where \(x_{k,t+1}\) is the portfolio weight for stock k at \(t+1\), \(x^_{k,t+1}\) is the portfolio weight before revision at \(t+1\), \({\mathring{T}T1}\) represents the length of the nonzero elements in total portfolio return and N, the number of stocks.
The introduction of transaction costs affects the overall profitability of a portfolio strategy. In practice, they lessen the net returns and diminish capital available for future investments. We assume proportional transaction cost for the purpose of this study (Please refer to Assumption 1 for more details). We investigate the impact of transaction cost on our portfolio strategy by computing Sharpe ratio with transaction cost. In particular, the optimal portfolios can be obtained by solving problems (17) and (18). To highlight the effect of transaction costs, we consider two situations with transaction costs, \(c^b_k=c^s_k=0\) and \(c^b_k=c^s_k=0.02\).
Computational results
Tables 2 and 3 shows the annualized outofsample metrics for different periods of each portfolio considered in this study: RSMt is the regularized meansafety first portfolio with transaction costs, RSMVt is the regularized mean–variancesafetyfirst portfolio with transaction costs. The others are selected alternative portfolio strategies in literature.
Analyzing portfolio risk, the regularized portfolio rules have lower risk than other portfolio selection strategies. This can be attributed to the downside risk measure, Roy safetyfirst principle considered in the regularized strategies. The variance and the safetyfirst principle as a consolidated risk measure in RSMVt have lower risk reduction compared to RSMt.
The portfolio turnover indicates how frequently assets in a portfolio are bought and sold. This performance measure is preferred to be small. In terms of turnover, the naive 1/N portfolio has the lowest turnover. The two portfolio selection strategies considered in this paper have relatively low turnover rates with RSMVt the preferred choice. The portfolio rules in selected from literature have a high turnover rate as compared to the regularized models in this paper. The highest turnover comes from TMN, MV and minVu portfolios leading to smaller Sharpe ratios. We observe that with large turnovers, feasible transaction costs lowers the monetary gains of many selected portfolios strategies in literature, as seen by Sharpe ratio deductions after transaction costs have been considered.
The Sharpe ratio allows investors to analyse riskadjusted returns in exchange for the level of risk they are assuming. The higher the Sharpe ratio, the more returns the investor gets per unit of risk. The lower the Sharpe ratio, the more risk the investor bears to get more returns. Comparing all strategies in this study, the regularized portfolios have the highest Sharpe ratios with RSMVt leading the way. With regards to sparsity, more than 30 % of stocks are selected by RSMt and RSMVt in all the evaluation periods. The \(l_1\)penalized mean–variance model selects at least 50 % of the stocks across the evaluation periods. The smaller set of sparse portfolio optimizes the budget allocation by focusing on stocks believed to foster diversification.
From Table 2, the samplebased portfolios i.e. minVu, minVc, MV and TMN perform worse due to a large number of stocks that increases the degree of estimation risk or overfitting. Apart from VD, all other nonregularized portfolios selected from literature (minVu, minVc, MV, TMN) perform worse than 1/N regarding both Sharpe ratio and turnover. We observe an increment in Sharpe ratio, lower risk reduction, lower turnover for post financial crisis subsample periods. Among the selected alternative portfolio strategies considered in this paper, \(l_1\)MV has a better outofsample performance than minVu, minVc, MV, 1/N, TMN and VD. In comparing periods 2006–2008 to 2009–2015, we observe that during the financial crisis period, the Sharpe ratio is lower and the risk reduction is much higher with turnover ratio also higher. This amounts to lower returns during 2006–2008 as compared to the other periods. The Sharpe ratio is much higher and the risk reduction is lower in periods 2009–2011 and 2012–2015 as compared to periods 2006–2008 and 2009–2015. In all the subsample periods 2006–2008, 2009–2011 and 2012–2015 including the period 2009–2015, our proposed models RSMt and RSMVt have better outofsample performance than the selected alternative models.
To further gain financial insights, we use S&P 500 index data from January 2009 to December 2015 as a benchmark portfolio. With a starting wealth value of $1, we compare the cumulative wealth of portfolio strategies RSMt, RSMVt and \(l_1\)MV to that of S&P 500 index. To provide evidential proof, Fig. 1 plots the cumulative wealth of the portfolios strategies relative to S&P 500 benchmark.
In Fig. 1, $1 is used as initial wealth and it grows at a monthly return of the portfolio selection strategies considered. The figure shows distinctly the higher performance of the regularized portfolio over S&P 500 benchmark.
Conclusion
In this paper, we seek nearoptimal sparse and stable portfolios to reduce the difficulty of portfolio management. Theoretical results are established to guarantee the stability and sparsity of our novel portfolio strategies. Computational evidence indicates that \(l_1\) squared \(l_2\) penalized meanRoy safetyfirst portfolio and \(l_1\) squared \(l_2\) penalized mean–varianceRoy safetyfirst portfolio are able to choose optimal sparse and stable portfolios while maintaining satisfactory outofsample performance.
We compare the performance of our proposed models (\(l_1\) squared \(l_2\) penalized meanRoy safetyfirst portfolio and \(l_1\) squared \(l_2\) penalized mean–varianceRoy safetyfirst portfolio) of optimal asset allocation relative to selected alternative portfolio strategies in literature (minVu, minVc, MV, 1/N, \(l_1\)MV, TMN and VD). Our results show that our regularized proposed models have a better outofsample performance with high Sharpe ratios and relatively low turnover rates. Except for VD, the Sharpe ratio of 1/N portfolio when compared to other selected nonregularized portfolio rules considered in this paper is higher, which shows that estimation errors in returns shrink gains from other selected classical portfolio strategies in literature. The norm penalty improves Sharpe ratio and turnover. As a result, \(l_1\)MV has a better outofsample performance than minVu, minVc, MV, 1/N, TMN and VD.
To gain more financial acumen, we compare our proposed models and the best performing portfolio strategy among the selected models from literature considered in this paper, to a benchmark, S&P 500 index. The results indicate that given an initial wealth of $1, the excess returns from \(l_1\) squared \(l_2\) penalized mean–varianceRoy safetyfirst portfolio is the highest. Our proposed models for optimal asset allocation are favourable since they overcome unsteady and extreme portfolio weights induced by estimation error due to parameter uncertainty.
References
Bawa VS, Brown SJ, Klein RW (1979) Estimation risk and optimal portfolio choice. NorthHolland Publ Co, New York
Bawa VS (1978) Safetyfirst, stochastic dominance, and optimal portfolio choice. J Financ Quant Anal 13(02):255–271
Borkovec M, Domowitz I, Kiernan B, Serbin V (2010) Portfolio optimization and the cost of trading. J Invest 19(2):63–76
Brandt M (2009) Portfolio choice problems. Handbook Financ Econom 1:269–336
BrittenJones M (1999) The sampling error in estimates of meanvariance efficient portfolio weights. J Finance 54(2):655–671
Brodie J, Daubechies I, De Mol C, Giannone D, Loris I (2009) Sparse and stable Markowitz portfolios. Proc Natl Acad Sci 106(30):12267–12272
Chang TJ, Meade N, Beasley JE, Sharaiha YM (2000) Heuristics for cardinality constrained portfolio optimisation. Comput Oper Res 27(13):1271–1302
Chiu MC, Wong HY, Li D (2012) Roys safetyfirst portfolio principle in financial risk management of disastrous events. Risk Anal 32(11):1856–1872
DeMiguel V, Garlappi L, Nogales FJ, Uppal R (2009) A generalized approach to portfolio optimization: improving performance by constraining portfolio norms. Manag Sci 55(5):798–812
DeMiguel V, Mei X, Nogales FJ (2014) Multiperiod portfolio optimization with many risky assets and general transaction costs. Available at SSRN 2295345
Dumas B, Luciano E (1991) An exact solution to a dynamic portfolio choice problem under transactions costs. J Finance 46(2):577–595
Dybvig PH (2005) Meanvariance portfolio rebalancing with transaction costs. Working paper, Washington University in Saint Louis
Fan J, Zhang J, Yu K (2012) Vast portfolio selection with grossexposure constraints. J Am Stat Assoc 107(498):592–606
Farrell JL, Reinhart WJ (1997) Portfolio management: theory and application. McGrawHill, New York
Goldfarb D, Iyengar G (2003) Robust portfolio selection problems. Math Oper Res 28(1):1–38
Green RC, Hollifield B (1992) When will meanvariance efficient portfolios be well diversified? J Finance 47(5):1785–1809
Haque M, Varela O, Hassan MK (2007) Safetyfirst and extreme value bilateral US–Mexican portfolio optimization around the peso crisis and NAFTA in 1994. Q Rev Econ Finance 47(3):449–469
Jagannathan R, Ma T (2003) Risk reduction in large portfolios: why imposing the wrong constraints helps. J Finance 58(4):1651–1684
Jorion P (1986) Bayes–Stein estimation for portfolio analysis. J Financ Quant Anal 21(03):279–292
Kellerer H, Mansini R, Speranza MG (2000) Selecting portfolios with fixed costs and minimum transaction lots. Ann Oper Res 99(1–4):287–304
Kirby C, Ostdiek B (2012) It’s all in the timing: simple active portfolio strategies that outperform naive diversification. J Financ Quant Anal 47(02):437–467
Kourtis A, Dotsis G, Markellos RN (2012) Parameter uncertainty in portfolio selection: shrinking the inverse covariance matrix. J Bank Finance 36(9):2522–2531
Ledoit O, Wolf M (2004) A wellconditioned estimator for largedimensional covariance matrices. J Multivar Anal 88(2):365–411
Levy H, Sarnat M (1972) Safety firstan expected utility principle. J Financ Quant Anal 7(03):1829–1834
Lobo MS, Fazel M, Boyd S (2007) Portfolio optimization with linear and fixed transaction costs. Ann Oper Res 152(1):341–365
Maringer D, Kellerer H (2003) Optimization of cardinality constrained portfolios with a hybrid local search algorithm. OR Spectrum 25(4):481–495
Markowitz H (1952) Portfolio selection. J Finance 7(1):77–91
Markowitz HM (1968) Portfolio selection: efficient diversification of investments, vol 16. Yale University Press, New Haven
Merton RC (1980) On estimating the expected return on the market: an exploratory investigation. J Financ Econ 8(4):323–361
Mitchell JE, Braun S (2013) Rebalancing an investment portfolio in the presence of convex transaction costs, including market impact costs. Optim Methods Softw 28(3):523–542
MoralEscudero R, RuizTorrubiano R, Suárez A (2006) Selection of optimal investment portfolios with cardinality constraints. In: IEEE congress on evolutionary computation, CEC 2006, IEEE; pp 2382–2388
Muthuraman K, Kumar S (2006) Multidimensional portfolio optimization with proportional transaction costs. Math Finance 16(2):301–335
Nagai H (2003) Optimal strategies for risksensitive portfolio optimization problems for general factor models. SIAM J Control Optim 41(6):1779–1800
Nawrocki DN (1999) A brief history of downside risk measures. J Invest 8(3):9–25
Pai G, Michel T (2009) Evolutionary optimization of constrainedmeans clustered assets for diversification in small portfolios. IEEE Trans Evol Comput 13(5):1030–1053
Platen E, Heath D (2006) A benchmark approach to quantitative finance. Springer Science & Business Media, New York
Rachev ST (2001) Safetyfirst analysis and stable paretian approach to portfolio choice theory. Math Comput Model 34(9):1037–1072
Rachev S, Ortobelli S, Stoyanov S, Fabozzi FJ, Biglova A (2008) Desirable properties of an ideal risk measure in portfolio theory. Int J Theor Appl Finance 11(1):19–54
Rockafellar RT, Uryasev S (2000) Optimization of conditional valueatrisk. J Risk 2:21–42
Roman D, DarbyDowman K, Mitra G (2007) Meanrisk models using two risk measures: a multiobjective approach. Quant Finance 7(4):443–458
Roy AD (1952) Safety first and the holding of assets. Econom J Econom Soc 20:431–449
RuizTorrubiano R, Suárez A (2015) A memetic algorithm for cardinalityconstrained portfolio optimization with transaction costs. Appl Soft Comput 36:125–142
Schultz R, Tiedemann S (2003) Risk aversion via excess probabilities in stochastic programs with mixedinteger recourse. SIAM J Optim 14(1):115–138
Sharpe WF (1966) Mutual fund performance. J Bus 39(1):119–138
Telser LG (1955) Safety first and hedging. Rev Econ Stud 23(1):1–16
Tibshirani R (1996) Regression shrinkage and selection via the lasso. J R Stat Soc Ser B (Methodol) 1:267–288
Tütüncü RH, Koenig M (2004) Robust asset allocation. Ann Oper Res 132(1–4):157–187
Tu J, Zhou G (2011) Markowitz meets Talmud: a combination of sophisticated and naive diversification strategies. J Financ Econ 99(1):204–215
Zou H, Hastie T (2005) Regularization and variable selection via the elastic net. J R Stat Soc Ser B (Stat Methodol) 67(2):301–320
Authors' contributions
EFEAM, the corresponding author, conceived, designed the methodology and wrote the paper. DY redesigned the structural framework of the paper and edited the content. XW and DY aided in the numerical test and analysis of the results. BY made a significant contribution to the methodological approach, analysis framework and supervised the study. All authors have read and approved the final manuscript.
Acknowledgements
We thank the reviewers for taking time to add their comments and suggestions to this paper. We also thank SpringerPlus for giving us the platform to contribute to literature. We acknowledge support by the National Nature Science Foundation of China (11571061, 71301017).
Competing interests
The authors declare that they have no competing interests.
Author information
Authors and Affiliations
Corresponding authors
Rights and permissions
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
About this article
Cite this article
Atta Mills, E.F.E., Yan, D., Yu, B. et al. Research on regularized mean–variance portfolio selection strategy with modified Roy safetyfirst principle. SpringerPlus 5, 919 (2016). https://doi.org/10.1186/s4006401626217
Received:
Accepted:
Published:
DOI: https://doi.org/10.1186/s4006401626217
Keywords
 Meanrisk portfolio
 Safetyfirst
 Stable portfolio
 Sparse portfolio