“The role of commodities investments in the decumulation phase of retirement”

This paper examines the role of commodity-related investments in the decumulation phase of retirement. Benchmarked against a balanced portfolio, the findings suggest that including commodities in a traditional portfolio improves the retirement outcomes at the lowest percentiles of wealth distribution. Furthermore, we demonstrate that downside protection is more pronounced by reducing allocation to equities (rather than bonds) to invest in alternatives. An equally weighted combination of passive and active commodity-related investments provides superior downside protection compared to a traditional portfolio at all levels of allocations used in the analysis. As a consequence, commodities may be employed as a portfolio diversification tool particularly in the decumulation phase of retirement.


Introduction
This paper examines the role of commodity-related investments in the decumulation phase of the retirement portfolio.Using a 20-year investment horizon and a 4 per cent drawdown rule, we investigate retiree's retirement outcome based on the investment strategy she chooses at retirement.Benchmarked against the traditional 60/40 asset allocation, we examine the performance of commodities long-only investment (S&P-GSCI), active commodity-related trend following (Commodity Trading Advisors, CTAs/Managed Futures) and a combination of passive and active commodities-related investments.Our findings suggest that, compared to a traditional asset allocation, a strategy with allocations to commodityrelated investments significantly improves retirement outcomes.This is, particularly, the case at the lowest percentiles of wealth distribution.Wiafe (2015) finds that whilst increasing stock levels of portfolios significantly increases the wealth outcome for the right tail of the distribution, it increases the chance of portfolio ruin which results from the overreliance on positive returns from volatile assets.In this paper, we argue that mean and median of the terminal wealth are inappropriate metrics for the measurement of retirement adequacy.As financial risk aversion is highest in pension relative to other types of investments (see Van Rooij et al., 2007), we posit that studies should focus on improving the worst outcomes (extreme left tail of the distribution) instead.Therefore, commodity-related investments are selected as the focus, given their well-received advantage for portfolio diversification (Bodie and Rosansky, 1980; Anson, 1999; Jensen et al., 2002;Erb and Harvey, 2006; Gorton and Rouwenhorst, 2006).Hoevenaars et al. (2008) examine the role of commodities longonly investments in the accumulation phase of retirement.They find that alternative asset classes add value for long-term investors.For the first time in the pension literature, we examine the performance of commodity-related investments, both long-only and active trading, in the decumulation phase.This paper presents three major contributions to the literature.First, we demonstrate that individuals have better downside protection (at the left tail of the distribution) during the decumulation phase by including commodity-related investments in their portfolios.At the 1 st percentile, increasing allocation to commodities up to 15 per cent increases terminal wealth.Even at 20 per cent, the 1 st percentile of terminal wealth is higher for the portfolio with alternatives than the traditional portfolio.At the 5 th percentile of terminal wealth, considerable difference exists in portfolio outcomes for the different strategies analysed.While increasing allocation to commodities generally decreases wealth at the 5 th percentile, the inverse is reported for the CTA/Managed Futures allocation.
Second, at both the 1 st and 5 th percentiles of the wealth distribution, our findings suggest that the downside protection is more pronounced by reducing allocation to equities (rather than bonds) to invest in commodity-related alternatives.This is explained by the high negative correlation between commodity-related alternatives and bonds,which is significantly higher than the correlation relative to equities.At the 1 st percentile, even when allocation to alternatives is increased with a reduction in bonds, there is still a significant increase in wealth levels with increasing allocation to CTA/Managed Futures, while the inverse is noted for allocation to commodities.
Finally, an equally weighted combination of the two commodity-related investments (long-only commodities and CTA/Managed Futures) provides superior downside protection compared to a traditional portfolio at all levels of allocations used in the analysis.These findings suggest that diversifying through alternative investments, indeed, provides better outcomes compared to the traditional portfolio in the decumulation phase of the retirement.This is, particularly, the case at the 1 st percentile, when downside protection is required the most.
For robustness, we investigate the worst outcomes for different allocation strategies by employing the Expected Tail Loss (ETL) and Value-at-Risk (VaR) metrics.We find that VaR and ETL generally decrease with increasing investment horizon.After 20 years in retirement, retiree VaR is, on average, lowest for a strategy an allocation to CTA/Managed Futures.The strategy with a combination of alternatives however underperforms the remaining strategies, having the highest VaR over the long term.These results confirm the findings on terminal wealth distributions, suggesting that allocating to commodity-related investments reduce the extreme losses retirees may experience during their retirement.
Our findings present important implications to the pension funds management industry.The remainder of the paper proceeds as follows.Section 1 summarizes the relevant literature.Section 2 outlines data and investment strategies employed in this study.Section 3 discusses the detailed results and the paper concludes.

Literature review
Appropriate investment measures play a crucial role in wealth accumulation.In the retirement literature, existing research has mainly focussed on developing well-diversified and sustainable investment portfolios during the working life of individuals (Basu and Drew, 2009;Milevsky, 1998).Findings by Oeppen and Vaupel (2002) reveal that female life expectancy in the last 160 years has been increasing steadily by almost three months every year.Increasing life expectancy highlights the importance of optimal investment strategies and the need for adequate measures to cater for elongated retirement periods (Krumholz et al., 2015;Blake et al., 2006).
The literature reveals a number of key variables that determine retirement outcomes for the modern worker: contribution rate (Basu et al., 2011;Blake et al., 2001), risk aversion level (Jagannathan and Kocherlakota, 1996;Hickman et al., 2001) and gender (Neelakantan and Chang, 2010;Sunden and Surette, 1998).Whilst the majority of the literature focuses on the accumulation phase, until recently, not much attention has been given to the decumulation phase.It is important in retirement that individuals take active measures to ensure that they do not outlive their available wealth.These include decisions regarding asset allocations, portfolio drawdown plans and longevity management.In many developed countries, the onus lies on the retiree to manage her risks in retirement, although there is the benefit of government sponsored pensions in countries such as Australia.

Data and methodology
For traditional asset classes, we obtain historical monthly returns data on U.S. equities and bonds between January 31, 1980 and May 31, 2014, spanning a period of 35 years sourced from the Global Financial Database (GFD).We employ the S&P 500 Total Return Index as the equities investment and US 10-year Government Bonds Total Return Index as the proxy for bond investment.
For alternative investments, we employ the S&P-GSCI (Standard and Poor's Goldman Sachs Commodity Index) Total Return Index as a proxy for passive long-only commodities investing.The S&P-GSCI consists of 24 actively traded commodities across energy (crude oil, brent, gas oil, natural gas, RBOB gas and heating oil), metals (aluminium, copper and zinc, lead, nickel, gold and silver), agriculture (Chicago wheat, Kansas wheat, corn, soybeans, cotton, sugar, coffee and cocoa) and livestock (live cattle, feeder cattle and lean hogs) sectors.The S&P-GSCI is selected for two reasons.First, it is one of the most widely used performance benchmarks in commodity markets by industry practitioners.Second, a considerable number of exchange traded products on the S&P-GSCI index are readily available.Furthermore, the Barclay CTA (Commodity Trading Advisors) Index is employed as a proxy for commodity-related trend-following strategies.Published and maintained by Barclay Hedge, the Barclay CTA Index is computed based on of 535 trading programs.Only advisor which have four years or more of prior performance history are included in the index.Additional programs introduced by qualified advisors are not added to the Index until after their second year.The Barclay CTA Index is selected as it offers the longest history available for CTA performance tracking 1