L&G have been much in the news recently, so too has our own Dean Buckner, with star appearances in the Financial Times, the Times and the Daily Mail (see also here) and all in two days.
The fur continues to fly over L&G’s planned June 4th dividend, but the attention is shifting subtly from the dividend itself to the underlying valuation methodology and the central issue is (one again) the Matching Adjustment.
There is also the issue of the firm’s ‘virus spread’ losses or, more precisely, Dean’s estimate that these could be up to £10.3 billion. Dean has stirred up a right hornets’ nest this time.
Here is my take.
Let’s start with L&G’s own numbers. According to their latest Solvency and Financial Condition Report, their Eligible Own Funds (EOF) is £16.9 billion. (EOF is the Solvency II definition of capital, which bizarrely includes sub-debt, but there is Solvency II for you.) Their Solvency Capital Requirement = £9.4 bn, so the coverage ratio = 16.9/9.4 = 179%, which is wonderful news because this coverage ratio equates to an annual failure probability of 0.0002% and an expected waiting time to default of over half a million years. But if you strip out the £16.8 billion capital created by MA and Transitional Measures for Technical Provisions (TMTP) then EOF becomes £0.1 bn, the SCR becomes £23.3 billion and the coverage ratio becomes 0.5%. This coverage ratio implies an annual failure probability equal to, er, 50% with an expected waiting time = 2 years.
My point is that a lot hinges on your view of the validity of the MA and TMTP. If you accept their validity, then the company is in fine shape. If you do not, e.g., because you regard applying MA as being equivalent to asking the bookie for ‘dead cert’ winnings before the horse race starts, then you might wish to temper that assessment somewhat.
Therefore, L&G’s EOF capital is £16.9 billion if you accept the validity of the MA and TMTP capital, or it is £0.1 billion if you do not, or it is somewhere between, if you accept one of MA and TMTP but not both.
What is their IFRS capital, you ask? The answer is that L&G report IFRS capital to be £9.4 billion in their 2019 Annual Report.
What happens to their IFRS capital if you strip out the MA and/or TMTP capital? The answer is that we don’t know because apparently L&G won’t say.
The counterargument, however, is that
“If a firm is in a strong financial position then there is no harm to shareholders in deferring dividend payment, which will be kept as retained earnings. If it is not in a strong financial position it would be wise, and in the interest of shareholders and other stakeholders, to conserve capital. A dividend that, in retrospect, could not have been afforded, causes major damage for the company and all of its stakeholders going forwards. In the current crisis, therefore, a dividend payment is not in shareholders’ interests.
“Based on the default sensitivities stated in L&G’s 2019 annual report, and the current level of corporate bond spreads, the mark to market losses could be as high as £10bn. However, spreads vary widely by sector, and without knowing the sectors they are exposed to, only the firm will know the precise number.”
In the absence of further disclosure, the best I can do is extrapolate from the information we have. If EOF is £16.9 billion and IFRS capital is £9.4 billion, then their difference is £7.5 billion. So if one subtracts this difference and the MA and TMTP capital from the original EOF of £16.9 billion, then one obtains an estimated IFRS capital of £16.9 billion – £16.8 billion – £7.5 billion = minus £7.4 billion.
It is important to emphasise that this number is an extrapolation: only Mr. Stedman and his colleagues can give us the true number and they ain’t sayin’.
All the numbers quoted so far apply to year end 2019.
Then there are L&G’s ‘virus spread’ losses which have occurred since then and so we come back to Dean’s famous £10 billion.
I have seen Dean’s spreadsheet and it looks reasonable to me: in essence it was a standard (i.e. respectable) delta extrapolation based on the information that L&G provided in their Annual Report. Apart from the fact that L&G don’t like the result, there is nothing controversial about it.
I did my own triangulation exercise as a cross-check. The number of L&G shares outstanding at the end of 2019 was 5.97 billion. Their share price at that time was 303p and their share price as of Monday was 187.6p. Therefore, their market cap at end 2019 was 5.97 billion times 303p or £18.1 billion and their market cap on Monday was £11.2 billion. Take the difference to be a rough estimate of the virus spread losses (like what else could it be?) and you obtain an estimate of the latter equal to £6.9 billion.
Dean’s estimate seems to have struck a nerve at L&G, however – Hosking’s article says “Over the weekend L&G put pressure on the UKSA to retract its estimate but it declined to do so.” They then publicly rubbished the Buckner/UKSA estimate. “We simply do not recognise these numbers,” said an L&G spokesman.
“It [L&G] has also stressed in briefings to analysts that very few of the bonds are issued by companies now bearing the brunt of the pandemic downturn such as retailers, the oil industry and hospitality,” said Hosking in his Times article.
L&G’s 2018 Annual Report sheds some light on these exposures: £3.5 billion in consumer cyclicals, £7.4 billion in non-cyclicals, £2.3 billion in energy/oil and gas, £3.5 billion in real estate which could be a lot of empty malls, £10 billion in various ‘infrastructure’, £1.8 billion in “Structured finance ABS/RMBS/CMBS/Other” and £4.7 billion in equity release, which, as everyone knows, is safe as houses.
Let’s take L&G at their word when their spokesman said: “Our decision to pay our recommended dividend is considered, prudent and affordable. It is the right thing to do for our thousands of personal shareholders, for the pension system and the economy.”
“… considered, prudent and affordable” is the key phrase. So why doesn’t L&G just release the evidence that leads them to this happy conclusion?
Absent that, there is a strong case that L&G’s forthcoming dividend should be gated, as Sir John Vickers and others have argued, on the grounds that there is no point locking the stable door after the dividend has bolted.