Partnership Assurance (PA.L) started trading on June 7th and is currently valued at over £1.8bn which is an trailing IFRS profit (EBIT) multiple of over 15. Whilst they are on the face of it a reasonable firm in a high growth sector, in this article I will put a case for why this valuation is concerning overstated.
The essence of my argument is that either the valuation is being performed on the wrong metrics or that those with access to full data have built in goodwill to the valuation is simply unjustifiable.
Valuing insurance companies is tricky at the best of times; and that assumes you have as much information as possible available. The difficulty revolves around the fact that:
a) Most of the firm's assets belong to the policyholders; and
b) The profits on a lot of life insurance business take a VERY LONG TIME to emerge.
Couple this with frankly bizarre accounting methods and the result is that valuing a life insurer on a multiple of its most recent account profits is potentially a massive mistake.
Partnership Assurance (PA.L) have an interesting and elegantly simple business model:
1) sell specialist annuities (regular payments of cash pensions, paid for life) to policyholders with medical or lifestyle features that shorten their expected lifetime. Whilst this is a niche of the market it is extremely competitive, these products are sold on extremely thin margins.
2) Sell some other life insurance and long term care products
3) Use the proceeds from 1) and 2) to provide equity release mortgages (less competitive, much higher margins); and to invest in corporate debt, and other instruments to support their annuity business
4) Sit back and watch (hope?) the whole lot unwinds to give you some lovely profits (oh, and obviously make everyone's retirement lovely too)
Ignoring rather subjective things like future new business (=Goodwill) for now; obvious risks include:
· Annuitants might end up living much longer than estimated (so PA may pay out in pensions more than the premiums they received and the investment income earned on those premiums).
· The embedded option risk in the equity release mortgages comes home to roost. This happens if house prices crash, if PA.L have been lending at high loan-to-value ratios (which they do, citing their expertise in mortality modelling to help mitigate the uncertainty) it may even be enough of a risk for house price inflation not to keep pace with the interest rolling up on the equity release mortgage.
· Lots of defaults in the corporate bond portfolio.
Let me repeat this again. Life insurance is a very long game played out over a decade or so.
It's simply no good to look at a few years performance to convince yourself it's a good investment (for example p4 of the IPO document showing the "amazing" increase in IFRS profits - we'll look at those a little more closely in a second). For this reason embedded value (NYSE:EV) was developed as a means of analysing the possible value of future profits over the long term, i.e. the shareholder value embedded in the firm. Don't forget that most of the balance sheet assets of an insurer essentially belong to the policyholders. So EV gives a measure of how much of those assets may flow to shareholders over the life of the firm.
EV is far from perfect, but it's a good start and despite all its problems, is used by just about every other life insurer in the UK in their public accounts - except PA. Odd, no? I'm not accusing them of withholding information, they don't have to publish EV. IFRS profits are short-termist and potentially rather volatile - mainly due to marking unrealised investment capital gains and losses to the profit and loss account.
If the EV were available one might (if you were an actuary at least, banking analysts may beg to differ) make your starting point for the valuation of an insurer as:
Valuation = EV + Goodwill
Goodwill = [EV value of the last year's new business] * [a low multiple maybe up to 7] + other goodwill
However, PA.L hasn't given us this (I seriously hope it's in the full Prospectus). So let's try to pull apart what they have given us. Where relevant we'll compare PA to their closest competitor: Just Retirement - who do produce EV figures, but have their year-end in June rather than December. Where I can I'll try to infer consistent figures.
New business premiums: have increased by 42% year on year. Wow, that's a big number. But they are in a growing sector and appear to have had a broadly stable share of that sector over the last few years. Oh, and let's not forget that the previous financial year was not that kind to insurers. There's probably a sound argument for those growth rates being unsustainable for more than a couple more years.
Earnings before interest and tax: £106 million, an increase of 64%, impressive. They and their competitors benefited in the run up to the year end from a rush of people buying pensions just before some new legislation in the UK (RDR and the introduction of the Gender Equality Directive). Splicing together year end accounts and interims, Just retirement produced IFRS EBIT = £137 million in the year to December 2012, up from roughly £50 million the prior year. More impressive still.
Let's think about the market context of these rather large changes - credit spreads down, interest rates largely flat. Why are these important? They form the basis for the way the firm's liabilities are valued. Very broadly an insurer with annuity liabilities will be able to discount the value of those future cash outflows by the "risk free rate" (government bonds or swaps) plus some proportion of the spread of corporate bond yield over the risk free rate.
Over the year to 31 December 2012, the graph above shows that the rate that PA.L are likely to have been using to discount their liabilities (and hence value the firm) decreased by around 1%.
When corporate bond spread decrease the value of those bonds increase, in this case, a lot.
A flick through the regulatory returns of PA.L (yes, I'm that geeky) shows they were carrying an above average amount of corporate bonds rated A or below. As 2012 progressed the value of these bonds would have increased significantly. IFRS accounting is such that regardless of whether or not they sold those bonds (they didn't) those (unrealised) profits on the capital gain on those bonds get to go through the P&L.
I'm no bond trader, nor am I a specialise bond fund manager, but I think it would be fair to say that the contraction in spreads over 2012 were pretty exceptional; yet this rather exceptional event appears in the IFRS profits that appear to be the basis for valuations.
Let's draw this to a close. If we were to try to estimate PA.L's EV it would probably be of the order of £500 million.
If you're interested, here I've looked at their regulatory excess capital FSA Form 2, line 13 = £327 million, and approximated the present value of their existing business (VIF) by scaling the VIF from Just Retirement's EV returns to the size of PA.L's balance sheet. EV = Shareholders Net Assets + VIF.
So we're looking at a valuation of goodwill in excess of £1 billion. Perhaps I'm missing something - or perhaps Morgan Stanley and BoA-ML just did an extremely good sales pitch. That amount of goodwill, or even that multiple of (a one off) IFRS EBIT is very stretching. My view (this is not a trade recommendation) is that PA.L will trade significantly down relative to the UK insurance sector over the next year.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.