Rumour has it, a senior property analyst told me last week, that at least one big UK bank is no longer issuing loans on commercial real estate, or even rolling over existing ones. If true, I am not particularly surprised. After all, the banks' exposure to the sector is at record levels and property values are sliding.

But this raises a wider issue. Suppose, even when the immediate credit crisis has subsided, the banks are too enfeebled to resume normal lending. Where does this leave corporate borrowers overall?

According to the UK's Association of Corporate Treasurers, the question is not pressing. Treasurers, they say, take a longer view than observers sometimes realise.

Until last summer, companies were filling their boots with cheap debt on the longest possible terms. Assuming their cash flow holds up - a crucial proviso - the ACT reckons they are safe for at least a couple of years.

But let us give the bearish screw another turn. Suppose in two years the banks are still in disarray and their lending still constrained. What then?

Presumably, the burden of financing corporates will fall on real money investors - pension funds, sovereign wealth funds and the like. In other words, corporate bond issuance will have to pick up sharply. And the investment funds will have to resume the habit of buying those bonds themselves. Indeed, there is evidence they are already doing so, through private purchase rather than on the market.

But for all that to work smoothly, the ACT argues, one condition is ultimately essential. The ratings agencies must be rehabilitated and get back to their proper job.

Whipping boys

On the face of it, this is a surprising notion. Throughout the crisis, the agencies have been whipping-boys. They got subprime wrong, and they messed around with structured credit. Above all they face a central conflict of interest, since they are paid by the entities whose securities they rate.

Some of that is fair, but not all. Their conflict of interest has at least the advantage of being simple and transparent, unlike those of the investment banks.

As for structured credit, criticism may be not so much wrong as misdirected. The agencies' job, as they themselves plaintively assert, is to assess the risk of default. And the actual default rate on complex derivatives has been no worse than predicted.

What has been catastrophic is the price. But it is not the agencies' job to predict market prices. If it were, they would be redundant.

In short, they stand accused of not performing more tasks than they are paid to. But as one agency executive points out, no self-respecting institution would buy equity in a company just because an analyst put a "buy" rating on it. That is only one piece of information. The institution has a fiduciary duty to analyse the stock further on its own account.

There is a degree of irony here. At the height of the credit boom, it was argued that the advent of credit default swaps (CDSs) did the agencies out of business. Corporate bonds themselves might not be traded often enough to give a reliable price. But the CDS price, being much more liquid, embodied all known information.

Today, by contrast, CDS prices are arguably telling us nothing beyond the fact that an awful lot of people want to buy credit insurance and very few want to sell it. So credit investors, thrown back on doing their own sums for a change, are venting their feelings on the ratings agencies.

But in fact, the agencies' position is arguably still central. Companies, it appears, actively prefer to discuss their affairs with the agencies rather than the banks. The latter are out to push other services on them and are also so internally conflicted that there is no guarantee what will happen to the information.

All the foregoing, of course, is based on a bearish premise. But there is a wider issue, to do with the future of securitisation and the banks' model of origination and distribution.

Central banks are not at all happy with the way this system has been working. Suppose, for instance, banks were obliged in future to retain the first-loss slice of a securitised loan, as they often did voluntarily in the past. The resulting risk weighting might be such that the securitisation game was no longer worth playing.

In other words, the next few years might see the banks retreat to old-style direct lending, on a necessarily smaller scale. Institutional savings, meanwhile, would be channelled directly to corporations through the bond markets. Old skills would have to be re-learnt on all sides.

And a good thing too. In the long run, what matters is that contact is re-established between ultimate borrowers and lenders. So long as it happens in time.