Investors should beware the insurance magic money machine

One of the strangest things in modern finance is the convention that allows some insurance companies to conjure capital out of thin air.

It came up the other day when Britain’s insurance regulator, the Prudential Regulation Authority, suddenly announced that it was not going to publish the outcome of the stress tests it set the UK’s life insurers last year, despite having previously promised to do so.

Some might quite like to have seen these figures, given the stresses the insurers have just been put under by the market turmoil surrounding the coronavirus outbreak. But, hey ho. 

While the PRA warned that ratings downgrades and defaults might increase in future, it said that life insurers’ solvency had been resilient in the first quarter, at least partly “helped” by this magical capital-creating convention, known as the “matching adjustment”. 

This had “significantly cushion[ed]” the impact of rising spreads on those insurers. Remember the normal effect of ballooning spreads is to drive down asset values on the balance sheet, thus eroding equity capital. (General insurers, who don’t get this benefit, face investment losses of $96bn (£79bn) according to Lloyd’s of London estimates.)

So how does this magical mechanism actually work? How can a mere accounting convention protect investors from a major market fallout?

Well, imagine a simple insurer with assets and liabilities worth £100 each. The liabilities are long-dated annuity cash flows and the assets are invested in UK government bonds. The entity has no net worth. But assuming the assets perfectly match the cash flows, the annuitants’ income is notionally safe.

The matching adjustment kicks in when it shifts some of that money into higher-yielding assets. In theory this should change things: higher yields carry more investment risk. So to continue protecting the annuitants, the insurer needs more loss-bearing capital than its present zilch.

But here’s where our convention really earns its corn. Using matching adjustment, our insurer can discount its liabilities at a higher rate, reflecting the extra return it hopes to make from those higher-yielding assets. This reduces its liabilities to, say, £90. So without anyone contributing a penny, or the company retaining any earnings, hey presto, its equity “buffer” has risen from £0 to the more substantial level of £10.

No less helpful is what happens when market turmoil strikes and spreads balloon. Then our insurer gets to discount its liabilities at even higher rates, creating more artificial capital and thus compensating for falling asset prices.

As the PRA acknowledges, many UK life insurers take full advantage of matching adjustment. And how! According to analysis of five large insurance companies by Dean Buckner, a former PRA employee, these entities had £28bn of this artificial capital on their balance sheets in 2019 and a further £29bn in artificially reduced regulatory solvency requirements. That represents £57bn of matching adjustment witchcraft. The five had total stated regulatory capital of £66bn. 

Why does the regulator let them? Mainly because it puts insurers into a special bucket: one marked “buy and hold” investors. As the annuitants cannot surrender their policies, there is no risk of insurers having to sell up to satisfy redemptions. So there is no need for so much loss-absorbing capital to guard against short-term fluctuations in asset values. 

Or, at least, that is the principle. But how prudent is it? One worry is that it allows insurers to crystallise upfront a large chunk of unrealised returns. That could permit the payment of dividends and bonuses based on those “earnings” long before they are ever received (assuming they ever are). It might also drive managers to invest, not in the most appropriate assets for annuitants, but ones that attract the juiciest matching outcomes.

Then there is a further concern, related to the present turmoil. In principle, matching adjustment should only compensate for increases in spreads related to short-term price fluctuations — not the increasing chance of borrowers defaulting. 

But deciding which part of the spread is which is frankly guesswork. What’s worse, the convention is not driven by prudence, but by rules that might spit out stupid answers.

Financial experts have long questioned the underlying principle that drives matching adjustment. Martin Taylor, a former member of the Bank of England’s Financial Policy Committee, described the “actuarial convention” by which the composition of an insurer’s assets should determine the size of its liabilities as “one of the weirdest emanations of the human mind”.

With all the present uncertainties, one might expect the PRA to gate dividends until the picture became clearer. But despite encouraging insurers to retain capital, it is prepared to stand aside and permit one of the country’s biggest life insurers, Legal & General, to make a payout in two weeks of £754m.

This sets far too much store by a rule which, for any intellectual justification, does little more than structurally flatter capital positions. More caution is warranted. Matching adjustment isn’t a real cushion as the PRA would have it; it’s a mask.

Source link

Add a Comment