Writeoff Forgiveness and Goodwill (ConocoPhillips)

September 15, 2009

Benjamin Graham wrote in Security Analysis that a great deal of financial analysis involves recasting financial reports to get a full sense of a firm’s baseline ability to produce earnings. A lot of the process involves removing nonrecurring or irrelevant events from a single year’s results, as we did with LINE’s profits on its derivatives. Over a longer period of analysis, like five or ten years, we leave the nonrecurring events in or perhaps even distribute them on average over the period. Nonrecurring events still tell us how unlucky or unskilled management is over time, and even though for a single year’s earnings we may forgive a large nonrecurring writeoff all at once, a writeoff is still money gone, and there is a limit to forgiveness.

Cooking the BooksMuch of this fiddling with nonrecurrent events is the fault of accounting rules anyway. When it comes to goodwill, writeoffs take on an unusual character. Goodwill represents the purchase price paid to acquire a company that is in excess of the assets of that company. It represents the excess returns available from buying an established company instead of just replicating it, or, if you believe Damodaran, some of it comes from the “growth assets” of a company – the assets it will one day acquire to produce its growth and future excess returns. A lot of analysts advocate ignoring goodwill entirely, when it first shows up as an asset and when it is inevitably written down when conditions deteriorate. After all, if there are excess returns in a business they will show up in the income statement anyway, so counting both the excess returns and the goodwill looks like double counting. Goodwill nowadays, instead of being linearly written down over time, is now written down whenever, in the opinion of accountants, the excess returns it generated are no longer there. Since this essentially trades one accounting fiction for another, I can’t say that it’s solved the problem of dealing with intangibles, but it does make the data more frequently subject to large, discrete abnormalities.

ConocoPhillips (COP) is an integrated oil producer and refiner that wrote down $25 billion of goodwill last year, plus more than $7 billion for their interest in a joint venture, on the grounds that with the price of oil and the economy down, returns on their business and its recent acquisitions are going to suffer. With our policy of ignoring goodwill, we might say that without the writedowns they earned $15 billion last year, and earned $12 billion the year before that and $15 ½ the year before that. And since the market cap is less than $70 billion, even looking at the worst case their PE ratio was less than 6.

However, as I’ve stated, goodwill writeoffs have to go somewhere, or to be precise they have to come somewhere since the company paid for that goodwill. This year, earnings have been running far behind what the previous years’ performance suggests they would have been, only $2 billion in the first two quarters when last year they were $9 ½. I should point out that unlike Linn Energy in the last post, COP does not hedge its commodity exposure , which is understandable because COP’s production is orders of magnitude larger and the integrated nature of its  business theoretically leaves them less exposed to price changes, as with Rayonier.

We can reconcile the irrelevance of goodwill with the fact that the company has run into a wall by considering return on equity, since equity is where goodwill writeoffs come from. If we double this year’s earnings, and divide it by the book value of equity at the end of last year, we get 7.8%, which looks pretty reasonable, although since COP trades at a moderate premium to book value that is not the return on a purchase at this price. If we put the writeoffs back in and look at last year’s return on equity, though, we get 16.9%. The year before, 13.6%, and the year before that, 19.1%. So, although the goodwill writeoff made more sense than leaving the goodwill in, it hasn’t accomplished what the rules intended it to, which was reconciling earnings power with asset values.

I’m not sure how much of this is temporary and can be blamed on the recession, but it seems simplistic to me that goodwill can always be ignored. Since it represents extra money paid for an acquisition, and since its being written off represents overpayment, I think the more conservative solution is only to look at it in a negative fashion; not counting it as an asset but definitely counting it as a writeoff, at least for a multi-year analysis. And if COP is representative, jumping in after a big writeoff is dangerous. And identifying where the low-hanging fruit isn’t is just as important as identifying where it is.

Leave a Reply