Back to search

Athens University Economics and Business (AUEB)

Essays on asset liability management and dynamic asset allocation

Abstract

dc:description

The recent interest in the area of strategic asset allocation for long–term investors stems from the adverse market conditions in the early 2000s and the financial crisis 2007-2008. The fall in equity market returns has devastated pension plans’ assets and at the same time the decline in interest rates has increased the present value of their liabilities (i.e. future obligations and contributions). In 2000, less than 20% of S&P 500 firms experienced underfunded pension plans; by 2010, about 64% had underfunded plans. In the second pension storm, a 63bn dollars pension surplus for S&P 500 firms in 2007 turned into a 308bn dollars deficit by the end of 2008. Firms that used strategic asset allocation practices to match pension assets with liabilities, coined Asset Liability Management (ALM) strategies, have been better able to protect their funded positions during the crisis periods.The objective of ALM is to consider the fund’s asset value compared with the fund’s liabilities, rather than in absolute terms. A first approach in the ALM literature consists in matching liabilities cash flows with assets cash flows of the same maturity, the so–called cash–flow matching approach. In this approach fixed–income instruments such as zero–coupon bonds or inflation–indexed zero-coupon bonds are used in the asset allocation decision policies. When such assets do not exist, another approach, less precise, relies on immunization, which allows an imperfect matching by equalizing the durations of assets and liabilities. However, a shortcoming of these approaches is their low expected return, since most of the assets held in the portfolio are fixed–income instruments. A second approach, called surplus optimization, seeks maximal expected return on the surplus, defined as the difference between expected returns on assets and liabilities (Sharpe and Tint, 1990) and results in an asset mix that is expected to produce higher expected returns. Finally, a related approach, called liability-driven investment (LDI), relies on the construction of two seperate portfolios: the first portfolio, the performance–seeking portfolio (PSP), seeks generating performance, as in standard strategic asset–only allocation approaches; the second portfolio, the liability–hedging portfolio (LHP) aims at hedging thepension fund liabilities. Leibowitz (1987) and Sharpe and Tint (1990) are the first papers that incorporate liabilities in a single period context. Leibowitz (1987) and Sharpe and Tint (1990) build under the assumption of a constant opportunity set and hence lead to myopic single–period investment decisions. However, there has been well documented empirical evidence that the asset premia are time-varying rather than constant (Campbell and Shiller, 1988a; 1991; 1998; 2005, Fama and Schwert, 1977, Fama and Bliss, 1987, Fama and French, 1989, Campbell et al., 2003). In a world of time–varying investment opportunities (risk premia, interest rates, inflation etc), optimal portfolio choice is horizon-dependent. Long term investors’ optimal portfolio strategy will be dynamic with frequent adjustment to their asset allocation policy in response to the changing opportunity set. Early studies in this directions were for instance proposed by Samuelson (1969) and Merton(1969; 1971; 1973) showing that the solution to a multi–period portfolio choice problem can be very different from the solution to a myopic portfolio choice problem.In Chapter 2 we develop an analytical solution to the dynamic multi–period portfolio choiceproblem of an investor with risky liabilities and time varying investment opportunities. An advantage of this solution is that it can be readily implemented for investment opportunity sets with any number of assets and state variables and stands out in terms of intuition, tractability, and potential for practical applications. Our model builds on a strand of literature that uses approximate analytical solutions under an asset-only framework (Campbell and Viceira, 1999; 2001; 2002, Campbell et al., 2003 and Jurek and Viceira, 2011). Closer in spirit to our model is the model proposed by Hoevenaars et al. (2008) who assume that an investor with risky liabilities chooses the optimal portfolio weights at the beginning of the investment horizon and rebalances to the initial allocation at a rebalancing time shorter than the investment horizon. In contrast, dynamic investors follow a dynamic investment strategy where the portfolio weights take into account changes in the investor’s investment opportunity set. We use the model to compare the asset allocation of investors who take liabilities into account, assuming time varying returns and a multi–period setting, with the asset allocation of myopic ALM investors. In the absence of regulatory constraints on asset allocation weights, there are significant gains to investors who have access to a dynamic asset allocation model with liabilities. The gains are much smaller under the typical funding ratio constraints faced by pension funds.Recently, commodities have been considered an increasingly important class of assets forinstitutional (i.e. pension funds, foundations, endowments funds) and individual investors. The main reasons for commodity investing have been threefold: first, commodities can potentially capture a commodity risk premium, second they can yield further diversification benefits when incorporated in a traditional–only portfolio, and third they are considered to be good hedges against expected and unexpected inflation; see for a discussion on the latter reasons the studies of Bodie and Rosansky (1980) and Gorton and Rouwenhorst (2006). However, over the past ten years we document an increased investment in commodities, which in turn has increased the correlation between commodities and traditional asset classes and as a result has questioned the diversification benefits associated with commodity investing (Tang and Xiong, 2012; Silvennoinen and Thorp, 2013; Lombardi and Ravazzolo, 2013; Daskalaki and Skiadopoulos, 2011).A widely used approach by commodity investors is to invest in a broad commodity marketindex, such as the Standard & Poor’s Goldman Sachs Commodity Index (S&P GSCI) orthe Dow Jones–UBS Commodity Index (DJ-UBSCI). However, while style–based investmentsand their role for portfolio allocation in equity markets have been widely studied (Eun et al.,2008, 2010; Ang et al., 2009; Bender et al., 2010; Ilmanen and Kizer, 2012; Blitz, 2012), there is considerably less knowledge about the portfolio implications of style investing in commodity markets. Furthermore, in practice investors make their asset allocation decisions in the presence of their liabilities. In Chapter 3 we shed new light on the role of commodities in strategic asset allocation for investors with and without liabilities who (a) believe that asset returns are time varying and predictable (b) have short and long term horizons and (c) have access, in additionto a standard passive commodity portfolio (i.e. the S&P Goldman Sachs CommodityIndex (S&P GSCI)), to commodity portfolios based on equal weights, momentum and the basis.We document significant benefits, in– and out–of–sample, from investing in factor–basedcommodity portfolios. The benefits from commodity momentum and basis portfolios are larger for investors who do not have access to inflation–indexed bonds. Furthermore, the importance of commodities in investors’ portfolios, albeit reduced, remains significant even when equity value and momentum tilted portfolios are included. We also confirm and extend the evidence on the negative role of commodity investments based on commonly used commodity benchmarks for investors with long horizons and liabilities.In Chapter 2 and Chapter 3 our analysis has been developed under a time varying investment opportunity framework. In Chapter 4 we consider a single–period (myopic) framework. We propose a straight-forward, intuitive, and easy to implement approach for a truly factor–based asset allocation model in an Asset Liability Management (ALM) framework. Our model combines the models developed in an Asset Liability Management framework that use as building blocks primarily asset classes (Leibowitz, 1987; Sharpe and Tint, 1990), and the recent factor–based allocation models developed in an Asset Management setting (Clarke et al., 2005;Eun et al., 2008, 2010; Ang et al., 2009; Bender et al., 2010; Ilmanen and Kizer, 2012; Page and Taborsky, 2011; Blitz, 2012). The objective is to construct the optimal asset mix that best mimics a portfolio of (fundameτnal) factors which determine the fund’s assets and liabilities.Thus, the proposed model is developed under a new modeling framework of the liabilities and assets of a pension fund. Modeling pension liabilities is a difficult task since it does not exist an asset class or portfolio that exactly replicate the pension fund liabilities. Pension liabilities are partially or fully indexed to the consumer price index and behave like bonds (Ezra, 1991). Thus, we consider pension liabilities to be represented by a non–fully (inflation) indexed bond (generic bond). We derive analytically the fundamental factors of assets and liabilities based on a total differential approach of the price of a stock and a non-fully inflation indexed bond (generic bond). Finally, this framework allows us to revisit the inflation hedging capacity of traditional assets (i.e. equities, nominal bonds and inflation–indexed bonds) and contribute to the mixed evidence on the relationship between asset returns and inflation (Nelson, 1976; Miller et al., 1976; Bodie, 1976; Fama and Schwert, 1977; Kaul, 1987; Fama and Schwert, 1977; Boudoukh and Richardson, 1993; Boudoukh et al., 1994). We show that inflation indexed bonds are by definition perfect inflation hedges. Our empirical evidence suggest that stocks are good hedges against inflation in the short and long run. Stocks are unrelated with changes in inflation expectations, which express the short–term component of inflation, and at the same time they are slightly economically affected by an inflation shock. Additionally, the inflation hedging properties of equities are closely related with the prefect inflation hedging properties of an inflation–indexed bond. On the other hand, nominal bonds are good inflation hedges in the long run, though they suffer from severe losses in the short term when changes in inflation expectations and unexpected inflation increase.Collectively, this thesis contributes to different strands of literature. First, we contributeto the dynamic multi–period portfolio choice literature within an ALM framework. We additionally take into account the regulatory framework in an asset allocation exercise for a defined–benefit pension plan. Second, we contribute to the literature on commodities from a strategic asset allocation perspective. In this respect, we investigate the implications of commodities investing for different types of investors (asset–only and asset–liability investors), commodities (standard–to–date benchmarks and commodity factor premia) and investment horizons (short– and long–term horizons). Third, we contribute to a growing literature on factor–based asset allocation models. We propose a straight-forward, intuitive, and easy to implement approach for a truly factor–based asset allocation model within a single–period Asset Liability Management (ALM) framework.

Degree

thesis:*
Grantor dc:publisher
Athens University Economics and Business (AUEB)
Year dc:date
2015

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Sakkas, Athanasios
  • Σάκκας, Αθανάσιος

Subjects

dc:subject × 9

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/35938

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Sakkas, Athanasios; Σάκκας, Αθανάσιος. Essays on asset liability management and dynamic asset allocation. Athens University Economics and Business (AUEB), 2015. http://hdl.handle.net/10442/hedi/35938