Self-graded exams: Insurers said to have low-balled 2016-19 loss picks

US insurers have likely whitewashed their reserving for peak-period losses mid-2018 to mid-2019 and now have the sufficient cover to gloss over reserves for the coming years, a key equity brokerage is warning investors.

“We believe that the industry remains under-reserved for the 2016-2019 accident years,” analysts at  Bank of America Securities (BofA) warned investors of insurer balance sheets and, in turn, the preferred valuation metrics built on book value.

“We tend to be fairly confident that, at minimum, the 2H18-2H19 periods will continue to evolve adversely,” analysts wrote. “That is to say, we do not believe that 3Q18 represents the peak loss ratio/underwriting trough moment.”

Analysts regularly walk in with heavy doses of scepticism.  BofA cites an age-old trend that underwriters “have been known to both suppress loss data” to pad earnings “but also, sometimes, to overstate losses” to smooth results.  The latest cycle has only the 2016-19 suppression, with no compensatory overstatement, they warn.

Press as hard as they like on investor calls, but analysts “have not been so successful in mining confessions” on low reserves. Small companies might sometimes be forced to. Large companies freely blend under-reserved lines within “a much larger and less precise set of data.”

So  BofA analysts have started plugging select management comments into their models.  BofA tracks timing of management comment about the onset of margin expansion vis-à-vis peers and vis-à-vis when they claim pricing starts trending in excess of loss cost.

Those models, built on signals that commercial P&C pricing did not begin to exceed loss costs until late-2019/early-2020 and then needed 6-12 months for new business to be earned in, say 2018 and 2019 loss picks will end up taking more reserves.

“The idea that loss ratios peaked/underwriting margins troughed in 4Q18 seems dubious in our view.”

From the comments culled from firms in its own research universe, CNA and W.R. Berkley may be under greater suspicion of low-balling reserves versus stalwarts  Arch,  Hartford or  Travelers. But that,  BofA admits, remains as speculative as everything else in discretionary accounting.

Visibility might only get worse. Actuarial models for judging reserve adequacy, never a precise model, will have been made nearly useless by the pandemic. That opens the door for much more potential gymnastics in discretionary accounting moving forward, analysts warn.

Analysts who might have tracked paid claims as a portion of reserves on prior year events won’t have good data again until 2024 as the pandemic-era shut-down of courts & settlements destroys data relevance,  BofA suspects.

"Previous so-called ‘paid-to-incurred’ loss trends have become largely irrelevant in assessing reserve adequacy," analysts wrote. “As such, we might be ‘flying blind’ with regards to the mathematics we typically use to guide our judgments.”

Did you get value from this story? Sign up to our free daily newsletters and get stories like this sent straight to your inbox.