Exchange Traded Funds in the General Accounts of Insurance Companies

Against a background of heightened volatility in fixed-income markets in 2020, the insurance industry has seen renewed interest in fixed-income ETFs. The National Association of Insurance Commissioners (NAIC) estimates that insurance company general accounts hold around $40 billion in fixed-income ETFs.

Special Accounting and Risk Treatment

Currently, bond ETFs have a lower capital charge level and are not considered equities for risk-based capital (RBC). In 2019, NAIC’s Capital Adequacy Task Force solicited feedback from insurers and asset managers on its decision to apply bond factors to all bond and preferred stock funds on NAIC’s approved list. They received many letters arguing for ETFs to have the lower capital factor applied to them.

As of September 2020, NAIC published a list of 166 ETFs permitted to receive special accounting treatment if held by an insurance general account. Most of these ETFs are “benchmark-hugging” funds issued by the largest passive managers: Blackrock, Vanguard, and SSGA. Then there are a handful of specialist funds, actively managing exposure to municipals, bank loans, currencies, and emerging market debt.

Issuers of ETFs Receiving Special Accounting Treatment

Sources: NAIC, FactSet

Many of these ETFs had a great year both from the returns and flows perspective. It’s important to note that most ETFs on the NAIC list are market-value-weighted funds.

Market-value-weighted bond ETFs may expose investors to companies with the highest debt issuance. Beyond that, passive ETFs can also expose insurers to bonds issued by their peers and counterparties.

Book Value Accounting

How should fixed-income ETFs be treated for book value accounting? NAIC-approved ETFs can be held at cost under the systematic value (SV) accounting method. It allows for the treatment of ETFs as single bonds where initial book yield can be calculated based on the future cash flows, which would be dividend income in the case of ETFs.

For example, if an insurance company purchased the fund on December 31, 2019, the last dividend can be assumed to be the future dividend and the internal rate of return of those forecasted dividends can be calculated. Applying this logic to AGG, the iShares U.S. Aggregate Bond ETF, would have produced a yield of around 2.5% on January 30, 2020, based on the historical 12-month dividends as a proxy of future cash flows.

Each month after that, the investor must review the actual income and adjust the book yield up or down based on the amount received.

The theoretical alternative, and probably a better option, would be projecting cash flows on the actual ETF holdings and using the trade-dated bond prices as the cost basis. This would have projected the book yield of AGG to be around 2.31% on January 31, 2020.

Using historical dividend information, we assume that the fund will be able to re-invest at the previous rate to achieve a constant yield. Holdings-based book yield assumes forward-looking cash flows and amortization/accretion to achieve a constant book yield; thus, the difference in starting book yield.

Passive ETFs continuously rebalance their holdings to match the index they track. If we are calculating the book price as the market price on the date of the ETF purchase, then we are at the ETF’s discretion on how book yield is managed. In our example, the net book yield contribution for a sample one-month period is 9bps as shown in the table below. The 9bps can be further reduced by the management fees of the fund, which are luckily not too high for index-tracking ETFs.

For an insurance company to decide whether ETF investing adds value, examining book value attribution can be illuminating. However, this requires daily or at least monthly updates to the positions. This can become operationally difficult due to a lack of data and technology.

Conclusion

The jury is still out on bond ETFs in the insurance general account. The $40 billion invested in them is not a large number comparing to the overall insurance AUM. ETFs are an easy approach to parking the premium inflows and don’t have a significant impact on capital charges. However, the unwanted exposure they introduce, the lack of alpha they add from the book yield perspective, and, of course, the operational headaches associated with accounting for them are all items to consider. Which of those pros and cons is meaningful remains open for debate.