A new measure of the scale of our big banks' recklessness in the boom years has been provided in the Treasury's annual report for the past year.

It says that it expects the public purse to incur a loss of £25bn in respect of the Asset Protection Scheme.

That's the safety net provided by the Treasury for Royal Bank of Scotland and Lloyds, which was announced on 19 January as an alternative to full public ownership of these banks.

It involves taxpayers providing insurance to RBS and Lloyds against possible future losses on £325bn of loans and investments made by RBS and £260bn of credit extended by Lloyds.

Negotiations are still in train on the fine print of this insurance arrangement. But the Treasury says that "on preliminary analysis of a sample of assets initially proposed to be included in the scheme", it estimates it could lose £25bn (as my colleague Stephanie Flanders has written about here).

However it adds that the estimate "could be subject to substantial revision (up or down) as further due diligence reports are completed" and that "altered economic and market conditions would clearly also effect estimated losses" (doh!).

Now you'll probably find this slightly odd, but I think this statement is more interesting for what it says about RBS and Lloyds than for what it says about potential costs to us, to taxpayers.

The reason is that it's impossible at this stage to forecast with scientific precision the shortfall between the loans provided by these banks and how much they'll ultimately get back from overstretched borrowers.

But here's the important point.

There would be no estimated loss to the Exchequer at all unless the Treasury was confident that the banks would use up the entirety of their excess - their liability to take the first loss - on this insurance scheme.

Or to put it another way, the Treasury would not and will not incur a penny of loss on the Asset Protection Scheme unless and until Royal Bank of Scotland suffers £42.2bn of additional eye-watering losses on those £325bn of loans and investments and Lloyds incurs £35.2bn of losses on its £260bn of insured assets.

It's worth saying that again. Royal Bank and Lloyds - which already reported record losses for 2008 - are expected by the Treasury to suffer further losses on their loans and investments of at least £42.2bn and £35.2bn, or £77.4bn in aggregate.

That's the kind of financial calamity that makes me feel slightly dizzy.

Of course, these losses will be incurred over a period of months and years. And they will be offset to some extent by profits that are being generated from the unpoisoned parts of the two banks.

But to call these losses the toxic fruit of reckless lending doesn't quite capture the magnitude of these banks' departure from sensible prudential standards.

And I ought to add that if the Treasury were right that the loss to the Exchequer turns out to be £25bn on top of these losses for the banks and their shareholders (that us again, of course, since taxpayers are set to own 84% of RBS and 62% of Lloyds), there would be further losses for the two banks of £2.5bn (losses over and above the first loss are shared 90:10 between taxpayers and banks).

As it happens, I can conclude with marginally better news for taxpayers.

Which is that the net loss for the taxpayer on the APS may eventually turn out to be zero.

Here's why.

It's true that there will almost certainly be a substantial gross loss for the public sector as insurer of those ill-judged loans and investments.

But, as with all insurance schemes, the banks are paying substantial fees for the protection. And in the case of Royal Bank, it has also agreed not to claim tax losses or allowances relating to its disastrous loans and investments.

Once those fees and tax benefits are taken into account, the taxpayer may well break even on the deal.

Which, of course, is not to argue that this is good business for the public sector.

We'd all have been better off if the banks' lending binge hadn't been so wild and prolonged.

But small virtue can be extracted from this hideous financial necessity.

Whenever I find myself in any gathering of bankers, business people or politicians, the question they ask more than any other is whether the government will be able to borrow all it needs from markets - or whether at some point big investors will lose their appetite for gilt-edged stock, the exchequer's IOUs.

A good person to tap on this - though one who is not without a vested interest - is the Treasury's banker, Robert Stheeman, the head of the Debt Management Office, whose job is to raise all those hundreds of billions of pounds to cover the gap between tax revenue and public expenditure.

This year he has to sell an utterly unprecedented £220bn of gilts. That's more than four times the finance that he's typically had to find in recent years. And it would expand the existing stock of gilts - or the size of the market - by more than a third.

And, what's more, with the government refusing to countenance significant spending cuts or tax rises, he'll have to raise a similar amount next year too (even if a new government were to massively reduce public expenditure, there would be a lag before there was an impact on borrowing).

So - I asked him, in an interview for the Today programme - how great is the risk that investors will sit on their hands? Does he lie awake at night fearing a re-run of the 1970's, when a Labour government had to be bailed out with emergency financial support from the International Monetary Fund?

Well, Stheeman doesn't have bags under his eyes and sees a funding crisis as a very remote danger - largely because the market for gilts is much wider and deeper than it was.

For example, it has become much more international, with a record 36% of gilts now held overseas.

Now, you might point out, he would say that, wouldn't he?

That said, he was much less dogmatic about whether the government might end up having to pay a much higher interest rate to borrow - which is hugely important, because an increase of one percentage in the cost of borrowing £200bn would be £2bn that wouldn't be available to spend on public services every single year till the debt is repaid.

If, for example, the UK lost its impeccable AAA debt rating, that would almost certainly push up funding costs - not least because some of those helpful overseas buyers are central banks which aren't permitted to hold sovereign debt rated at less than AAA (though it was striking that Stheeman told me that he didn't think a downgrade of just a notch would make it significantly harder for him to raise what he needs).

But here's one reason why it's so difficult to judge where the cost of borrowing for the government will settle in the coming few months: the Debt Management Office has sold fewer gilts than have been bought by a separate part of the public sector, the Bank of England.

In April, May and June, Stheeman and his team flogged £57.9bn of gilts, while Mervyn King's traders waded into the market to buy £77.7bn of UK government debt.

The Bank of England is buying as part of its so-called Quantitative Easing programme to increase the stock of money in the economy and cut the cost of credit.

But the Bank has almost disbursed the £125bn allocated in total to the scheme - and was somewhat equivocal a couple of weeks ago about whether it will increase its gilt-purchasing budget.

Investors seem persuaded that the Bank of England will buy a bit more - although we'll have the first test in a gilt auction this morning of whether the Bank's equivocation is seriously unsettling investors.

Gilt prices have been falling after the large penny dropped in markets that there may not be many more weeks before the Bank of England transmogrifies from a massive net buyer of gilts into a potential seller.

Where the gilt price settles then is - as Stheeman implied - not something that can be predicted with scientific certainty.

And here's the great and painful paradox.

If the Bank of England stops buying at a moment when investors become a bit more confident about prospects for the global economy and our economy - if they become less averse to risk and more interested in buying assets other than AAA sovereign debt - well, then it might become altogether more tricky and expensive for the government to borrow.

The Conservatives' "plan for sound banking" (as they call their 52-page policy document on reforming the banking system to be published tomorrow) differs from government policy in a number of significant ways.

First, as I've written about here several times in the past few weeks, a Tory administration would transfer to the Bank of England responsibility for preventing banks, building societies and insurers - both individually and collectively - taking excessive financial risks.

A Financial Regulation Division, responsible for regulating and supervising financial institutions, would be created at the Bank and it would be headed by a new Deputy Governor for Financial Regulation.

This would strip from the City watchdog, the Financial Services Authority, one of its most important functions.

What would be left at the FSA would be its consumer protection functions and its responsibilities to ensure that financial institutions conduct business in a fair and proper way.

So the Tories would rename the FSA as the Consumer Protection Agency. And this Consumer Protection Agency would absorb the bit of the Office of Fair Trading which licences and regulates consumer credit businesses.

These institutional changes would be part of a broader initiative to shift the balance of power from banks and insurers to consumers.

Thus the Tories would force credit card companies and banks to provide much more information directly to individual customers about their charges and terms for a variety of products, from credit cards, to mortgages and savings accounts.

George Osborne, the shadow chancellor, wants financial institutions to provide this information in a form that could instantly be uploaded into product comparison websites, so that customers can instantly see whether they are getting a good or bad deal from their banks.

This is an idea that Osborne has borrowed from one of the pioneers of "nudge" behavioural economics, Richard Thaler.

Also, in the event that the Tories win the next election, Osborne would - in the words of the policy paper - "ask the Office of Fair Trading and the Competition Commission to conduct a focussed examination of the effects of consolidation in the retail banking sector".

Such a competition probe would look at whether choice for consumers has been reduced in a seriously damaging way by the takeover of HBOS by Lloyds, the acquisition of Bradford & Bingley and Alliance & Leicester by Santander, and the withdrawal from the UK of various overseas institutions and assorted small firms.

Since the opinion polls indicate that the Tories will form the next government, the likes of Lloyds and Royal Bank of Scotland will doubtless instruct lawyers almost immediately to prepare their cases for why they haven't acquired excessive market power in recent months.

But these banks will note with some trepidation that Osborne says the findings of these competition enquiries would determine his strategy for how to dispose of the government's huge stakes in Royal Bank and Lloyds, together with its 100% ownership of Northern Rock. There is an implication that he would break up RBS and Lloyds.

Finally, the Tories are going further than the government in detailing plans for what's called macro-prudential regulation, which is about preventing the kind of lending binge that took place in the three or four years before the summer of 2007 and precipitated the worst financial and economic crisis since the 1930s.

The paper talks about creating "a powerful new Financial Policy Committee within the Bank of England, working alongside the Monetary Policy Committee", which would have tools - such as varying how much banks can lend relative to their capital resources - to prevent banks in general from taking excessive risks.

However, although the Bank of England's size and power would be massively increased by these reforms, the Tories want to curb the personal power of the governor of the Bank of England.

The paper says that "the new structure will... reduce the institutional reliance on the position of Governor, with a collegiate approach to policy on financial stability and more use of external expertise".

The point is that responsibility for the supervision of banks and for limiting risks in the financial system would be vested in the Financial Policy Committee, which - like the existing Monetary Policy Committee that sets interest rates - would consist of Bank executives and outsiders.

So the governor of the Bank of England could not dictate - as he can in limited areas now - how the bank will act in a financial crisis.

What is striking is that - as a backstop to the powers of this Financial Policy Committee - the Tories are committing themselves to the introduction of a leverage limit, or a simple ceiling on how much banks can lend in gross terms relative to their capital resources.

But what's also conspicuous is something of an internal contradiction in the Tory analysis.

On the one hand, the paper is quite clear that certain banks - such as Royal Bank of Scotland - borrowed and lent far too much and became far too complex for the health of the economy. They became instruments of economic destruction.

The paper says that there are "some valid arguments in favour of some degree of structural separation between the riskiest banking activities and deposit taking institutions [or those that look after individuals' savings]".

However, it thinks that unilateral action to force banks to divest these allegedly dangerous activities would not be feasible and would damage the City as a financial centre.

So it's relying on the Bank of England to impose the equivalent of a punitive tax on banks that do considerable trading for their own account of a speculative kind, by forcing such banks to hold disproportionately higher capital reserves.

And the Tories believe that this approach "would result in the separation of the riskiest activities if retail banks find the additional costs of engaging in them are too great".

Well, we'll see. Interestingly, this is one of the few policy areas where the government and the Tories are somewhat converged in their thinking.

There's a good deal more to say about the Tory paper. And given that most in the City assume - rightly or wrongly - that the Tories will be in power next year, it'll be studied in minute detail in the City over the coming weeks.

But I'll restrict myself to four additional observations.

• First, the Tory paper is vague about what will happen to the FSA's oversight of regulated markets and listed companies, and its powers to investigate market abuse and insider trading. That said, I understand that these roles might well end up - in the longer term - merged with the Takeover Panel and the Financial Reporting Council as part of a new regulatory authority with broader responsibilities for the conduct of publicly listed companies.

• Second, the management of the FSA now faces a nightmare few months: given the high probability that its days are numbered, retaining and recruiting staff will not be easy.

• Third, breaking up the FSA in the way that the Tories want to do is easier said than done, in that recent reforms have created unified teams within the FSA of those who look after consumer protection and those who are responsible for checking that banks and insurers are lending and investing prudently.

• Fourth, on the assumption that Lord Turner, the chairman of the FSA, becomes the new Deputy Governor for Financial Regulation (my sense is that the chief executive of the FSA, Hector Sants, is planning to leave the regulatory fray next year), we'd see the start of a compelling contest to succeed Mervyn King as Governor in 2013.

This would probably be a death match between Turner and the Deputy Governor for Financial Stability, Paul Tucker.

Which, of course, is to trivialise these weighty policy matters, but is not uninteresting.

A Treasury-sponsored review has today recommended substantial reforms to the structure and behaviour of banks' and financial institutions' boards, to restrict the freedom and the incentives for senior executives to take reckless risks.

Sir David Walker, who is a senior adviser to the US investment bank Morgan Stanley and is a former director of the Bank of England, believes that last year's financial crisis, which helped to precipitate the worst global recession since the 1930s, was in part a consequence of "failures in governance in banks and other financial institutions".

After five months of analysis, he has concluded that:

1) the boards of big banks didn't understand the scale of the risks their organisations were running;

2) that non-executives of big banks did too little to rein in the excesses of the executive directors;

3) that shareholders in banks also failed to curb reckless gambling by financial institutions, that the owners didn't "exercise proper stewardship",

4) and that bankers were paid in a dangerous way which encouraged them to speculate imprudently.

One recommendation which is likely to alarm some banks is that boards' remuneration committees would set the pay not only of executive directors but also of executives below board level whose "total remuneration" might be "expected to exceed the median compensation of executive board members".

In the case of Barclays, for example, which owns a substantial investment bank, this would lead to the board setting the pay of hundreds of bankers paid as many millions each year as those on the Barclays board.

The pay of these so-called "high end" executives would also be disclosed "in bands" in the banks' annual reports, although the executives' names would not be published.

In an interview with me, Sir David said he was fully prepared for protests about this degree of disclosure on pay from the big banks, who are likely to complain that they would be revealing commercially sensitive information that could put them at a competitive disadvantage.

Like the Financial Services Authority, the City watchdog, Sir David also wants a significant element of bonuses or performance pay to be handed to relevant bankers only after several years have elapsed, up to five years, or enough time to verify that the deals triggering the bonuses aren't toxic.

Other proposals are:

a) new risk committees should be set up on boards, separate from the audit committees, which would be chaired by a non-executive;

b) these risk committee would overseas all substantial transactions and would have the power to block those deemed too dangerous;

c) non-executives would devote 30 to 36 days each year to the affairs of a bank or financial institution, up from 20 to 25 days at present (many of you probably won't believe they earn their fees of £100,000 or so a year for four to five weeks of work);

d) non-executives would be better trained, they would be scrutinised more rigorously by the FSA and they would be encouraged to hold the executives to account, in a way that would probably end the "collegial" nature of bank boards;

e) the chairmen of banks or other financial institutions would commit no less than two-thirds of their time to the business, they would have significant and relevant "financial industry experience", and they would face re-election by shareholders every year;

f) boards would monitor more closely whether their big shareholders were selling shares and would take steps to learn why these shareholders had lost confidence in their businesses;

g) the FSA would also "be ready to contact major selling shareholders to understand their movitation";

h) institutional shareholders would sign up for a new set of "principles of best practice in stewardship", to encourage them to be more actively engaged in the affairs of companies, which would be overseen by the Financial Reporting Council.

When I spoke to Sir David he stressed that he was acting in an independent capacity when making these recommendations and that the Treasury was under no obligation to implement them.

Some may feel that his reforms would lack teeth, because he does not want them enshrined in legislation. Instead he wants them enforced through the Combined Code, the voluntary code on UK boardroom practices that is overseen by the Financial Reporting Council.

Update, 15:35: Barclays has told me two interesting things: first that its board already approves all group pay packages worth more than £500,000 a year; second that it's not worried about disclosing in its annual report how many of its staff receive pay of that magnitude and greater.

Sir David Walker, a former regulator and someone who in the past would have been described as a City grandee, will today publish his prescription for how the so-called governance of banks can be improved.

Governance is the sententious word for the structures and rules for institutions that are supposed to prevent them doing the wrong thing.

And since banks all over the world in the years before the credit crunch did the wrong thing on a scale that was without any precedent, there must surely have been a collective failure of governance.

Or to put it another way, the owners and non-executive directors of banks from UBS in Switzerland to Merrill Lynch and Lehman in the US to Royal Bank of Scotland and HBOS in the UK were useless at preventing those banks borrowing and lending in a reckless and dangerous manner.

So presumably that means the owners and non-executives were either under-qualified nitwits, cowards or unhealthily close to greedy manic chief executives.

Which would mean - surely - that new codes of conduct for shareholders and initiatives to improve the quality of those who sit on bank boards would make a world of difference.

That may be so.

But it's worth reminding ourselves what actually happened at Royal Bank of Scotland, for example, before we conclude that accidents can always be prevented if only the right people are in the right jobs.

Because the non-executives at Royal Bank were an impressive bunch - on paper.

They included three former bankers, an erstwhile treasury official whose responsibilities included financial regulation, the one time boss of an insurance giant, and the titular head of Goldman Sachs in Europe.

At the time, these were not individuals who would have been described as either ignorant of finance, shrinking violets or nincompoops.

Some have alleged that the non executives were terrified of the steely chief executive of the time, Sir Fred Goodwin.

This, I have to say, is less than compelling. I know some of these non-execs. And I can tell you that they are materially tougher and less pliable than old boots.

Then there's the question of whether they knew as much as they should have done about what was going on.

That is is a moot point. Some non-execs say they didn't know all the relevant details about RBS's holdings of lower quality housing loans in the US.

But it wasn't non-executive ignorance of those risks - or perhaps of any risk - that led to its collapse into the arms of the state.

What really did for RBS was its record-breaking takeover of the rump of the Dutch bank ABN Amro in the autumn of 2007.

This was the wrong deal, at the wrong price and at the wrong time.

The non-executives were not ignorant of the risks RBS was running in buying ABN. However they thought these risks, which turned out to be suicidal, were worth taking.

All of which simply says that even smart, well-qualified people can be gripped by irrational exuberance - and that therefore we shouldn't get carried away with the idea that governance can be a perfect protection again catastrophe.