Enhancing capital efficiency with Matching Adjustment

Insurance companies in Singapore are obliged to meet regulatory metrics¹ termed Capital Adequacy Ratio (“CAR”) and Fund Solvency Ratio (“FSR”) by holding sufficient capital against their calculated risks known as the Total Risk Requirement (“TRR”)². The TRR can be further broken down into liabilities (insurance) and asset risk requirements, referred to as C1 and C2 pillars respectively within the directive³. To reduce their risk requirements, insurers can apply a Matching Adjustment (“MA”) to a defined set of liabilities using eligible assets.

The MA is a mechanism that was introduced under the Risk-Based Capital 2 Framework (“RBC 2”) in March 2020 by the Monetary Authority of Singapore (“MAS”) for Singapore-based insurers.⁴ Designed in part to acknowledge insurers’ intent and ability to hold assets to maturity in order to match long-term insurance liabilities, the MA allows them to reduce the amount of capital they are required to hold against their long-term obligations. Consequently, insurers can enhance capital efficiency and improve profitability, whilst safeguarding the interests of the policyholders.

While assets are valued at market value, the valuation of the liabilities and its corresponding risk figures are determined by discounting the liability cashflows by the risk-free rate⁵. The MA allows for a positive adjustment (“MA spread”) to this spot risk-free discount rate, bringing down the ascribed value of the liabilities and by extension, the capital requirements⁶.

One of the tenets of MA is to recognise the economic reality of asset-liability management where assets purchased to back the liabilities are expected to be held to maturity, and that there is reasonable⁷ expectation for the assets to not default. The cashflow of these assets – usually bonds – are exact and predictable⁸ like their projected liability counterparts. Unlike liabilities however, the assets employed tend to have a credit spread component and are discounted at the asset yields for valuation (as opposed to risk-free rates), leading to significant difference in interest rates sensitivities between the two sides of the balance sheet, and subsequently volatility in the CAR and FSR metrics when interest rates or credit spreads move. By referring to the actual assets held by the insurers, the MA aims to provide an appropriate adjustment that allows the liabilities to be further discounted by a rate similar to the credit spread of the assets, less an adjustment for possible defaults and downgrades. This effectively allows liabilities to mirror changes in the market value of assets that are not related to default or downgrades, dampening balance sheet volatility and providing a more meaningful representation of the economic position of the insurers.

Fig. 1. Illustration of how the MA spread is derived from the credit spread of the bonds

Source: Eastspring Investments. January 2025.

Through the lens of asset managers, the MA Portfolio is viewed as a portfolio construction problem. Insurers must meet a list of conditions⁹ to apply MA to their liabilities and these have a direct implication on the portfolio construction process:

Fig. 2. Impact on portfolio construction

Source: Eastspring Investments. January 2025.

Certain conditions stated in the framework have indirect impacts as well, leading to MA portfolios being naturally biased towards having diversified holdings¹⁶ (in terms of issuer, sector, country and credit concentrations), which shows up in the portfolio construction process as weight constraints.

The usage of the terms “constraints” and “objective” in the table above are very much deliberate as the cashflow matching portfolio construction process is an optimisation problem that can be formulated as a linear program¹⁷, which is an optimisation specification that contains linear objectives and constraints¹⁸. Put together, the aforementioned specifications allow one to solve for a minimum value portfolio that meets all of the requirements and gives the insurers an MA spread that can be applied to their liabilities with the least amount of capital allocated. However, like most things in life, the devil is in the details: the setup of the linear program requires significant amount of data – a challenge compounded by a potentially large asset universe, and a well-formulated problem that will be able to account for nuances such as turnover constraints, cashflow reallocation¹⁹, market impact and hedging can result in a solution space that grows quickly, making it difficult to solve.

Optimisation is a systematic portfolio construction method because it is completely rule and logic based, i.e. deterministic. The output of the process is a set of bonds that can, under ideal conditions, be turned into a portfolio. However, conditions are never ideal, and it is still prudent for the fund manager to re-evaluate the bonds considering market developments to assess whether it is suitable for inclusion in the portfolio.

The above is an extract of a whitepaper “Practical guide to Matching-Adjustment portfolio construction”. Please reach out to an Eastspring sales representative if you would like to have more information or if you would like to have a discussion on how a MA portfolio can be implemented to meet your specific needs.