Insurers reliance on ex-post risk metrics to determine investment risks concerning - research

Written by Adam cadle

04/12/2019

Insurers are stating they have a “good understanding” of investment risks in their investment portfolios, despite most having a heavy reliance on ex-post risk metrics, new research has shown.

New research conducted in September 2019 by bfinance with members of the Association of Financial Mutuals (AFM) in the UK and Ireland, showed that while the vast majority believe that permanent capital loss is the most important definition of risk, more than three in five “consider historic realised volatility to be a good proxy for current portfolio risk”.

“Historic data on realised risk has a useful role to play in assessing investment risk,” the paper stated. “Metrics based on ex-post data are straightforward to estimate, since the only input is historic returns. Yet their convenience cannot make up for their lack of robustness: inaccurate estimation of risk is likely, due to the reliance on a single sequence of historic outcomes. Historic measures also take a long time to build up enough data for risk analysis. For monthly reporting, one month only means one more single observation.

“In contrast, implied metrics are typically more accurate measures of risk that seek to consider a wider range of outcomes at any point in time by simulating thousands of implied portfolio returns, Implied risk measures allow us to form a current view of risk based on current portfolio exposures."

The paper added that insurers running multi-asset portfolios to generate appropriate returns must go further and additionally be able to examine risk exposures in terms of ‘risk premia’: long-term drivers of return. These include market risk premia, macro risk premia, style premia, to name but a few.

Related Articles