Jeremy Warner: What do we want to do with our banks?

Outlook: This week's Treasury committee report on the banking crisis reserves some of its more biting observations for UKFI, the supposedly independent organisation established to hold and manage the taxpayers' various investments in the banking industry.

UKFI is described as an enigmatic organisation which is both insufficiently transparent about what it does, may not be as independent of the Treasury as it pretends, and has failed to set out an adequate exit strategy for its investments.

These strictures very much reflect the cross-party make-up of the Treasury Select Committee. Some opposition MPs condemn UKFI as little more than a creature of the Treasury, and taking their cue from the chief executive, John Kingman, a former Treasury high-flier who they see as a Brownite crony, condemn its supposed arm's-length credentials as a charade.

Others come at their criticism from the polar opposite point of view – that the banks ought to be run not as arm's-length commercial organisations but as if wholly owned by the taxpayer in the public interest

Underlying it all is the belief that UKFI should be more transparent and vigorous on bankers' pay. Some believe it to be a national scandal that any bonuses are still being paid at all, and blame UKFI for failure to impose a crackdown.

These differences go to the heart of the debate about what to do with the banks. UKFI tolerates the continued payment of bonuses to key staff because attempting to ban them would severely curtail the banks operationally. If banks are to stay in business and therefore eventually command a value for taxpayers, they have to remain competitive on pay, even though they don't seem to deserve a penny.

To publish lots more information about what the traders are being paid, as demanded by some MPs, would similarly put these organisations at an immediate commercial disadvantage.

For better or worse, the Government has decided against outright nationalisation of the banks. It also hopes eventually to sell them back to private investors for a profit. If the most profitable bits are undermined in the meantime, reducing what remains to low-paid utility operations, selling for a profit is obviously going to be quite a challenge.

Nor is there much point in UKFI being specific about its exit strategy. Who knows when public trust in the banking system will be sufficiently restored to allow for privatisation, or indeed when the optimum conditions for flogging off these investments might be? There's no point in UKFI boxing itself in, though it would obviously go some way to answering demands for more information if in the forthcoming annual report, UKFI spelt out some of the preconditions for resale to private investors.

All of which brings me back to the disturbingly humongous estimates recently trotted out by the International Monetary Fund of losses still to be realised in the banking system, the cost to the taxpayer of keeping the banks alive, and the amount of new capital still needed to repair the damage. If the IMF is right, then UKFI can forget about reprivatisation. Next stop will be outright nationalisation. Vince Cable and like-minded thinkers will finally have got their way.

In fact these estimates, though given considerable credibility by the British press, are almost certainly wrong. Indeed, they are so far out of kilter with national estimates that it is being said only half in jest that it must all be a part of the IMF's campaign for extra resources. How can you expect such a cash-starved organisation to get these things right?

The IMF first estimated the likely costs to UK taxpayers of measures to date as 13.4 per cent of GDP. After furious calls from the Treasury, it was later persuaded to reduce this forecast to 9.1 per cent, but even this looks far too high. The IMF's estimate of writedowns to come and the amount of new capital the banking system still needs to raise to return ratios to normalised levels are similarly over the top. In any case, if they are right, then we are all doomed, for as if the country wasn't already bankrupt enough, this would certainly tip us over the edge.

Fortunately, the IMF seems to have got its assumptions largely wrong. Its modelling, particularly of the housing market, is so simplistic as to be barely worth performing at all. I don't want to say that there is no possibility at all of the IMF being proved right. As we have learnt over the past two years, anything is possible. But to be correct, the losses on the Government's asset protection scheme, where more than £600bn has been insured, will have to be off the scale.

There is no point in pretending that the assets insured are all going to come good. The whole purpose of the scheme was to relieve the banks of liability for their worst assets, so that markets could again have confidence in the strength of banking balance sheets. The Treasury is conducting a second due diligence of these assets to ascertain just how toxic they really are.

But just as the IMF assumes the worst, it's actually no more implausible to suggest that after fees and first loss charges are taken into account, the scheme may end up costing the taxpayer nothing at all, or at least not very much. I hope the Treasury can be persuaded to publish details of its due diligence. Any informed view of likely outcomes also requires access to the assumptions used by the Financial Services Authority in "stress-testing" the banks against likely future write-offs.

The US Treasury promises considerably more transparency on stress testing than we have had here in the UK. This looks like the way to go. One of the reasons the Japanese banking crisis of the 1990s lasted so long is that it was more than a decade before anyone had any proper understanding of what was inside these banks. Bad loans went unrecognised and undisclosed. The more information we have, the more possible it is for banks to have renewed confidence in each other's balance sheets and for liquidity to start flowing again.