the king of credit indices casts doubt on its own products

the credit crunch has been good to markit group. the company, spun off from toronto dominion, a canadian bank, in 2003, has built a commanding position in the credit-derivative indices used to value fiendishly complex assets—hence the name. as trading in such products has dried up, investors have come to see those indices—such as the ABX (for subprime mortgages), the CMBX (commercial mortgages) and LCDX (leveraged loans)—as oases in a liquidity desert. banks use the indices to help them calculate write-downs, or to hedge their exposures. the ABX is popular with hedge funds keen to take a view on the housing market.

the indices' relentless fall has added to pressure on banks, such as mer and c, with big mortgage-related holdings. citi's shares slid to a nine-year low last week on talk of a further, $18 billion write-down. banks that mark assets far from where the indices trade incur the ire of their auditors.

the indices are undoubtedly useful, but some people think banks are putting too much emphasis on them. they capture only a narrow slice of the market: the ABX references 20 securities, for instance. and they are prone to distortion (mostly downwards) by heavy speculation. "they are not liquid enough to take the weight of short-selling heaped on them,” says one fund manager, who adds that the ABX “will probably be remembered as one of the most crowded hedge-fund trades in history.” one version of the CMBX implies losses more than n 30 times greater than those suffered to date, a multiple that strikes some people as implausible. dick bove, a veteran bank-watcher with punk ziegel, recently denounced markit's indices as “fallacious”. he, as do i, expects to see write-ups as their flaws become apparent.

intriguingly, markit agrees with the sceptics, up to a point. the firm is working on a report that will highlight the limitations of the ABX and champion other inputs, such as new issuance and recent trades in similar securities, that can help value credit instruments, “two years ago we had to tout [the ABX's] virtues. now people consider it to be more relevant than it should be. they are panicking, over-reacting,” says ben logan, markit's head of structured finance. he hears frequent complaints from banks on both sides of the atlantic saying that their auditors want to see entire portfolios marked to markit indices.

yet the accountants want to avoid any impression that they are going soft on clients. they fear that would open them to the sort of criticism—and legal action—that engulfed them after the dotcom and telecom crash. they are already under fire for allowing eager profit-booking in mortgage securitisations. sticking with a widely used index is safer than waving through valuations based on banks' own models. hopefully markit's research will persuade the bean-counters to take a broader view. some regulators also urge them to lighten up, especially in their treatment of structured products. relying on a single data point is lunacy - but these are crazy times.

markit's self-criticism may seem odd, but bear in mind that it is majority owned by a group of 16 banks, including such walking wounded as c and ubs. by and large, their internal models paint a less depressing picture re than the ABX and CMBX do. if the auditors can also be persuaded that the indices are unduly pessimistic, the banks may be able to get away with smaller write-downs. if ever there was a time for banks to pick holes in one of their own products, this may be it.