Why do shares in insurance companies trade in relation to the net assets these companies hold on their balance sheets? The short answer is because no one has a clue how much profit they actually make so it would be stupid to value them in relation to their earnings. Using net assets as the benchmark is the make-do alternative.
For this, we can blame the nature of insurance. It is what happens when a company is in the fortunate position of getting its income upfront yet faces the uncertainty of not knowing the costs of acquiring that income, or when – and even whether – those costs will be incurred.
So an insurance company receives a premium of £1,000 on day one of a three-year policy and has money in the bank. In the absence of a claim, the money remains in the bank at the end of years one and two and only at the end of year three does a valid claim for £5,000 come in. As a result, a policy that was beautifully profitable for all but a few of its 1,095 days ends up as a loss maker. Meanwhile, the insurance company is simultaneously selling many similar policies whose outcomes range across the gamut of expectations. A combination of experience and the law of large numbers tells the insurer what its aggregate profit from all these is likely to be. That’s a help, but likelihood and reality never perfectly match and sometimes get hopelessly out of line. Thus profit is another word for intelligent guesswork and net assets become the choice for valuation purposes.
Better accounting standards may be able to solve this inherent difficulty, but don’t bet on it. The world’s chief accounting-standards setter, the IFRS Foundation, is tackling the issue with IFRS 17: Insurance Contracts. Yet the standard has been 14 years in the making and its implementation has been postponed twice – it is now due to take effect from January 2023. This suggests insurance companies are as keen to implement it as they are to insure long-tail health hazards.
Meanwhile, the insurance industry is changing anyway. There is a trend for insurance companies to be less like conventional all-purpose insurers. Granted, this does not obviate the need to find better ways to account for insurance contracts, but it may mean that some insurance companies – in their changing forms – will become easier to value.
At least since the US sub-prime crisis of 2008-09 there has been a trend towards specialisation in insurance brought about by a combination of ultra-low interest rates and tighter regulatory requirements, which effectively mean more capital is needed to back a given amount of risk underwritten. As a result, globally the insurance industry looks increasingly like a giant version of Lloyd’s of London where players are neatly divided into their own niche – those who provide the capital, those who decide which risks to underwrite and those who provide the market’s specialist functions, which increasingly get lumped under the headings, ‘managing general agent’ or ‘program manager’.
Which is where Randall & Quilter Investment Holdings (RQIH) comes in and Bearbull’s interest gets personal if only because its shares are held in the Bearbull Income Portfolio. Randall & Quilter’s bosses label their company “a unique global speciality insurance company”. More to the point, it splits its functions into two. The first is underwriting so-called legacy insurance, where it buys books of existing policies – hopefully for less than their ultimate realisable value – from those who want to quit the underwriting game. Second is program management, which is essentially the role of a full-service broker, who can do pretty well anything that’s needed to run an insurance operation.
It is difficult to say which is the more profitable of the two sides because they operate according to different swings in the insurance cycle. This might be good for diversification but it makes assessing performance tricky, a dilemma that seemed to be summed up on page 5 of the company’s latest annual report. Juxtaposed yet highlighted on the same page are the operating profits of legacy insurance shown in sterling and those of program management displayed in US dollars. Confusing or what?
Still, the good news is that the confusion may be drawing to a close. Randall & Quilter (henceforth R&Q) is in the process of changing the nature of its underwriting operation to one where the capital and the risk are shovelled into a so-called ‘sidecar’. In this, outside investors put up the capital and the collateral in return for premium income but without the bother of having to clamber through various regulatory hoops.
In R&Q’s case, it is forming Gibson Re with $300m of outside capital (about £220m), enough for the vehicle to hold $2bn-worth of insurance reserves (ie, potential claims). So, for the next three years Gibson will re-insure 80 per cent of the new legacy insurance R&Q takes on and, in return for managing this operation, R&Q will receive annual fees worth 4.25 per cent of the reserves.
The effect should be that the group’s profits are increasingly driven by predictable fee income. That is already the case for the comparatively new program management side, which made its first profit – $3.4m-worth – in 2020. Eventually fee income – rather than volatile underwriting profits – should be the norm for legacy insurance, too (though R&Q will still bear 20 per cent of future underwriting risk). The company’s executive chairman, William Spiegel, suggests that, if all of Gibson’s capital is deployed by 2023, then the legacy insurance division should be generating annual fee income of about $50m. Extend that conjecture to the whole group and management suggests that fee income could be $140m in 2023, producing operating profit of $90m.
That could transform the value of R&Q’s shares. In crude terms – and at current dollar exchange rates – that could feed through to earnings per share of about 15p. The question is, on what sort of rating should those earnings be capitalised? That depends on investors’ perception of R&Q. Should it be rated as an insurance broker, an asset manager or maybe just a speciality insurer?
Think of it as an insurance broker – a mini Marsh & McLennan (US:MMC) – and a multiple of 20 times earnings or more might seem plausible. In which case, the share price could be around 300p compared with the current 167p. If that rating seems a bit too rarefied, try that of an alternative asset manager, putting R&Q in the same bracket as, say, Blackstone (US:BX). Even then, the rating – and implicitly the share price – would only be a touch lower. Last, and least glamorous, R&Q might simply be rated as a speciality insurer, something like London-listed Hiscox (HSX) or Beazley (BEZ). In that case, we would be back to estimating value in relation to tangible net assets, probably around two times. But even that multiple would raise R&Q’s per-share value to about 230p, 35 per cent above the present level.
Sure, there is a contrary view – since all of this value concealed within R&Q is known (how else could I be writing about it?), then why isn’t it better reflected in the share price already? After all, juggling around with the vehicle into which assets and liabilities are dropped may affect the timing of cash flows, but it should not affect their eventual quantity. R&Q simply plans to swap its lumpy and hard-to-predict cash flows for smoother, more predictable ones. But if the discount rates for risk are correctly attuned, then the present value of the future cash flows should be the same for both cases.
Besides, there is no suggestion that R&Q’s legacy insurance operations are failing to meet expectations. Its management targets a 15 per cent return when it mulls the purchase of a book of policies, yet over the past five years it claims a 20 per cent return on the transactions it has made. That’s an excess return that few would sniff at.
But even if R&Q’s bosses are playing a game of smoke and mirrors, what shouldn’t be in doubt is that the new model will require less capital for R&Q than the old one. For shareholders, this should mean the slightly tiresome sight of seeing their interest diluted by regular issues of new shares becomes a thing of the past. It should also free up more retained profits to be distributed as dividends. True, the dividend is now well below the level of the late 2010s, but management aims to build on the 4p paid in 2020. In the long run the aim is to distribute something between a quarter and a half of each year’s operating profit. For 2022, City analysts reckon the level will be a touch over 5p, producing a yield of 3.0 per cent or so at the current 167p.
Granted, that does not put the shares in high-yield territory, so I might ask why are they still in the Bearbull portfolio? Partly it’s because I don’t have alternative uses for the £22,000 tied up in the R&Q holding; partly because the income portfolio is on course to generate a yield of about 4 per cent this year anyway, so R&Q’s short fall matters less. And it’s not as if the investment has been any sort of disaster. In the three-and-a-half years that the income portfolio has had the holding, its total return has averaged almost 7 per cent, which is not far short of my target return. With apparently-hidden value about to hove into view, now might be an odd time to sell.
Besides, there is the more pressing matter of finding a home for the £18,000 to be re-invested when the cash from private equity’s approved offer to buy drinks distributor Stock Spirits (STCK) is distributed late this year or early 2022.
One flippant thought might be to put that money into a life insurer. At the very least it would mean I could discuss how IFRS 17 tackles the complexity of accounting for life-insurance policies. If you like intellectual challenges, there’s one to test the powers of concentration. Perhaps the idea isn’t even that silly since shares in, say, Legal & General (LGEN) or Aviva (AV.) are hardly short on yield. On 2021’s likely payout, Aviva’s yield tops 5 per cent and Legal & General’s exceeds 6 per cent. Then again, somehow I think there will be better candidates out there. One to be discussed in the coming weeks.
