JP Morgan is taking the contrarian stance to the new-fangled view that insurers should be valued on tangible net asset value (TNAV).
Here’s the bank’s explanation of the TNAV valuation phenomenon:
Valuation. The market looks at the bank sector and sees that it is trading below tangible NAV. It then looks at the insurance sector and sees that in most cases it is trading above tangible NAV (stripping out DAC). This valuation disconnect obviously raises some questions.
Solvency. The market is very worried about the capital position of financials, which is understandable given that we are seeing severe financial system stress. Tangible NAV is seen as the true “loss absorbing” part of the insurance capital base, so a low tangible NAV is a concern in an environment when losses are increasing.
JPM broadly agrees with the new non-focus on tangibles. Tangibles are impaired, they say.
NAV — the difference between the carrying value of the assets and liabilities of the company — has become a focus for investors because it basically represents a valuation floor. It’s essentially what an insurance company would get if it were liquidated in bankruptcy proceedings, or forced to sell its business in the current market.
As with banks, TNAV strips out intangibles like goodwill, or in the insurance sector’s case specifically, deferred acquisition costs (DAC). This makes sense in the current environment — intangibles are essentially a claim on future earnings (the banks deferred tax assets being a good example of this). Earnings are less certain now, so why not strip out the intangible assets that rely on them?
But that doesn’t necessarily mean insurers are worth TNAV, according to JPM:
Stripping out all of the intangible assets, including the DAC, is like valuing a retailer with zero value for its inventory. Even if it were the best measure, then the subsequent RoNAVs (sometimes as high as 200- 400%) would surely anyhow mean a higher than 1x P/NAV multiple! In truth, we don’t see much wrong with the traditional approach of book value less goodwill and maybe some haircut on DAC — this still penalizes the companies for their unrealized losses on bonds.
And a low TNAV doesn’t necessarily mean insurers are bust or about to launch a rights issue, JPM says:
We believe that the TNAV has fallen mostly due to accounting reasons — i.e. the full marking to market of assets. Remember that insurance companies mark-tomarket all of their assets as they are almost entirely financial and traded (or at least used to be!). This is in contrast to a bank where the loan book is the major asset on the balance sheet, and the loan book is obviously carried at cost, not market value. This means that an insurance company shareholders’ equity is much more volatile than that of a bank, even if economically there is not much difference (i.e. the assets are basically the same, just the accounting treatment different) and, if anything, should be less volatile due to lower leverage.
The reason why book value, and thus TNAV, has fallen so much is because of the marking to market of the bond portfolio. This has meant that there is currently a very large level of unrealized losses on bonds sitting within the shareholders equity – within Aegon, for example, it is 118% of the reported shareholders’ funds. Some part of these unrealized losses will be realized, at some future point in time, but some part will not, but will instead reverse as these bonds reach maturity. In our models we assume 30-50% will be realized .
If TNAV is low because of accounting treatment, this doesn’t necessarily mean the company’s insolvent. Firstly they say, insurers match their assets and liabilities — unlike banks. Therefore they’re not overly-concerned about movements in the value of the assets from liquidity, only credit (JPM sees the current movement in the bond market as split 50:50 between illiquidity and credit). Secondly, some regulators don’t include unrealised losses on bonds in their solvency calculations. Ultimately, it’s the regulators’ opinions of the insurers’ capital position that count, JPM says.
These are two key issues. The fact that regulators don’t look at unrealised losses on bonds has been key in keeping insurers alive. Notably however, it is regulators in France, German and the US who ignore those losses, not the UK. As we noted earlier, the FSA seems to be taking a more stringent view of insurers’ capital position than before.
Plus insurers don’t necessarily mark-to-market their corporate loan books. Prudential, Legal & General and Standard Life for instance, are still clinging to IFRS rules, which allow them to mark-to-market their holdings, and then discount much of the market impact by using their own assumptions of the bonds’ long-term performance potential. A new accounting standard, MCEV, would be more stringent.
Others like Aviva and Old Mutual have already embraced MCEV. In fact, Old Mutual commented last week:
Andrew Birrell, chief actuary, said Old Mutual’s profits would have been “at least 25pc higher” if it had used the old standards. Old Mutual said there was a fear that companies not reporting under the new standard were hiding something.
A quick look at Aviva’s recent results shows an £885m after tax loss under IFRS and a £7.7bn after tax loss under MCEV. That’s a wide margin, to put it mildly.
The thrust of what we are getting at is this – TNAV is a somewhat simplistic and pessimistic way of valuing insurers. But, given current market conditions -the potential for a sudden wave of corporate defaults and continued confusion over mark-to-marketing – it may not necessarily be the worst way of valuing them either.
If you’re less conservative or more optimistic, then, as JPM suggest, by all means look at insurers trading low relative to NAV. But if you’re more of the bearish persuasion, avoid them altogether, or else look at the ones trading close to 1x NAV.
Related links:
Valuing insurers – FT Alphaville
SELL: Insurers - FT Alphaville