The American Council of Life Insurers wants to see the effects of residential mortgage-backed securities losses on risk-based capital tailored to reflect the expected severity of the losses.

Today, RBC rules rely on RMBS ratings, and the ratings reflect only the likelihood that there may be some kind of issuer default, not the total amount of expected losses, the ACLI, Washington, says in a proposal submitted to the Valuation of Securities Task Force at the National Association of Insurance Commissioners, Kansas City, Mo.

“A more logical approach for RBC purposes, and an approach more consistent with the goals of the RBC calculations, would be to determine an estimated or modeled loss from a base set of consistent assumptions applied to the individual characteristics of each security,” the ACLI says, according to a copy of the proposal posted on the VOSTF section of the NAIC’s website.

The NAIC’s Securities Valuation Office could hire an outside firm to model losses for the RMBS securities actually held by the insurance industry, the ACLI says.

“The modeling should be conducted on a security level basis and should use assumptions generally accepted by market participants for prepayments, home price levels, expected defaults, severities of loss, and performance of loans in good standing, along with other assumptions including interest rates,” the ACLI says.

The ACLI proposal is an interesting idea, according to Andrew Hopping, chief financial officer of Jackson National Life Insurance Company, Lansing, Mich., a unit of Prudential P.L.C., London.

But, “as a result of the various assumptions and models that need to be derived, the proposed solution is a very subjective approach,” Hopping writes in a comment letter. “It raises questions about the incentives and goals of the modeler and how those incentives will influence how the process is implemented.”

Using a custom-made “analytical measure” might work better than using an off-the-shelf rating, because regulators could reduce the potential for conflicts of interest by paying for the measure calculations themselves, and regulators could have the measures updated whenever new data comes in, according to Andrew Davidson, president of Andrew K. Davidson & Company Inc., New York, an MBS risk analysis firm.

“One advantage that ratings may have over analytical measures is that they are widely available and subject to market scrutiny,” Davidson writes in a comment letter. “Therefore, we believe that analytical measures should be computed by firms that have sufficient standing in the market, that the methods and calculations are known, and that the regulator should review the analytical measures produced for consistency with other publicly available forecasts and analysis.”