Singapore, China and South Korea have taken the lead in advancing risk-based capital (RBC) rules in their markets. Singapore launched its upgraded RBC2 regime in March 2020; China upgraded its C-Ross II rules at the start of 2022; and Korea introduced its K-ICS this year.

Other markets in the region are working to their own timelines. Hong Kong is aiming to introduce an RBC regime for the first time in 2024; Malaysia also plans to upgrade its RBC rules next year; Thailand is looking to re-calibrate its Value at Risk (VaR) confidence level in 2025¹; and Japan and Taiwan plan to move to a regime based on the global Insurance Capital Standard (ICS) in 2025 and 2026, respectively.

But it’s not just the timelines that vary. The differences in rules for calculating capital requirements are dizzying. We set out the factors used to calculate capital requirements for equity and credit risk in four Asian markets (see table). But note, these are just a fraction of the factors that insurers need to reflect in their overall solvency calculations.

Even though RBC regimes are highly fragmented in Asia Pacific, they have a similar impact. Whenever they have been introduced, insurers’ solvency adequacy ratios have either dropped or are expected to drop, significantly in many cases.

Almost half of insurers (46%) that responded to Hong Kong’s QIS2 survey noted their solvency ratios under the new rules would be below the regulatory minimum of 100%.

After Thailand upgraded to RBC2, the risk charge for stocks listed on local exchanges jumped from 16% to 25% – a rise of more than 50%. If Malaysia drives up its VaR confidence level from its present level of 75% to 99.5%, its equity risk requirement charge would likely triple.

In Korea, 19 insurance firms – more than a third of the market’s total – asked the Financial Supervisory Service to postpone implementation of its K-ICS regime owing to the pressure it would impose to meet capital adequacy requirements.

Different RBC approaches

According to a report from US consultancy Milliman, insurance risk charges accounted for between 10% and 40% of total risk charges for insurers in Asia Pacific (see table). Investment-related risks accounted for most of the rest.

Source: Milliman, July 2022

It follows, therefore, that investment departments within insurance firms ought to take the lion’s share of responsibility for capital and risk management.

Given regulators’ increasing focus on capital adequacy, insurance CIOs worldwide are considering whether to factor RBC capital requirements into their investment processes, on top of conventional risk and return factors.

Some asset managers came up with a framework to triangulate returns, volatility and capital charges to determine the optimal mix of assets that insurers should hold. Their theory is to compare an asset class’s return relative to its capital requirements and its return relative to its value at risk. It only deems an asset as capital efficient if its return divided by capital requirement is greater than its return divided by VaR – on the understanding that the regulatory charge is less than the market risk as measured by VaR suggests.

Insurance services multi-national Willis Towers Watson looks at things differently. It proposes that insurers use risk-adjusted return on capital (RAROC) to factor in capital requirements.

Most often applied to fixed income investments, RAROC is used to determine an asset’s attractiveness. Calculated as a discount rate, it compares the cost of buying a fixed income instrument and holding associated capital requirements against the present value of relevant cash flows and potential variation in capital requirements.

The mathematical formula is as follows³:

Unlike scheduled cash flow, which generally will not change unless there is a default, the capital requirements for a fixed income instrument will change over time for a variety of reasons, such as a change in the term to maturity caused by the passage of time.

If capital requirements rise in the future, it should be reflected by a higher discount rate R to maintain the equality.

Both approaches take account of RBC considerations. The former seeks to benefit from arbitrage between the capital cost required by regulations and market risk, while the latter sees capital requirements as a type of cost and so targets additional return to compensate for it.

At an academic level, it seems reasonable that insurers could reference either approach when making investment decisions. However, it may be impractical to apply them to a real investment process since they have two key shortcomings.

Firstly, both approaches apply on a standalone basis, with capital requirements calculated using asset-side information only, with neither taking liabilities into account. That may work in some scenarios, when risk charges are calculated by applying a risk factor to asset value.

However, in other scenarios, such as calculating interest rate risk charges, most RBC regimes including those in Singapore and Hong Kong, require a balance sheet shock-based approach – with the effect on assets and liabilities considered together.

Where insurers are trying to work out regulatory capital requirements, it would be a mistake to apply a method that considers only asset-side information when this is not how the solvency regime works.

For example, short-dated bonds will always be viewed as more capital efficient (i.e. lower capital requirements) than longer-dated bonds when viewed on a standalone/asset-only basis. However, when liabilities are also taken into account (i.e. where risk is measured at a balance sheet level), capital requirements are lowest when short-dated bonds are bought to cover short-dated liabilities and long-dated bonds are bought to cover long-dated liabilities (see table).

Secondly, the objective of both approaches is to find the best assets from a return vs. capital requirement perspective. But under orthodox portfolio optimisation, either from a return vs. capital requirement or a risk vs. return perspective, you can’t achieve an optimal portfolio simply by selecting the “best assets” and putting them in sequence.

Conventional portfolio optimisation almost never starts with trying to find out the “best” asset class. Instead, a holistic, mean variance optimisation is used, and that should be the case for RBC optimisation, too. In other words, finding the asset class with the best RBC Sharpe ratio is not helpful in a real investment process.

Integration of RBC optimisation

Insurance companies can apply RBC optimisation at both a strategic asset allocation (SAA) level and/or at the underlying security level.

SAA optimisation

Intuitively, the graphic below appears to resemble a conventional return/risk scatter chart. The approaches used to construct a conventional efficient frontier and an RBC-efficient frontier are similar. Optimisation can incorporate investment return, liability duration and other requirements as constraints. Insurers can then select portfolios on the frontier to meet capital-efficient return targets.

Source: abrdn, October 2022. Based on Hong Kong RBC QIS3.

However, when it comes to portfolios that seek to support liabilities with long durations, this approach needs to be modified since it usually relies on index-level information, such as average duration, to determine RBC risk charges, especially for interest-rate risk.

Given that commonly used fixed income indices have significantly shorter durations than typical liability durations of life insurance companies, this approach won’t generally produce an optimal outcome.

To solve this, insurers can create a “liability-matching portfolio” in which they seek to match liability cash flow as much as possible. They can then treat this portfolio as a separate asset class and run an SAA optimisation across other asset classes.

Security level optimisation

Optimising under RBC at a security level works best for fixed income portfolios. Under Singapore’s RBC2 regime, three factors determine the RBC risk charge for a corporate bond: the extent to which its cash flow matches liability cash flow; its term to maturity; and its credit rating.

For a single bond, these three factors can be fixed at any given point of time. However, at a portfolio level, each factor can be adjusted based on different bond allocations. Much like yield, a fixed income portfolio’s total risk charge will change in line with differing allocations to the underlying bond. In this way, optimisation can find an allocation to maximise the ratio of yield to RBC risk charge.

Need to rethink

The emergence and evolution of risk-based capital regimes means that insurers across Asia Pacific must rethink their investment processes.

At abrdn, we have significant experience of partnering insurers to modernise the way they invest. With deep expertise of optimising portfolios under Solvency II in Europe – the world’s most mature risk-based regime – abrdn is well-placed to meet the needs of insurers in Asia Pacific as they look to optimise portfolios for returns, volatility and capital efficiency under local RBC regimes.