8 Responses to “The big lessons from Northern Rock”

Comments

1. Tim Congdon: Although – as ever – Mr Wolf was interesting and thought-provoking in ‘Big lessons from Northern Rock’, the message of his article was wrong. The ‘big lesson’ of the Northern Rock affair is not that most of the blame falls on Northern Rock’s management, but that the UK’s system of banking supervision is unsatisfactory.

   The main difficulty with Mr Wolf’s argument is simple, that
   1. Northern Rock was deemed to be a solvent and satisfactorily regulated bank by the Financial Services Authority, and
   2. Northern Rock was found by its auditors to have been profitable and to have increased its capital in the first half of 2007.
   If a solvent, profitable and well-regulated bank were not entitled to a lender-of-last-resort loan from the Bank of England in a liquidity crisis, then the Bank of England would have ceased to be a central bank in the usual sense of the term. To make this statement is not to say that Northern Rock’s management and the FSA are beyond reproach.

   Further, if the FSA and Northern Rock’s auditors were correct that the bank was solvent in early 2007, there is
   1. little risk that a large amount of public money will be lost, and
   2. no justification for the interference with private property rights that compulsory nationalisation would involve.
   Again, this statement does not mean that the FSA and the auditors were correct in the assessment they made in early 2007, and it does not rule out the possibility that Northern Rock’s interim report, and its Annual Report and Accounts, were misleading.

   Of course if the FSA and the auditors ‘fell down on the job’ (i.e., the bank’s various financial statements were misleading), much larger questions are raised.

   Two features of the UK’s system of banking supervision were largely responsible for causing the Northern Rock crisis. Both date from the 1998 Bank of England Act, although the subprime crisis was the trigger for the timing of the problems. The two features were
   1. the absence of a rule specifying a minimum liquid assets ratio (such as had existed for decades until 1998) on UK ‘banks’, where the word ‘banks’ is understood to include all institutions that might reasonably expect lender-of-last-resort assistance from the Bank of England in a crisis, and
   2. the separation of the banking supervision at the FSA from the ability to make a LOLR loan at the Bank of England. (Banking supervision is necessary for judging a bank’s suitability for a LOLR loan.)

   ………………………………………………

   There is no room here to review every paragraph of Mr. Wolf’s piece. But a discussion of the just the first three paragraphs shows that he should have presented his argument with much more care.

   Paragraph 1. According to its interim report (published on 25th July) Northern Rock had equity capital of £2.3b. at 30th June 2007 and had made pre-tax profits in the preceding year of over £600m. Despite the high cost of its loan from the Bank of England, it is inconceivable that it has had operating losses since then that wipe out its entire capital. It ought therefore still to have positive shareholders’ funds and to have some value on the stock market. (An important reservation here is the contingent liability on the Granite financing vehicle. This liability may wipe out shareholders’ funds. See the comment below on SPVs.)

   Paragraph 2. Mr. Wolf says that Northern Rock “is the only significant UK bank whose financing imploded during the crisis” overlooks that every UK bank has been hanging on to its cash in recent months. (In other words, all banks have been reluctant to lend to other banks.) It also overlooks that the share prices of all UK banks have tumbled in the last nine months. Alliance & Leicester and Bradford & Bingley had similar business models to Northern Rock, with their assets focussed on residential mortgages and their liabilities including a high proportion of wholesale money. They also have seen large share price falls and it is conceivable that they also will have to seek LOLR help from the Bank of England.

   Paragraph 3. Mr. Wolf suggests that the FSA must take some blame, but wonders what it was doing wrong. As noted above, since the 1998 Bank of England Act the UK’s banks have not been required to maintain a buffer of liquid assets, which could be sold to the Bank of England when they are short of cash. The FSA’s executives ought to have realised years ago that this was a serious omission from the Act. (It also appears that the FSA gave its blessing to the growth of “special purpose vehicles”, such as Granite in Northern Rock’s case. These SPVs were represented by their sponsoring banks as being independent, in terms of capital and funding requirements, but in fact there was [and still is] a large contingent liability if the SPVs could not fund themselves. The FSA should have made clear – two or three years ago – to the UK’s banks that it could not accept the potential balance-sheet strain that might arise if the contingent liabilities became actual.)

   Posted by: Tim Congdon | November 19th, 2007 at 8:07 pm | Report this comment
   

2. Willem Buiter: Martin’s absolution of the Bank’s liquidity management policies is incorrect on two counts.

   First, if the Bank had, at its discount window (standing lending facility), accepted the kind of collateral routinely accepted by the ECB and the Fed, then the whole Liquidity Support Facility for Northern Rock would have been redundant. The discount window, available at a penalty rate of 100bps over Bank Rate and on demand against suitable collateral, would have provided Northern Rock with the £20bn to £30bn worth of financing it required. And the identity of the borrower at the discount window is confidential! The chances of survival of Northern Rock would have received a further boost if, like the Fed, the Bank of England had extended the duration of its discount window loans from overnight to up to one month.

   The Bank’s ludicrously restrictive eligible collateral set and its insistence on only overnight lending created the need for the purpose-built Liquidity Support Facility. This Liquidity Support Facility did, in the glare of publicity, what a properly functioning discount window would have done discreetly.

   The discount window would have been even more effective if the Bank, jointly with the FSA, had run an effective campaign to de-stigmatise discount window borrowing. They should do so as yet.

   Second, the amount Northern Rock would have had to borrow at the (properly managed) discount window of the Bank could have been much smaller if the Bank, in its sale and repurchase operations (repos), had not insisted on the same ludicrously restrictive list of collateral as it did for borrowing at the discount window. By only accepting as collateral in repos securities that were already liquid (UK and European Economic Area sovereign debt instruments, the debt of a few highly rated international organisations like the World Bank, and, in exceptional circumstances, US Treasury securities), its liquidity-enhancing repos were quite pointless during a liquidity crisis. The ECB accepts as collateral in repos private debt instruments, including illiquid and nonmarketable securities, asset-backed securities, mortgages and mortgage-backed securities, rated at least in the A category.

   The Bank relented in its collateral policy for repos (and in its willingness to try to influence the spreads of 3-month Libor, and of other maturities longer than overnight, over the market’s expectation of the policy rate over the relevant horizon) only after Northern Rock had hit the rocks. And then its change in procedures and policy was half-hearted.

   The Bank confused lender-of-last-resort-operations to assist individual banks, (actions which are inherently subject to moral hazard which must be mitigated by charging a penalty rate for the loans), with market-maker-of-last-resort operations, which support markets for securities that have become illiquid. Provided the collateral that is offered in repos is valued properly (say through auctions organised by the Bank if no other market price is available) and subject to proper liquidity haircuts on this ‘fair’ value, no moral hazard or other incentive distortion is involved with market-maker-of-last-resort operations There is no need to charge a penalty rate if the collateral is valued appropriately and subjected to an appropriate short back and sides. Yet when the Bank offered to organise three auctions for 3-month repos against prime mortgages, it set a penalty floor for these loans (100bps above Bank Rate). Not surprisingly, nobody came to the auctions.

   Correcting market failure (illiquidity is a manifestation of loss of confidence and trust and constitutes market failure) is why we have public institutions like the Bank and the FSA. The Bank is the only institution that can create any amount of liquid assets for the private sector, at any time and at no cost. The private provision of liquidity in times of crisis is not an option. Requiring private institutions to hold intrinsically liquid securities on their balance sheets in sufficient amounts to allow them to live through periods of market turbulence, disruption and illiquidity is possible but socially ineffecient. Liquidity is a public good. The private provision of public goods is as silly and inefficient as the public provision of private goods.

   So the Bank failed in its discount window operations and it failed in its liquidity-oriented repos. We can never be absolutely certain that Northern Rock would have survived if the Bank had had the discount window and repo policies of the ECB or the Fed, but it would seem to be pretty darn likely.

   Martin quotes with approval the statement by Mervyn King that “The role of the Bank of England is not to do what banks ask us to do; it’s to do what’s in the interests of the country as a whole.”

   My response to the application of this truism to the Northern Rock situation is not Martin’s ‘Amen!’, but ‘Never mind, stick to the issue’. The Governor, in his eagerness not to please the banks when pleasing the banks would not have been in the public interest, ended up acting as though he believed that just because banks asked him to do something, that meant the requested action could not possibly be in the public interest. However, even self-serving arguments made by deeply biased self-interested parties can be correct. In this case they were.

   It’s not enough to be tough. You also have to be right.

   Posted by: Willem H. Buiter | November 19th, 2007 at 11:28 pm | Report this comment
   

3. Patrick Honohan (guest): Although a minimum liquid assets ratio, suggested by Tim Congdon in response to Martin Wolf’s piece, might sound attractive, it is not as easy as that to prevent casualties occurring when short-term markets seize up (in as dramatic a fashion as occurred this year). At what percentage should such a ratio be pitched (and expressed as a ratio of what)? Think of a number and ask yourself whether it would have been enough to provide Northern Rock with all the liquidity it needed to replace the borrowings it could not roll-over. And what about the offsetting transactions, liquidity puts and the like, that a determined bank could easily adopt to undo the effects of the regulation? There has clearly been reckless disregard for liquidity risks–and not only by Northern Rock. But simplistic rules offer no secure solution.

   Posted by: Patrick Honohan | November 20th, 2007 at 10:34 am | Report this comment
   

4. I apologise for the delay in responding to the posts by Tim Congdon and Willem Buiter. I have the greatest respect for both of them and recognise that they have both devoted more attention to these issues over their careers than I have. For this reason, I wished to think about the points they raise carefully.

   I have tried to do so, though I recognise that I have not (yet) reached conclusions with which I am entirely satisfied. But I do know why this is so. It is, I admit, partly emotional: I find it hard to accept that the public sector should be forced to rescue the financial system from the consequences of its own greed and stupidity, again and again. It is also intellectual: defining the proper limits of such assistance is hard to do. Reasonable people are bound to differ on what these limits should be.

   The financial system makes much of its money out of financing investments in risky assets with short-term borrowings that it promises to redeem at par. Just putting this down should make any reasonable person shudder, since this is obviously an accident waiting to happen. And it does, repeatedly. But it is also a profitable accident waiting to happen, since the public sector is willing to help the financial sector redeem its borrowings (through the so-called “lender-of-last-resort” function) and, in extremis, to guarantee the value of a large part of its outstanding loans (i.e. deposits).

   Evidently, given its explicit and implicit guarantees to this rickety structure, the public sector has to regulate the capital and risk-taking of these institutions and, in the past at least, also the liquidity of the assets it holds. The financial sector, in turn, tries to get round these restrictions by placing some part of its business outside the regulatory net. The more complex the institutions, the easier the latter becomes.

   The reason I have spelled out these largely obvious points is that it only then that one can understand how difficult the position in which the authorities constantly find themselves turns out to be. They wish to prevent crises now, without increasing the likelihood of crises in future. It is not easy to make the judgements required to do this. But one point seems to me clear: the absence of insolvencies does not prove that the authorities are doing their job well. It could just as well prove the opposite. So the fact that Northern Rock ended up in huge difficulties does not prove the authorities got it wrong. It could just as well mean they got it exactly right.

   Now let me turn to the points raised by Tim Congdon.

   Tim argues that Northern Rock was entitled to lender-of-last-resort assistance from the Bank of England. I agree. He also insists that it is solvent, taking as evidence the auditors’ report and the view of the FSA. From this he concludes that there is little chance that any public money will be lost and so no reason to nationalise the bank. On this I disagree.

   The truth is that we have no idea whether Northern Rock is solvent. There are two ways of thinking about the solvency of Northern Rock.

   One – the simplest one – is whether Northern Rock could meet its obligations as they came due. We know the answer to this quite precisely: it could not do so. That is why it needed to be rescued. If it were not a financial firm that would have been enough to ensure that it was put into administration. Of course, shareholders might have ended up with something at the end of this process. What is clear is that we have the lender-of-last resort to prevent this from happening to financial institutions, with all the dangers that implies for the latter’s mismanagement of their liquidity.

   The second definition, necessarily forward-looking, is whether the present value of the stream of its receipts (cash inflow) exceeds the present value of the stream of its outgoings (cash outflow). The truth is that neither the auditors nor the FSA knows the answer to this question, since they cannot now know the future receipts and outgoings of the institution.

   Accounts are economic history. They tell one little about the relevant future. In this case, what matters is the future of the economy, of house prices and of employment. It seems to me to be perfectly possible that Northern Rock is insolvent. The evident unwillingness to refinance its borrowings and the recent offers made to purchase Northern Rock both suggest that informed people are not at all convinced the institution is solvent. I find these doubts understandable.

   At present, Northern Rock continues in business solely because of public sector interventions – the guarantee of its deposits and the loan from the Bank of England. In these circumstances, leaving decisions over the sale of the bank to shareholders amounts to giving the latter the upside and the public at large the downside.

   Since I am unpersuaded that there is no downside, this does not seem to me to be an acceptable state of affairs. The idea that such public sector guarantees will be transferred to new owners, at no charge, seems to me to be utterly unacceptable. They are, after all, immensely valuable. Why should the shareholders benefit from such free insurance from the public? At the least, if private shareholders are to take over the bank, the open-ended deposit guarantee must be withdrawn. My guess is that this would restart the run at once and so force the bank into administration. Maybe, at that point, we can agree that it should have been nationalised, instead.

   I agree with Tim on the need for some sort of liquidity regulation of banks (though, as Mr Honohan notes, this is hard to do). But it is one of the lessons of this crisis. (I will discuss this further when I turn to Willem Buiter’s comments below.)

   I am not persuaded, however, that the separation of the regulation from the lender-of-last-resort function was as grievous an error as Tim argues. As Charles Calomiris of Columbia University has noted in this forum, such separation is perfectly normal. The advantage of it is that the central bank will not be captured by its own history as a regulator. Any regulator is bound to believe that an institution it regulates deserves such lending. The Bank of England can take a more dispassionate position. In this case, it did provide the loan, anyway. So the idea that the failure of co-ordination was decisive seems to be at the least exaggerated. A more interesting question is whether the FSA did a decent job of regulating Northern Rock. I sympathise with Tim’s doubts on this.

   Now let me turn to Tim’s specific comments:

   On Paragraph 1, let me just repeat what I said above on Northern Rock’s solvency. On Paragraph 2, it is possible that other banks will need assistance from the Bank of England, but it has not yet happened. On Paragraph 3, I agree entirely with Tim’s insistence on the mistake made in approving so-called “special purpose vehicles”. Indeed, it would be fair to argue that we really do not know what regulation of banks is aimed at achieving. This is what I suggested in my original column.

   I plan to respond to Willem’s lengthy post in the very near future.

   Posted by: Martin Wolf | November 24th, 2007 at 4:01 pm | Report this comment
   

5. Martin Wolf: Let me now comment on Willem’s response to my column. Of course, Willem has written a great deal on the liquidity crisis and Northern Rock, in his blog Maverecon and the Economists’ Forum. But here I will focus on the specific comments above.

   Willem’s first point is that if the Bank had taken the sort of collateral accepted by the Federal Reserve and the European Central Bank, the operation to support Northern Rock would have remained confidential.

   I am quite sure that this is wrong if, as Mervyn King has argued, Northern Rock needed £20bn-£30bn. Maybe thirty years ago it would have been possible to keep such an enormous operation secret for months. But one thing I do know more about than Willem is how the press works (remember that we are talking about the global press here, not just the British press).

   Rumours would have swirled around about who was taking all the money. Reporters would have called every eligible bank. Naturally, all those not taking the billions of pounds would have promptly denied that the money was going to them and, hey presto, Northern Rock would have been “outed” in a few days, at most. Indeed, the Bank was unable to hide even £300m advanced to Barclays.

   The only way to conceal the identity of the stricken borrower would have been for the Bank to have advanced so much money that the £20bn was a relatively small proportion of the total. But if the Bank had advanced, say, £100bn through its discount operations (and so vastly more than the ECB has advanced), outsiders would have concluded that the entire British banking system was in a desperate state. Given the limited deposit insurance available, there would almost certainly then have been a run on the banking system as a whole. That would have been a calamity, indeed.

   In short, I am convinced that Willem is wrong on this point. There was no possible way of concealing an operation of this scale in favour of Northern Rock.

   Willem’s second point is that Northern Rock would have had to borrow far less if the Bank of England, in its ordinary sale and repurchase operations, had not insisted on the same restrictive list of collateral as it did for borrowing at the discount window. Similarly, he argues, the Bank failed in refusing to influence 3-month Libor and other maturities that are longer than overnight.

   I understand the argument that the Bank should be prepared to take a wider range of collateral. It makes some sense. But the argument made here seems to me to be much less compelling than critics believe.

   First, the eligible debt instruments (sovereign debt of the European Economic Area, of international organisations and US treasuries) are hardly scarce! On the contrary, my back-of-the-envelope calculations suggest that outstanding value of these instruments must be close to £15 trillion!

   Given that the Bank’s list of eligible securities was known in advance and given the risks inherent in funding relatively long-term assets (such as mortgages) in the commercial paper or inter-bank markets, it is hardly unreasonable to expect banking institutions to hold substantial quantities of these assets to meet short-term funding requirements in the event of a liquidity crisis of precisely the kind we have seen (or arrange some other form of insurance). That is, so it seems to me, a simple matter of managerial competence.

   Second, it is untrue that the UK money markets have shown signs of stress fundamentally different from those in the eurozone. This is not surprising, since, as I understand it, the total liquidity provision by the ECB has been modest, and the latter’s operations at three-months have been substitutes for operations at shorter maturities, not additions to them.

   Here then is some evidence on spreads in relevant markets.

   A. The average deviations of overnight Libor from the relevant policy rate since the first of August 2007 have been 5 basis points for the euro, 15 basis points for sterling and 16 basis points for the dollar. The main reason the deviation in euro rates has been so much smaller is that, quite frequently, the deviation in euro rates has been negative. It is not clear to me why that should ever be a sensible policy objective.

   B. The average deviations of three-month Libor from the policy rate (over the same period) have been 38 basis points for the dollar, 64 basis points for the euro and 67 basis points for sterling. The sterling deviation was relatively large in August and early September and peaked at 115 basis points on September 12th 2007. Is that deviation really such a catastrophe? Moreover, from September 20th to November 15th the deviation in sterling rates was smaller than those in the euro.

   C. As the latest Inflation Report shows, in Chart 1.2, the relationship between three-month Libor and future expected policy rates has been very similar for the eurozone, US and UK. Indeed, for July 2008, the UK’s deviation is the smallest (at close to zero).

   My conclusion on this point, then, is that it does make sense for the Bank to accept a wider range of collateral, though changing the set in the midst of a crisis would have created a serious credibility issue. But I am entirely unpersuaded that the Bank’s error (if error it was) was a material factor either in the functioning of the UK’s money markets, in general, or to the fate of Northern Rock, in particular. On the contrary, if Northern Rock had been a eurozone bank, the ECB would almost certainly also have found it necessary to advance exceptional amounts of money.

   Willem’s third argument is that the mistakes allegedly made by the Bank show that it fundamentally misunderstands what a central bank is for. As I understand it, Willem’s view, expressed here and elsewhere, is that the authorities should ensure liquidity in markets for all financial assets at all times, as “market-makers-of-last-resort”.

   I strongly disagree. It is not the job of public authorities to make liquid markets in whatever junk imaginative bankers choose to create. Some markets should not be liquid, because the relevant assets are impossible to understand or value and have large credit risks attached to them. This point is particularly relevant to this crisis, because those are precisely the markets that have become illiquid. Markets in bonds, including even quite risky bonds (such as emerging market bonds) have remained quite liquid throughout.

   Willem’s argument is that illiquidity is a market failure, because it is a manifestation of a loss of confidence and trust. Governments exist to rectify market failure and so should intervene by purchasing the relevant assets after a suitable haircut. My view is that this is indeed sometimes the case. But sometimes the loss of trust is rational: indeed, it shows the market is working (at last).

   In the present case, markets are coming to their senses over both risk premia and asset quality. This is not a market failure, but a market correction. Moreover, central banks are not in a better position than the markets to work out what those assets languishing in illiquidity are worth. But if one does not know that, one cannot know what a sensible “haircut” of the kind Willem calls for would be.

   I agree that liquidity has public good aspects. But it also has private good aspects. The liquidity of their assets is, therefore, something investors should take into account in their decisions. Do we, to take an extreme example, want the Bank of England to make a market in housing when the next big downturn comes (possibly quite soon). If not, why not? The illiquidity of the housing market in a downturn is, after all, a truly significant social and economic problem.

   So I have a narrower view of the proper role of the authorities during the market correction (which is, at bottom, a necessary economic adjustment, not an unaccountable panic) that has created illiquidity in some markets (mostly, the ones where illiquidity was called for) and (well-deserved) difficulties for financial institutions that used special purpose vehicles as a way of holding poor-quality assets off balance sheet.

   What are those proper functions?

   First, the authorities should prevent a financial crisis from damaging the wider economy. This can be achieved by lowering interest rates, when called for.

   Second, they must ensure the continued functioning of the payment system.

   Third, they must (in today’s circumstances) guarantee the safety of the deposits that the public has reason to believe will hold their nominal value in all circumstances.

   Finally, they must prevent a collapse of the financial system as a whole.

   The British authorities have - messily, it is true - achieved all these aims, at least so far. So, quite honestly, I do not understand what this great calamity is supposed to be. Northern Rock is about to disappear (or should be about to disappear: but that is a subject for another column). So what? This should make other institutions quite a bit more careful in future.

   I remain very happy Mervyn King has not gone out of his way to please the British banks. I accept that this does not prove he is right. But it does not prove he is wrong either.

   Posted by: Martin Wolf | November 25th, 2007 at 3:44 pm | Report this comment
   

6. Tim Congdon: My reply to Martin is in two parts. The first part is to protest about the way - in my opinion - Martin misuses words and misunderstands the rules of engagement in a discussion of this sort. The second part is to note the larger and more important issues of public policy that really matter. I am afraid these larger issues cannot be covered in detail here.

   Misuse of words and misunderstanding of the rules of engagement:

   The words ‘insolvency’ and ‘illiquidity’ have fairly definite meanings in the banking context. A bank is insolvent if it is has no capital, i.e., the entry for ‘shareholders’ funds’ on the liability side of the balance sheet is zero or negative. (A bank may be solvent, but have inadequate capital relative to a regulatory standard, but we are not at the moment talking about capital adequacy.) A bank is illiquid if it is unable to repay creditors in full and on time. Critically, it is illiquid if it is unable either to repay retail depositors with cash or (in the case of a non-clearer such as Northern Rock) to repay the holders of its bond and note liabilities (as they come due) with credits from the account it holds with a clearer.

   Banks are unusual compared with other organizations, in that most of the items in the balance sheet have a value close to par (i.e., £100, for simplicity). Let us - for the sake of illustration - think about a non-clearing bank with a balance sheet of £100b. and significant retail deposits. In normal circumstances, non-equity liabilities (deposits, bonds, notes, etc.) amount to £95b. and are less than assets of, say, £100b. with the difference being due to the £5b. of equity capital. (Retail deposits might be £30b. and wholesale liabilities £95b.) All being well, a non-clearing bank with this structure would be regarded as liquid if it is had vault cash of, say, £1b. and a positive balance in its account with a clearer of £1b., i.e., its assets consist of £2b. of cash and near-cash, £10b. of fairly liquid assets run by the bank’s Treasury with a view to ensuring that the cash and near-cash is always positive, and £88b. of earning assets.

   Insolvency refers to the inadequacy of the assets relative to the value of non-equity liabilities; illiquidity refers to the inadequacy of cash as a proportion of assets and so is a statement about the composition of the assets side of the balance sheet.

   Insolvency is about the value of assets relative to liabilities; illiquidity is about the composition of assets.

   It is perfectly possible - in the example under discussion here - for the risky assets to be worth £88b. in a normal, business-as-usual, ‘going-concern’ sense and yet that the bank ‘runs out of cash’. What happens is that - as, say, £12b. of wholesale liabilities fall due - the bank sells its £10b. of liquid Treasury investments and deploys its £2b. of cash and near-cash, and then has nothing liquid left. Instead of cash, near-cash and liquid assets being 12% of assets, they are zero. Assets and liabilities have both contracted by £12b., from £100b. to £88b., and the bank is illiquid and cannot pay back any further deposits with cash. However, when the auditors prepare the accounts, the bank still has £5b. of equity capital, because if has no loan losses on its £88b. portfolio of earning assets. (Its non-equity liabilities have fallen from £95b. to £83b. because of the redemption of the £12b. of note liabilities, but the £5b. of equity is untouched.)

   There is no doubt that Northern Rock became illiquid at some point in late July/early August 2007. Martin evidently believes that illiquidity implies insolvency. But it does not. Martin equates illiquidity with insolvency, but that is not - in my experience and, I believe, the experience of most people involved in financial markets - how words are used in this subject.

   If the mortgages were likely to be repaid at £95 in the £100, then Northern Rock’s assets would not cover its liabilities and it would be insolvent. But that is not the situation. There is no suggestion that Northern Rock’s assets - pretty bog-standard UK mortgages - are performing badly. To repeat the point, Northern Rock is illiquid, not insolvent.

   Martin then makes all sorts of statements which are described, politely, as ignoring the usual rules of engagement in a discussion of this kind. He says, “Accounts are economic history. They tell one little about the relevant future. In this case, what matters is the future of the economy, of house prices and of employment.” He should check what Northern Rock’s accounts actually say and, whether he likes it or not, they are relevant to the future. Vitally, the loan-to-asset ratios on Northern Rock’s whole mortgage book are 60% and on new mortgages 80%. In other words, house prices have to fall by over 20% before Northern Rock’s borrowers lose an incentive to repay in order to keep their homes. Even then Northern Rock would not be insolvent, because most people with negative equity would continue to service their debts and in any case Northern Rock has a cushion of equity.

   In any worthwhile discussion of these issues a bank’s latest Report & Accounts (and in this case its Interim Report) are basic. But Martin does not refer to them at all. Has he bothered to read them?

   Martin might defend his statement that Northern Rock is insolvent by claiming that, since it could not find willing lenders in the wholesale banking markets, these lenders regarded the bank as bust and so it must be bust. But - if so - he would be assuming that which has to be proved. Only time will tell whether the auditors’ and FSA’s judgements on the quality of Northern Rock’s assets were wrong. (A fundamental point here is that investors in bank floating-rate-notes, CP, etc. are not the same as investors in bank equities. I suppose Stiglitz would call this an ‘information asymmetry’ or something.)

   The convention in banking is that - if a non-clearing bank cannot fund its assets (which was the problem here) - it goes in the first instance to its clearer (evidently Lloyds TSB in this case) and, if the clearer cannot help, the clearer goes to the central bank. Both Northern Rock and Lloyds TSB were given a clean bill of health by the FSA in early 2007, and both organizations were and are well-known to both the FSA and the Bank of England.

   The arrangements whereby banks go to the central bank if they have a cash problem are fairly straightforward, and should not involve a lot of drama and passion. It is interesting that the solvency issues at present appear to be worse in the USA and the Eurozone than in the UK, but American and European banks are not complaining about the Federal Reserve and the ECB. As I said at the start, if a solvent, well-regulated UK bank cannot borrow from the Bank of England when it has a liquidity problem, the Bank of England has ceased to be a central bank in the traditional sense. I am afraid the blame for this affair lies in two places,

   1. in a tactical sense, very heavily at the Bank of England and (sadly, but the verdict cannot be escaped) with Mervyn King who should have been far more pragmatic and flexible when the issue blew up, and

   2.in a strategic sense, with the 1998 Bank of England Act which redistributed responsibilities in a silly way by splitting bank supervision (FSA) from the organisation (B of E) that had the power to make a loan (but this split was very fashionable at the time in the academic literature), scrapped liquidity supervision entirely, and made no statement about the criteria for and modalities of lender-of-last-resort lending.

   I agree that the FSA, along with other international regulators, should have been tougher on special investment vehicles etc., and I agree with King and others that the UK banking system (like other banking systems) had been far too greedy in economising on capital and liquidity in the years leading up to the 2007 crisis. But the right way to deal with this crisis - as in any crisis of this sort - is to deal with the tactical problem immediately and to resolve the strategic issues later. (When a doctor is confronted by a 50-year old squash player who has had a heart attack, he should deal with the heart attack first. The advice ‘stop playing squash’ can come later.)

   Martin Wolf - well-known as a supporter of the free market - apparently favours the nationalisation of solvent banks that have funding problems. No, that is not the way market economies have developed and it is a very bad idea. Market economies try to preserve incentive structures and so have central banks to help solvent commercial banks that have funding problems. But central banks do need to behave as if there were concerned with central banking, with the emphasis here on the word “banking”.

   Wider issues of public policy raised by the Northern Rock affair

   The exchange here - to which Professors Buiter and Honohan have also made valuable contributions - has raised three important sets of questions,

   1. Is there any social value of the provision of liquidity services by the central bank, where the phrase ‘liquidity services’ refers partly to the lender-of-last-resort facilities that a central bank provides to the commercial banking system? And how does this value arise? More concisely, should central banks exist? Or should the note issue function be shifted to the finance ministry and the banking operations of the central bank scrapped?

   2. How should the value of these liquidity services be distributed among private sector agents? What must a private sector agent - in practice, a bank - do in order to qualify for lender-of-last-resort facilities, and

   3. How should liquidity services be conducted more specifically?

   My quick answers here are,

   1. Yes, the provision of liquidity services by the central bank is virtually costless to society, but it enables the commercial banks to economise on cash (and perhaps capital). This economisation on cash and capital reduces the cost of the loans, and increases the flexibility of the services, provided by the commercial banks to non-banks,

   2. The liquidity services should be provided only to commercial banks that are prepared to be supervised by the central bank in respect of both solvency and liquidity, as used to be the case in the UK before the era of the Basle rules. As a broad generalisation, only commercial banks heavily involved in retail deposit-taking, money (i.e., cash) transmission and cash clearing are likely to be prepared to tolerate regulation of both solvency and liquidity, and to have a business justification for maintaining a good relationship with the central bank, and

   3. Bagehot was broadly right. Substantial loans (to a fairly high multiple of the offending banks’ capital) should be available at a penal rate, assuming the collateral is good and the bank solvent. I regard a rate as ‘penal’ in this context if it is ¼% above inter-bank. (1% is preposterous. Because banks are geared 20 times, a 1% rate bears comparison to 20% to a non-bank. I have to say the more I think about this problem, the better it would if the rules governing LOLR loans were agreed publicly and perhaps even included in legislation. Central bankers’ ‘constructive ambiguity’ is silly hocus-pocus.)

   I think Martin Wolf’s answer to the first question is ‘the banking functions of the central bank should be scrapped and the note-issue function can be handed over to the Treasury’. There is a long tradition in monetary economics which is very much on his side (Ricardo, the Currency School, the 100% reserve banking advocates, etc.). Mervyn King also, I would guess, belongs to the Currency School - and has it showed! I have always sided with the Banking School in these debates. I think Willem Buiter belongs to the Banking School, too, but may have got him wrong.

   The subjects are vast and unresolved. Monetary economics needs its Ronald Coase.

   Posted by: Tim Congdon | November 29th, 2007 at 4:46 pm | Report this comment
   

7. Roger Sandilands (guest): What puzzles and perplexes me is that the government seems willing to let Richard Branson take over the Rock with the aid of £11 billions borrowed from RBS and Citigroup – the wholesale money market – when opprobrium has been heaped on NR for using just that to finance so much of its business.

   Willem Buiter says that NR was reckless in its financing strategy but implies that its strategy would not have been reckless if the authorities had had a more sensible view as to their proper role in dealing with a liquidity crisis. As you (and Buiter to a lesser degree) rightly emphasise, NR’s problem has been a liquidity crisis not a solvency crisis, and I fully support your criticism of Martin Wolf on that point. (Alastair Darling declared that the Rock was solvent when he authorised the loan facility, and this had a positive effect on NR’s share price at the time.)

   I have read that Northern Rock’s mortgage book has had a significantly lower delinquency or late-payment record than the industry average. It is paying a penalty rate on the LLR facilities it is getting from the Bank [and I take your point about just how grotesque a penalty rate that is against the Rock’s capital]. There should be no risk to the taxpayer. If there is such a risk it will be the risk that the government’s macroeconomic policies will have been rubbish. And that won’t have been NR’s fault. Meanwhile, there is no taxpayer subsidy. On the contrary, the Bank is making a tidy profit from the current arrangement – not that its own profit should be an overriding central bank objective.

   If, when the central bank provides liquidity to a particular institution experiencing a haemorrhaging of deposits, those deposits are flowing to other institutions then the Bank can always offset any inappropriate increase in the overall money supply through offsetting sales of securities to the market (and reverse things as and when the central bank loans are repaid).

   I agree with you (and Willem Buiter) that “the provision of liquidity services by the central bank is virtually costless to society”, but I do not think the main benefit is that “it enables the commercial banks to economise on cash (and perhaps capital)” thereby reducing the cost of loans. I think the lower the cash reserve ratio the more difficult it is to macro-manage and control the money supply because of the greatly increased money multiplier from small variations in the cash reserve ratio.

   More important is that borrowing short and lending long encourages saving and enhances the efficient allocation of the nation’s savings. To this end it is the job of the monetary authorities to ensure that the financial system is protected from excessive swings in the provision of overall liquidity (and, indeed, its sectoral distribution), hence in the circular flow of income and spending. I am not convinced by Wolf’s and Buiter’s hard-nosed view that NR is not too big to fail and to hell with both its depositors and shareholders.

   I think it would be grand larceny if the Bank of England nationalises Northern Rock and acquires nearly £100bn of good assets while completely wiping out the shareholders (i.e, Soviet-style nationalisation without compensation), and I also think it would be grand larceny if Branson gets his way.

   PS: In your balance sheet illustration on page 1 of your note, I think you should have written that if assets are £100b (including £5b capital) and retail deposits are £30b, then wholesale liabilities will be £65b, not £95b.

   Professor Sandilands is at Strathclyde University.

   Posted by: Roger Sandilands | November 30th, 2007 at 3:37 pm | Report this comment
   

8. Geoffrey Wood: Much of the above discussion puzzles me. It lacks context.

   In 1802, in his “Paper Credit”, Henry Thornton wrote as follows: “If any bank fails, a general run upon the neighbouring banks is apt to take place, which if not checked in the beginning by a pouring into the circulation of a very large quantity of gold, leads to very extensive mischief.” (Thornton, 1802, p182).

   And who was to “pour in” this gold? The Bank of England:
   “…if the Bank of England, in future seasons of alarm, should be disposed to extend its discounts in a greater degree than heretofore, then the threatened calamity may be averted.” (Thornton, 1802, p188).

   This approach was not incompatible with allowing some individual institutions to fail: “It is by no means intended to imply that it would become the Bank of England to relieve every distress which the rashness of country banks may bring upon them: the Bank by doing this, might encourage their improvidence … The relief should neither be so prompt and liberal as to exempt those who misconduct their business from all the natural consequences of their fault, nor so scanty and slow as deeply to involve the general interests” (Thornton, 1802, p188). Concern should be with the system as a whole.

   Following those procedures - letting the initial unsound institution fail, but providing liquidity so that there was not a subsequent run - worked in the UK and many other places for about 200 years in preserving financial system stability. It was the procedure which the Bank of England seems to have been seeking to follow. This was complicated by the desire to ensure that no depositor lost money, but aside from that what has changed so that we now have to have central banks protecting all banks and preserving liquidity in all financial (and maybe other) markets at all times?

   Geoffrey Wood is professor of economics at London’s City University

   Posted by: Geoffrey Wood | December 2nd, 2007 at 11:56 am | Report this comment
   

The FT Economists' Forum is a discussion among some of the world's top economists. As a general rule we accept comments from invited members only, but submissions from others will also be considered.

If you are a non-member submitting a comment, please include your relevant academic or financial background.

Post a comment

Name (required)

Mail (will not be published) (required)

As a final step before posting the comment, please type the two words you see in the image beloweight numbers in the audio clip; this test is to prevent automated robots from posting comments.

Type the two words: Type the eight numbers: Unfortunately you provided the wrong reCaptcha answer. Please try again: