I structured this latter discussion around four questions. What went wrong? Is the worst over? What are the lessons for financial institutions? What are the lessons for governments? Here then are my current answers to these questions.

What went wrong? The short answer: Minsky was right. A long period of rapid growth, low inflation, low interest rates and macroeconomic stability bred complacency and increased willingness to take risk. Stability led to instability. Innovation – securitisation, off-balance-sheet financing and the rest – has, as always, proved a big part of the story. As Minsky warned, undue faith in unregulated markets proved a snare.

This is the rake's progress enjoyed by the US over the past decade. But it has not been alone. Both equity and house-price bubbles affected parts of Europe, as well. But they were particularly important for the UK.

Is the worst now over? Certainly not. Unwinding of excesses on such a scale involve four giant processes: the fall of inflated asset prices to a sustainable level; de-leveraging of the private sector; recognition of resulting financial sector losses; and recapitalisation of the financial system. Making all this worse will be the collapse in private sector demand, as credit shrinks and wealth falls.

None of these processes is even close to completion. Some have barely begun. In particular, property prices are still falling, even in the US. Similarly, the adjustment in the real economy, particularly the inevitable rises in household savings rates in the US and UK are at an early stage. Because even uninformed people understand how uncertain the outcomes are, fear is pervasive. This is demonstrated by, among other things, the high spreads on interbank loans over expected official rates.

The biggest outstanding question is whether government-led rescues of undercapitalised financial systems will be needed. This is now looking increasingly likely. In today's world, governments rescue such crisis-hit economies in four ways: they offer generous lender-of-last-resort liquidity, via central banks; they run huge fiscal deficits, to offset the shift of the private sector into financial surplus; they substitute public debt for private debt, in order to recapitalise undercapitalised financial systems (often after outright nationalisation); and they may adopt inflationary erosion of the value of private (and public) debt. All of this is now likely, including, alas, even the last.

What then are the lessons for financial institutions? Stable doors are being shut after herds of horses have bolted. The Institute for International Finance has, for example, produced an excellent report on the things the financial industry ought to do (or, better still, ought to have done).**

This report focuses, properly, on risk-management (which was a disaster), compensation (which was grotesquely irresponsible), the originate-and-distribute model (which was rife with irresponsibility and fraud) and so on and so forth. No doubt people scarred by this crisis will take such advice seriously, for a while. But some years from now – 20, if we are lucky, less than 10 if the fallout is contained by the authorities – it will be ancient history. In deregulated financial systems crises are inevitable, like earthquakes on a fault zone. Only timing is uncertain.

What, finally, are the implications for governments now? The questions are two: how to restructure regulation for the long haul; and how much of their crisis tool box to use now.

John Kay argues that regulation must be restricted. His argument is based on two propositions: first, the payment system is the core financial utility; and, second, regulators cannot successfully second-guess the decisions of huge institutions staffed by better-paid and more highly motivated people than themselves.

Governments should, he argues, not even pretend they can make the financial system stable. They must, instead, try to “insulate the real economy from consequences of financial instability”. The latter, he suggests, can be achieved by insuring small deposits, creating a special resolutions regime for banks and making the deposit insurance scheme a preferred creditor.

I find John's position both attractive and unrealistic. A compelling reason for the latter view is that governments rightly define provision of financial intermediation and insurance as essential utilities in the modern economy. Another is that it is impossible to protect the real economy from a breakdown of the credit system. For this reason, governments cannot credibly promise to wash their hands of a financial breakdown. This is the lesson of at least a century of financial history.

Greater regulation is, alas, inescapable, even if doomed to be imperfect. Two steps must be taken.

One is to look for simple rules to improve the operation of the system as a whole, the obvious one being counter-cyclical capital requirements.

The other and far more controversial step is a shift in the psychology of supervision away from the presumption that institutions know what they are doing. In particular, far more attention must be paid to behaviour that may appear rational for each institution, but cannot be rational if all institutions are engaged in it at the same time. Financing house-price bubbles with loans equal to 100 per cent of the barely appraised value, because prices only go up, comes to mind.

Today, however, the authorities must also ask themselves whether what they are doing will make the system safer after the crisis is over. By these standards, the decision not to bail out Lehman looked right. But it was also risky, because we have to get through the crisis. Let us hope the decision proves to be part of the solution, not an aggravation of the challenges we face. I would now take no bets on this benign outcome.

* http://podcast.ft.com; ** Final Report of the IIF Committee on Market Best Practices, www.iif.com