Portfolio Construction: Benefits in being ‘insurance-like’
Sorca Kelly-Scholte and Michael Buchenholz say that pension funds can build better portfolios by adopting strategies used by insurers
The investment problem for the annuity liabilities of insurers and corporate pension liabilities is similar but in practice the investment strategies diverge considerably. Examining aspects of insurance investing that can be practically implemented by corporate pension funds may help pensions to construct more resilient fixed-income portfolios and extend their use of the credit markets. In particular, we think that pensions should take a leaf from insurers on:
• Using investment-grade credit as a venue for duration hedging.
• Adopting buy-and-maintain strategies, focussing on low turnover stable cash flow portfolios that match liabilities and minimise transaction costs.
• Diversifying across the full credit spectrum, to diversify systematic risk exposures, reduce single name concentration and lower interest rate risk.
Many European pension funds, either through regulation or generally accepted practice, value their liabilities by reference to government yield or swap curves. Against this measure of liabilities, investing in duration-matched credit can introduce substantial funding-level volatility as credit spreads vary through market cycles, as shown in the figure (using US data where credit markets developed earlier and there is the longest available data set).
For this reason, and probably also as a result of a historical paucity of long-dated investment-grade credit in Europe, many pension funds restrict their liability-hedging portfolio to government bonds and related derivatives and concentrate their risk budget in the equity market. However, as the figure also shows, the tracking error arising from equity has been substantially larger than that of holding credit.
The strategy remains rational for those who assume that the potential return per unit of risk is better in equity than in credit, and there is little benefit in diversifying your risk exposure across equity and credit. The former is arguable – our long-term capital market assumptions suggest similar liability-relative risk/return ratios in equities and investment-grade credit. The latter is empirically difficult to defend, and there is much benefit in diversifying risk across credit and equity markets. To maintain expected levels of return, this will mean allocating to credit from liability-hedging portfolios as well as from growth portfolios, and shaping the credit portfolio to hedge liability duration in the way that insurance companies do.
Referring again to the figure, a notable feature is the trade-off between funding-level volatility and the ending funding level across various portfolios. Of the credit strategies, long corporate Baa bonds have the highest returns and ending funding level over the period, but also exhibit the highest tracking error to the pension liability. In contrast, by reducing basis risk and moving to a higher-rated portfolio, funding-level volatility is significantly reduced but the deterioration in funding level over the period is magnified owing to lower returns.
One central inference from this exercise is that a portfolio of high-quality long-duration bonds alone is unlikely to earn sufficient return above the liability to merit the associated funding level volatility. To keep pace, either the inclusion of higher returning assets (and a higher tracking error) or sponsor contributions will be necessary. However, the dispersion in performance across credit qualities suggests that active bond management can alleviate some of this historically persistent drag on returns by allowing portfolio managers to apply credit-research insights.
Buy-and-maintain strategies focus on actively selecting high-quality bond securities that can be held to maturity, based primarily on the underlying financial strength of their issuers and their perceived remoteness from severe negative credit migration and default. Under this approach, bond sales are generally only made if the portfolio manager forecasts credit deterioration that is not reflected in market prices.
For example, in a downgrade situation, most of the price depreciation comes prior to the announcement as the market acts quicker than rating agencies. Passive strategies that are forced sellers of out-of-benchmark exposures will suffer, as they will typically sell at the lowest point. Barclays estimates that a buy-and-hold investment-grade portfolio (all maturities, not long only) from 1993-2015 has earned 75bps of annual excess returns. If the strategy is forced to sell bonds downgraded to high yield, rather than hold them to maturity, the annual excess return drops by 20bps to 55bps.
In contrast to the systematic index rules embedded in most pension credit, insurers’ use of buy-and-maintain credit strategies allows them the flexibility to hold on to downgraded issues. Our analysis of insurance statutory filing data in the US indicates that most bonds downgraded from investment grade to high-yield (fallen angels) are held by insurers for at least one year (or to maturity) as opposed to automatically selling on downgrades.
In addition to active management, another approach to reducing the impact of downgrades and defaults is portfolio diversification, both among risk-factor exposures and issuers. There are multiple levers fixed-income investors can use to generate diversified sources of risk and return but, in practice, many corporate pension funds allocate to traditional government mandates in their liability-driven investment (LDI) portfolios, which are dominated by duration risk, and to broad market-credit benchmarks which are dominated by duration and credit risk. Expanding the solution set to include a more diversified set of risk premia, such as liquidity, structured credit and leverage risk, presents an opportunity to add incremental yield to the portfolio and enhance risk-adjusted returns. Once again, insurance portfolios offer useful insights into these methods.
Insurers, and specifically those whose statutory reserves are largely backing annuity-based liabilities (much like pension funds), tend to invest less than half of their fixed-income portfolios in investment grade credit. The remainder is diversified across multi-sector fixed-income classes such as commercial mortgage-backed securities (CMBS), asset-backed securities (ABS), mortgages (agency and non-agency, collateralised and pass-through, and public and private), bank loans and private credit assets. Historically the credit markets in Europe have not offered this degree of diversity, but since the 2007-08 financial crisis the disintermediation process in Europe has opened up new opportunities. Insurance companies are also seeking out opportunities more globally, and we expect pension funds to follow suit.
Insurers face many more constraints than pension investors in determining their investment strategies, which can result in less efficient portfolios overall than pension portfolios. Nonetheless, as pension funds gradually turn themselves into annuity books, and accordingly continue to derisk, our analysis shows that diversifying away from corporate credit into a more ‘insurance-like’ portfolio, has much merit. Combining the hedging techniques of insurance companies with the investment freedom of pension funds can ultimately can translate into improvements to total return, funding level volatility, or both.
Sorca Kelly-Scholte is EMEA head of pensions advisory and solutions, and Michael Buchenholz is a pensions strategist, JP Morgan Asset Management