For what is described - too glibly at times, I think - as a blue...

For what is described - too glibly at times, I think - as a blue chip stock, QBE is quite an enigma. For starters, like all general insurers, the way it accounts for profits is somewhat arcane, in my view. In essence what an insurer does is it sells call option to its customer, whereby if that insured experiences a problem, it exercises its option (the claim) and the insurer pays out. Otherwise the insurer keeps the premium and also the returns it makes on it. There are a couple of subtle - and not-so-subtle - nuances that lie interdispersed within this "option transaction". Firstly, often the Insurer chooses not to be the only guy in the room when the inevitable claims start rolling in. It has funky proprietary systems that tell it how to parlay off some of the risk to another insurer - called the re-insurer. So, the original insurer now in turn becomes the insured in respect of a certain part of its book. In the P&L the total amount received from selling insurance (Gross Written Premiums, or GWP in insurance analyst parlance), is reduced by the cost of re-insurance to leave Net Earned Premium, (NEP). Pretty simple so far, I hope. Its a akin to a net revenue line for a "normal" company, after paying for sub-contractors. Now it comes to the equivalent for the Cost-of-Goods-Sold line for the insurance industry, or the claims line. In any given year, claims roll in (Gross Claims), which the insurer needs to settle. But remember that some of this has been re-insured, so these Gross Claims are reduced by Reinsurance Recoveries, to yield Net Claims. Then there need to be deducted "Cost-of-Doing-Business" type items such as commissions paid to brokers, and fixed costs such as the employee costs, costs of maintaining offices around the world, and then advertising, travel, etc. Deducting these costs as well as Net Claims from NEP results in the Underwriting Profit. This is a measure the profitability or the core insurance part of the business. Next comes the Investment Income on policyholders funds (or what Warren Buffet calls "Float" income, i.e., the return that the owners of insurance companies get from the free use of other people's money). Investment Income is normally very material for an insurer relative to Underwriting Profit (for example, in QBE's case Investment Income averages from 55% to 65% of Underwriting Profit). So then Underwriting Profit plus Investment Income equals Insurance Profit. Added to Insurance Profit is Investment Income on Shareholders Funds (as opposed to Policyholders' Funds), to give Profit Before Tax.

Students of accounting might argue that this is not that complex at all, and is indeed quite intuitive. However, this simplification overlooks numerous aspects of an insurer's P&L that make it, in my opinion, almost impossible to forecast net profit for insurance companies, with any degree of certainty.

Let's examine some of the the individual line items in the P&L which present the greatest challenges in terms of "forecastability":

1) GWP: is derived not only from the number of policies sold/written in any given financial period, but also on the rate achieved, or the premium, which in turn is a function of the competitive dynamics of the market place at any given time. Insurance products tend to be commoditised in nature, so price is invariably the differentiating feature in the eyes of the customer, making pricing power virtually impossible. So forecasting GWP is possible, but is not overly reliable.
2) Re-Insurance Rates: similar for 1) above; the marketplace is constantly fluid, which makes for dynamic and variable re-inurance rates
3) Claims: This is by far the hardest item to forecast not just because of the unpredictibility of inusrance events, such as natural disasters and commercial catastrophies, which is the space that QBE occupies, but also because provision that are made for claims on historical policies are continuously revised as actuaries change their assumptions based on the most recent actual claims events both within the company itself, and also across the wider industry landscape. So, if claims trends show declining tendency versus actuarial expectations, the outstanding claims would be reduced in the balance sheet, and the P&L would be credited accordingly. And vice versa. This is governed by something called the Probability of Adequacy (PoA) (ie. based on the actuarial probability of a certain occurence, how well is the balance sheet provisioned to service the claims that would arise as a result). For context, QBE carries some $16bn in outstanding claims at a PoA of around 85% (conservatively set given APRA's minimum is 75%). Just a 1% move in the PoA in any period would change outstanding claims by $160m, which is a meaningful number compared to QBE's $600m to $700m in Underwriting Profit, or $1.1bn in Insurance Profit and even its $1.3bn in Profit Before Tax for 2009. Often claims releases add or detract several hundreds of millions of dollars to pretax profit for QBE. By way of a manufacturing analogy, consider it like a factory producing widgets which are sold for $x.x dollars today, but for which the exact production cost will only be known over the next 5 or 10 years. Interesting investment proposition, isn't it? Which is why I won't pay mroe than 10x P/E for QBE.
4) Return on Funds (Policyholders' and Shareholders'): Needless to say, accurately forecasting investment returns is almost impossible given capital market vagaries. Despite QBE's stated conservatism in investing policyholder and shareholder funds overwhelmingly in cash and short-dated fixed income securites, broking analysts stll struggle to come up with with good forecasts for this item, in my observations.

So, in summary, I am sure you can see that the P&L for an insurance business is a bit of a black-box outworking, I'm afraid.

Now, why do I own QBE, given all this obscurity? (For the record, I first acquired the stock after the 09/11 event when it was technically insolvent due to its exposure to the World Trade Centre and had to raise new equity to recapitalise the banace sheet. I never bought a single stock between then and April 2009, and I have resumed buying in late June this year, below $18.00)

In principle, my investment theis for QBE lies in a long-standing belief in reversion-to-the-mean. In this case it applies to some of QBE's main earning's drivers. I think that reported profitability for QBE over the past 18 months is artificially low. Normally, the forces that drive the P&L move in counter directions to others which has a smoothing effect on earnings. However, over the last two or so years, everything has moved in the same direction, for the worst, notably:

1) The commerical insurance rate cycle globally has been very subdued for a longer-than-normal period.
2) The claims experience has been unusually high.
These factors have placed pressure of the Insurance Margin which has fallen at the bottom end of management guidance.
3) Investment income has been depressed due to the reduction in asset values, but more importantly interest rates globally, and this has affected QBE becasue the bulk of its float sits in cash.
4) The acquisition strategy has paused - not completely, but to a degree - in order to sustain and improve the excess capital position. This means the virtuous cycle of High-P/E-Buys-Low-P/E-Reinforces-High-P/E has been interrupted.

My view is that some of these earnings anchors will - simply because they are currently at extreme cyclical positions - turn around at some stage in the future (presumably sometime during my lifetime, although I can't predict for certain precisely when). In the meantime I am paying sub-10 times P/E while I wait for the normalisation to occur.

So while specific forecasting is not possible, as I argue earlied, general forecasting is, I believe. Consider it in realy simple terms, QBE today reports less profit than it did in 2007, despite having undertaken $5bn in acquisition since then, which is not an insignificant amount in the context of a company valued at some $18bn today.

Finally, I don't believe General Insurance, in isolation, is what defines the QBE business model. The way QBE has really created value for shareholders over time has been by acquisition, prsided over by the company's brains trust in the form of CEO Frank O' Hallaran, CFO Neil Drabsh and Chief Operating Officer Vince McLenaghan.

Hisorically, this was effected by a P/E arbitrage mechanism whereby QBE raised capital a high P/E in order to acquire GWP on low effective P/E multiples. This worked fine during the 1990s and the 2000s, when the stock enjoyed market darling status, trading at mid-teen P/E multiples. I believe this dynamic will be restored once the stock's re-rating occurs.


Here endeth the chat.
Happy and Prudent Investing

PS. This is NOT intended to be investment advice, so do not treat it as such; rather it is intended as a contribution on the wider debate on investing in the stock market in general.