If there is one thing that buyers and sellers in structured finance can still agree on it is that mark-to-market accounting has utterly failed. The fact that investment banks have clubbed together to create a fund to mop up forced sales from ABCP conduit and SIV vehicles indicates the extent of the destruction it has caused. Much as the banks involved do not want this action to be compared with the bail-out of LTCM, the objectives are the same: to prop up the value of assets and avoid a fire sale.
This move is prompted not only by the banks desire to avert a total wind-down in the SIV sector but also by the fact that they know the assets are worth much more than their marks suggest. SIVs bought at the very top of the capital structure and (with a couple of exceptions both of which have paid the price for it) had limited exposure to US sub-prime. But, according to Moodys, SIV capital Nav fell from 102% in June to 85% at the beginning of September.
To say that SIVs were not modelled to deal with the type of market disruption that has occurred is an understatement; they were structured to withstand a maximum of five days disruption in the CP market. But because of the liquidity crunch in the ABCP sector they have been faced with selling high-quality assets at an average price of about 98.7%, according to Fitch. And they have been faced with selling a lot of these assets: Moodys reckons that SIV holdings fell by about $75 billion to $320 billion in September and the ABCP market has fallen by $275 billion on a seasonally adjusted basis since July.
That is a very big hit for the market to have taken and it has taken it because of where those assets marks are. And those marks are where they are not because of credit risk but because of liquidity risk which no one in the market had deemed sufficiently significant to worry about. So how should this risk be priced? One suggestion is that in future the marks need to reflect jump-to-liquidity risk in the same way as the corporate credit market takes into account jump-to-default risk. Few now argue that there is no need to factor in the risk of liquidity changing suddenly before the market has had time to factor the increasing risk into current spreads, as that is exactly what has happened. Before this crisis, no one was being paid for jump-to-liquidity risk, as no one was asking for it. ABS spreads now need to reflect that risk and investors should be paid for taking jump-to-liquidity risk that, as we now know, is all too real.