About two weeks ago, the European Banking Authority announced their standard definitions of impaired loans (i.e. non-performing loans) and other asset quality standards, which aim at harmonising the various definitions in place throughout Europe for their upcoming asset quality review. Today, I won’t even be mentioning the odd fact to see a regulator getting in a bank for a few days and basically telling the bank that it knows its loan book better than the bankers themselves. No, today I will only speak about the harmonisation issue.
Banks have different ways of classifying past-due loans, impaired loans, loans in forbearance and so on, not only in between countries but also within countries. For instance, in the UK, I know some banks that will classify all loans in forbearance as impaired, artificially pushing up their headline bad loans ratio, while others do not, looking better as a result to the untrained eye. Most British banks will automatically classify loans 90 days and more in arrears as impaired, whereas most French banks will only apply the impaired definition when loans are 180 days and more in arrears.
So a standardisation seems to be a good thing as data becomes comparable. Well, it is, and it isn’t. To be fair, standardisation within a country is probably a good thing, although shareholders, investors and auditors – rather than regulators – should force management to report financial data the way they deem necessary. However, it makes a lot less sense on an international basis. Why? Countries have different cultural backgrounds and legal frameworks, meaning that certain financial ratios should not be interpreted the same way from one country to another.
Let’s take a few examples. In the US, people are much more likely than Europeans on average to walk away from their home if they can’t pay off their mortgage. Most Europeans, on the other hand, will consider mortgage repayment as priority number 1. As a result, impaired mortgage ratios could well end-up higher in the US. But US banks know that and adapt their loan loss reserves in consequence. Within Europe, legal frameworks and judiciary efficiency are also key: UK banks often set aside fewer funds against mortgage losses as the legal system allows them to foreclose and sell homes relatively quickly and with minimal losses. In France on the other hand, the process is much longer with many regulatory and legal hurdles. Consequently, UK-based mortgage banks seem to have lower loan loss reserves compared with some of their continental Europe peers. Does it mean they are riskier? Not really.
Another (abstract) example: in country A, the local culture pushes people to pay off their debt at all costs, whereas in country B, most people, once they stop paying back when they run into short-term trouble, never resume payment. In country A, banks consider it safe to classify a loan as impaired only after 180 days without payment. In country B though, banks know that the loan will never be paid off as soon as it is 30 days past due. Standardisation would make both countries use the same classification. Why not, but it doesn’t bring much: analysts will still have to take into account local variations (just in a different way). However, it might spark an unnecessary panic in country A when figures suddenly look much worse to the untrained public.
This is the issue with harmonising. Some standardisation may be welcome, but most analysts and investors already know the differences in reporting between countries and don’t take headline figures at par value, making the EBA exercise relatively pointless. For the less-well informed individuals, the EBA harmonisation could also bring a false sense of safety: figures look comparable, but in reality, they’re not entirely. In the end, harmonised reporting or not, adjustments always have to be made…
The same apply to most internationally-applied regulations. Basel rules, for example, effectively apply standardised capital and liquidity requirements throughout the world (with some local implementation differences). Banks in higher-risk countries have to comply with the same capital ratios as banks in lower-risk countries, the adjustment being made through risk-weighted assets…which are easily gamed. But analysts know well enough that a 17% regulatory Tier 1 ratio (a key bank capital safety ratio) is actually poor for an African bank, despite looking high by Western standards and way above official requirements.
In the end, standardisation makes particular sense in geographical areas where both culture and legal frameworks present only minor differences, such as the US. Europe is still a pretty fragmented continent, and it doesn’t look like things are going to change overnight. The EBA hasn’t been very clear about the exact implementation of its criteria. Let’s just hope they keep the issues discussed in this post in mind.
What Walter Bagehot really said in Lombard Street (and it’s not nice for central bankers and regulators)
(Warning: this is quite a long post as I reproduce some parts of Bagehot’s writings)
Bagehot is probably one of the most misquoted economist/businessmen of all times. Most people seem to think they can just cherry pick some of his claims to justify their own beliefs or policies, and leave aside the other ones. Sorry guys, it doesn’t work like that. Bagehot’s recommendations work as a whole. Here I am going to summarise what Bagehot really said about banking and regulation in his famous book Lombard Street: A description of the Money Market.
Let’s start with central banking. As I’ve already highlighted a few days ago, Bagehot said that the institution that holds bank reserves (i.e. a central bank) should:
- Lend freely to solvent banks and companies
- Lend at a punitive rate of interest
- Lend only against good quality collateral
I can’t recall how many times I’ve heard central bankers, regulators and journalists repeating again and again that “according to Bagehot” central banks had to lend freely. Period. Nothing else? Nop, nothing else. Sometimes, a better informed person will add that Bagehot said that central banks had to lend to solvent banks only or against good collateral. Very high interest rates? No way. Take a look at what Mark Carney said in his speech last week: “140 years ago in Lombard Street, Walter Bagehot expounded the duty of the Bank of England to lend freely to stem a panic and to make loans on “everything which in common times is good ‘banking security’.”” Typical.
Now hold your breath. What Bagehot said did not only involve central banking in itself but also the banking system in general, as well as its regulation. Bagehot attacked…regulatory ratios. Check this out (chapter 8, emphasis mine):
But possibly it may be suggested that I ought to explain why the American system, or some modification, would not or might not be suitable to us. The American law says that each national bank shall have a fixed proportion of cash to its liabilities (there are two classes of banks, and two different proportions; but that is not to the present purpose), and it ascertains by inspectors, who inspect at their own times, whether the required amount of cash is in the bank or not. It may be asked, could nothing like this be attempted in England? could not it, or some modification, help us out of our difficulties? As far as the American banking system is one of many reserves, I have said why I think it is of no use considering whether we should adopt it or not. We cannot adopt it if we would. The one-reserve system is fixed upon us.
Here Bagehot refers to reserve requirements, and pointed out that banks in the US had to keep a minimum amount of reserves (i.e. today’s equivalent would be base fiat currency) as a percentage of their liabilities (= customer deposits) but that it did not apply to Britain as all reserves were located at the Bank of England and not at individual banks (the US didn’t have a central bank at that time). He then follows:
The only practical imitation of the American system would be to enact that the Banking department of the Bank of England should always keep a fixed proportion—say one-third of its liabilities—in reserve. But, as we have seen before, a fixed proportion of the liabilities, even when that proportion is voluntarily chosen by the directors, and not imposed by law, is not the proper standard for a bank reserve. Liabilities may be imminent or distant, and a fixed rule which imposes the same reserve for both will sometimes err by excess, and sometimes by defect. It will waste profits by over-provision against ordinary danger, and yet it may not always save the bank; for this provision is often likely enough to be insufficient against rare and unusual dangers.
Bagehot thought that ‘fixed’ reserve ratios would not be flexible enough to cope with the needs of day-to-day banking activities and economic cycles: in good times, profits would be wasted; in bad times, the ratio is likely not to be sufficient. Then it gets particularly interesting:
But bad as is this system when voluntarily chosen, it becomes far worse when legally and compulsorily imposed. In a sensitive state of the English money market the near approach to the legal limit of reserve would be a sure incentive to panic; if one-third were fixed by law, the moment the banks were close to one-third, alarm would begin, and would run like magic. And the fear would be worse because it would not be unfounded—at least, not wholly. If you say that the Bank shall always hold one-third of its liabilities as a reserve, you say in fact that this one-third shall always be useless, for out of it the Bank cannot make advances, cannot give extra help, cannot do what we have seen the holders of the ultimate reserve ought to do and must do. There is no help for us in the American system; its very essence and principle are faulty.
To Bagehot, requirements defined by regulatory authorities were evidently even worse, whether for individual banks or applied to a central bank. I bet he would say the exact same thing of today’s regulatory liquidity and capital ratios, which are essentially the same: they can potentially become a threshold around which panic may occur. As soon as a bank reaches the regulatory limit (for whatever reason), alarm would ring and creditors and depositors would start reducing their lending and withdrawing their money, draining the bank’s reserves and either creating a panic, or worsening it. This reasoning could also be applied to all stress tests and public shaming of banks by regulators over the past few years: they can only make things worse.
Even more surprising: the spiritual leader of all of today’s central bankers was actually…against central banking. That’s right. Time and time again in Lombard Street he claimed that Britain’s central banking system was ‘unnatural’ and only due to special privileges granted by the state. In chapter 2, he said:
I shall have failed in my purpose if I have not proved that the system of entrusting all our reserve to a single board, like that of the Bank directors, is very anomalous; that it is very dangerous; that its bad consequences, though much felt, have not been fully seen; that they have been obscured by traditional arguments and hidden in the dust of ancient controversies.
But it will be said—What would be better? What other system could there be? We are so accustomed to a system of banking, dependent for its cardinal function on a single bank, that we can hardly conceive of any other. But the natural system—that which would have sprung up if Government had let banking alone—is that of many banks of equal or not altogether unequal size. In all other trades competition brings the traders to a rough approximate equality. In cotton spinning, no single firm far and permanently outstrips the others. There is no tendency to a monarchy in the cotton world; nor, where banking has been left free, is there any tendency to a monarchy in banking either. In Manchester, in Liverpool, and all through England, we have a great number of banks, each with a business more or less good, but we have no single bank with any sort of predominance; nor is there any such bank in Scotland. In the new world of Joint Stock Banks outside the Bank of England, we see much the same phenomenon. One or more get for a time a better business than the others, but no single bank permanently obtains an unquestioned predominance. None of them gets so much before the others that the others voluntarily place their reserves in its keeping. A republic with many competitors of a size or sizes suitable to the business, is the constitution of every trade if left to itself, and of banking as much as any other. A monarchy in any trade is a sign of some anomalous advantage, and of some intervention from without.
As reflected in those writings, Bagehot judged that the banking system had not evolved the right way due to government intervention (I can’t paste the whole quote here as it would double the size of my post…), and that other systems would have been more efficient. This reminded me of Mervyn King’s famous quote: “Of all the many ways of organising banking, the worst is the one we have today.” Another very interesting passage will surely remind my readers of a few recent events (chapter 4):
And this system has plain and grave evils.
1st. Because being created by state aid, it is more likely than a natural system to require state help.
3rdly. Because, our one reserve is, by the necessity of its nature, given over to one board of directors, and we are therefore dependent on the wisdom of that one only, and cannot, as in most trades, strike an average of the wisdom and the folly, the discretion and the indiscretion, of many competitors.
Granted, the first point referred to the Bank of England. But we can easily apply it to our current banking system, whose growth since Bagehot’s time was partly based on political connections and state protection. Our financial system has been so distorted by regulations over time than it has arguably been built by the state. As a result, when crisis strikes, it requires state help, exactly as Bagehot predicted. The second point is also interesting given that central bankers are accused all around the world of continuously controlling and distorting financial markets through various (misguided or not) monetary policies.
For all the system ills, however, he argued against proposing a fundamental reform of the system:
I shall be at once asked—Do you propose a revolution? Do you propose to abandon the one-reserve system, and create anew a many-reserve system? My plain answer is that I do not propose it. I know it would be childish. Credit in business is like loyalty in Government. You must take what you can find of it, and work with it if possible.
Bagehot admitted that it was not reasonable to try to shake the system, that it was (unfortunately) there to stay. The only pragmatic thing to do was to try to make it more efficient given the circumstances.
But what did he think was a good system then? (chapter 4):
Under a good system of banking, a great collapse, except from rebellion or invasion, would probably not happen. A large number of banks, each feeling that their credit was at stake in keeping a good reserve, probably would keep one; if any one did not, it would be criticised constantly, and would soon lose its standing, and in the end disappear. And such banks would meet an incipient panic freely, and generously; they would advance out of their reserve boldly and largely, for each individual bank would fear suspicion, and know that at such periods it must ‘show strength,’ if at such times it wishes to be thought to have strength. Such a system reduces to a minimum the risk that is caused by the deposit. If the national money can safely be deposited in banks in any way, this is the way to make it safe.
What Bagehot described is a ‘free banking’ system. This is a laissez faire-type banking system that involves no more regulatory constraints than those applicable to other industries, no central bank centralising reserves or dictating monetary policy, no government control and competitive currency issuance. No regulation? No central bank to adequately control the currency and the money supply and act as a lender of last resort? No government control? Surely this is a recipe for disaster! Well…no. There have been a few free banking systems in history, in particular in Scotland and Sweden in the 19th century, to a slightly lesser extent in Canada in the 19th and early 20th, and in some other locations around the world as well. Curiously (or not), all those banking systems were very stable and much less prone to crises than the central banking ones we currently live in. Selgin and White are experts in the field if you want to learn more. If free banking was so effective, why did it disappear? There are very good reasons for that, which I’ll cover in a subsequent post on the history of central banking.
I am not claiming that Bagehot held those views for his entire life though. A younger Bagehot actually favoured monopolised-currency issuance and the one-reserve system he decried in his later life. I am not even claiming that everything he said was necessarily right. But Bagehot as a defender of free banking and against regulatory requirements of all sort is a far cry from what most academics and regulators would like us to believe today. Personally, I find that, well, very ironic.