« Mechanical metaphors for monetary policy | Main | Banks, Aggregate Demand, and Aggregate Supply »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Makes sense. This isn't just banks, either, but accounting in general. In Canada, we work by GAAP - generally accepted accounting principles - which are just that, principles. In the end, the principles boil down to "accurately report your complete financial situation, don't fudge it so you look better than you are." The US uses a much more detailed rule book, and each rule change can be very substantial in its impact on accounting practices. This, I think, is where all the recent hoopla about the mark-to-market rules come from. In the US, if you can't sell it, it's worth nothing...until they changed the rule to give exceptions when the market is - entirely subjectively - in distress. In Canada, something like ABCP could probably be reclassified as a hold-to-maturity investment, (y'know, so long as you actually intend to hold it to maturity) which would make its market value irrelevant.

At the end though, the anecdote you put in here is saying that there's a drastic difference in style between the two regulators. But when people say that Canadian regulations are more stringent, they're not talking about what the enforcement folks do, they're talking about what the rules actually are. I also don't know much about banking-specific rules, but it seems to me that a set of principles can actually be more stringent than detailed rules, since there isn't wiggle room in a principle.

Recent discussions on this board on China, gold and China's announcement today....


The paper on the voxeu site was educational.

Will the U.S. have the same experiences as Argentina had?


When it comes to China, you have to learn to read between the lines.

1) The better condition of the Canadian banking system has been put down to supposedly better/stricter regulation and to our generally more cautious temperament. Having spent the last 25 years in the world of telecom regulation and policy, I tend to be sceptical that regulation can be both "stricter" and "better" (although "strict" can have a variety of meanings, I suppose), unless it is also narrower or simpler. Further, regulation is unlikely to be “better”, or even ultimately sustainable, if it undermines or contradicts natural market incentives. It would be interesting therefore to see a good comparison of US and Canadian banking systems from the perspective of moral hazard. The presence of less moral hazard might have provided a rational basis for our more cautious behaviour, obviously.

2) Further to that, there were no bank failures in Canada between 1934 and 1966 (http://www.cato.org/pubs/journal/cj16n1-3.html). Deposit insurance was instituted in Canada in 1967. There were 43 failures between 1967 and 1996 (according to Wikipedia). IIRC, Canada didn’t have a central bank until 1935(ish). As far as I am aware, Canada doesn’t have a history of “too big to fail”.

I would be inclined to explain the numbers as being due to the GD being fresh in everyones memory rather than to the deposit insurance bogey man. Besides, I'm not sure that I buy the moral hazard argument against deposit insurance. The possibility of losing other people's money doesn't seem to provide much incentive for responsible behaviour by bankers. Even with deposit insurance when a bank fails shareholders get wiped out, management still loses their jobs, and bond holders take a haircut.

The only moral hazard created by deposit insurance is that depositors might do business with an institution that they might otherwise stay clear of. Given the information asymmetries that exist between the institution and potential depositors this doesn't strike me as being a big problem since depositors mostly wouldn't have a clue anyway.


I’ve been trying to think of something useful to say on this subject. It’s difficult, given that few people are in a position to compare the two regulatory systems close up. I suspect the article has it about right.

My impression of the Canadian system is that it’s pretty solid. OSFI works through the gamut of each bank’s most senior executives for each major operating division and the risk management division. The risk management division tends to be a focal point, since it is the logical internal epicentre for many of the issues that a regulator would be most interested in. Extremely detailed information is presented. Questions are asked. Requests for more information ... etc. etc., on it goes.

There is one aspect to the larger process that should be highlighted. There are a number of important interest groups within a major bank that converge on the same set of interests from their own perspectives and agendas. I’m thinking here of OSFI, the Board of Directors, senior executive management, the risk management division, the internal auditors, the external auditors, and the rating agencies. Every one of these groups is interested in roughly the same set of information about the bank’s risk. Just as one example, the internal bank audit function can be ferocious. Depending on the personality, ambition, and style its leadership, it can very demanding in setting standards of excellence on other divisions, such as risk management. This sort of dynamic tension plays out before the senior executive team and the CEO, and then affects the nature of the standards that are presented to outside adjudicators such as OSFI and the rating agencies. That’s just one example. Perhaps that’s all generic, and I don’t know how it plays in the US, but it is a considerable source of checks and balances in the Canadian system. My impression is that because of this tension, the important and critical stuff routinely gets fully aired to the top executive and to OSFI.


“One colleague who used to work in banking told me he never actually met a regulator, there are so few. It's more a difference in style, or method?”

Not that surprising; might not have been at the right meetings. They spend much of their time working their way comprehensively through grilling interviews with the executives, of whom there are a lot at a Canadian bank. That’s seems like sort of the inverse of the babysitting approach that the article suggests in the case of the US.

Nick Rowe and the Globe ROB Rara Perkins, are spot on. As a former Canadian banker where I was a Loan Manager and then Mortgage Manager and then Commercial Credit Officer at the 4th largest branch of BMO, I can state that we never ever saw anyone from OSFI or its predecessor. However, we were audited by Internal Bank Audit from Head Office on average every 12 months. These auditors were NOT accountants analyzing debits and credits. The internal auditors analyzed the credit quality, due diligence, and documentation associated with the credit files selected for analysis. The bank audit rules were so strict we were ordered to avoid speaking with or engaging with the internal auditors - unless they spoke to us and requested documents. We were absolutely prohibited from fraternizing with or going for a drink with the bank auditors. I can state that never before or since in my life, was I so terrified as I was during a 2 week internal bank audit.

The audit function was not "rules-based" but principles-based - which was far more frightening and worrisome from a career perspective. While rules are precise and measurable and ascertainable (you broke the rule or you did not), a principles-based approach provided vastly greater latitude to the internal auditors. This was extremely well understood by bank credit managers and made each of us intrinsically conservative.

By contrast, American bank managers were always understood to be more "aggressive". We characterized them as "cash flow lenders" or sometimes, "go-go bankers", while we understood very clearly we were "balance sheet lenders" i.e. security based. The latter is intrinsically more conservative - because if the business had no assets or collateral e.g. new business startups, by definition they would not qualify for the loan.

It should also be noted that as late as the mid 1970s, we were still using British banking training materials. And until the first American in Canadian history was hired as a Canadian bank CEO in 1977 (William Mulholland the first Canadian bank CEO to have any university degree - Harvard MBA), the Canadian banks only hired from grade 12. They would not hire new employees with a university degree as they wanted - like the Jesuits - to recruit them young and then mold and shape them.

The words that were drilled into each banker from the first day were: probity, prudence, due diligence, judgment, rectitude, moral. The bank culture was not legalistic and rule bound nor was it "process-driven" place. It was very hierarchical but not similar to the military. Rather, the hierarchical culture was in many ways similar to the academy - unspoken but understood. "Flashy" or outspoken behaviour was looked down upon and strongly discouraged.

While it is very easy to sneer at bankers - as has always been fashionable in Canada by members of the academy - the Canadian banks have not failed while American and European ba nks have failed miserably. Moreover, the Canadian banks had a code of conduct and honour and morality that everyone in the bank understood and followed. Commercial bankers were remunerated by a flat salary with no bonuses. By contrast, investment bankers never possessed nor understood nor followed any similar code. For any person to label a Canadian banker as e.g. “greedy”, merely demonstrates an impressively comprehensive ignorance about the culture and practices and history of Canadian banking.

Final anecdote. My father was born in England and worked for a bank in London shortly before the war. He told me that on the very hottest days (no air conditioning in those days) on very special occasions when it was unbearable, the office director would stand up and announce: "gentlemen: perhaps we can doff our coats today".

Fast forward to 1979 at BMO Ottawa Main (old marble bank on Sparks Street mall opposite Parliament and beside National Press Club) on a very hot muggy day when the air con broke down. The temperature was over 35 C in the branch and consequently I removed my suit coat. Shortly thereafter, the Administration Manager of the bank came to me and whispered that he wanted to speak to me. We went into a private room and he said that the Vice President of the branch thought that "perhaps you may want to put your suit coat on". I protested that it was insufferably and unbearably hot. He smiled sadly and said yes, the Vice President understands this, but we are a bank and bankers do not act like that.

It's great to see the comments here finally coming in, as I knew they would. (Hi Ian!).

I'm going to largely let you guys roll, since again, this is not my area.

Just a couple of questions:

1. I'm still not fully clear on the rules based vs principles based distinction.

2. Ian makes an interesting distinction between cash flow vs balance sheet lending. This is applied to the businesses to whom the banks were lending, I take it. Several months ago JKH and I were comparing the cash flow vs balance sheet approach to looking at central bank solvency. I, as an economist, argued for a cash flow approach. JKH, as a Canadian banker, argued more for a balance sheet approach (correct, JKH?). There's central banks at one end, ordinary borrowers at the other end, and commercial banks in the middle. What matters for a comercial bank? Cash flow or balance sheet? Any US vs Canada differences there?


The question of solvency is complex for any economic entity, in my view.

The fact is that capital and capital adequacy is clearly defined as a balance sheet concept in the case of commercial banks.

That said, interpretations of “solvency” can be more subjective and presumptive than regulatory definitions of capital.

It is quite reasonable to look at the current capital position of a bank, and then consider the operating income that may be expected over a given future time period, in order to get a picture of longer term “solvency”. It becomes shaded as to whether one is doing a capital change simulation with income assumptions, or a “solvency’ test. What you’re really doing in this case is projecting a change in capital position from today to some future point, based on current balance sheet risk and expected future operating income. To the degree that you’re willing to notionally grant that future income some sort of present value, you’re making a statement of interpretation about current solvency. But it’s a subjective statement, and highly dependent on the assumption you’re making about both balance sheet risk and expected net cash flows in the future.

An example of this is the upcoming stress test results for US banks. I expect they will be framed in the context of the ability of banks to withstand economic risk, given their current capital position, the balance sheet risks they face now and may continue to face, giving some consideration to their ability to generate net operating cash flow over the stress test time period, and then projecting their capital position at the end of that period. So the test becomes an expected change in capital position that incorporates expected operating cash flow and balance sheet risk. It’s both. And all of that will be wrapped up in some loose assessment of “solvency”. It’s possible the results of the test may even shy away from using the word "solvency". If they do, they'll define it.

But at the end of the day, current balance sheet is the priority. I doubt there's a difference between US and Canada.


In the case of central banks, I objected to the approach of defining solvency on a formulaic basis, using the present value of seigniorage. That said, there’s a stronger argument to do this with central banks that with commercial banks, because seigniorage is closer to risk free than commercial bank net operating income. But I’m still not comfortable with it, because it shouldn’t be a formula. It’s an interpretation. Moreover, I dislike the concept of solvency per se for a central bank, because the bank’s true capital structure is normally an extension of government deficit financing (except with credit easing), so on that basis I'm uncomfortable with either the balance sheet or the cash flow approach. I think it’s economically artificial to segregate central bank solvency, although it’s more meaningful I suppose when you consider the issue of political independence, although that’s not clear either.


CalculatedRisk has a couple of blog posts on bank solvency. I'd be interested in your take:

Part 1:

Part 2:


I’ve only just taken a quick look at the Calculated Risk posts. They look like a good primer, and really quite good at explaining a complicated sequence of events in the credit crisis. Not surprising. He’s one of the best financial analysts of the economic bloggers.

I should clarify something, though. There are probably three different ideas involved in looking at solvency, as it relates to what we've covered off here, rather than two. The first two are the alternatives that I discussed where solvency is viewed either as current balance sheet condition or current balance sheet condition plus expected operating earnings over some specified time period. I know this was a distinction Nick and I discussed at one point as it related to central banks.

Calculated Risk also discusses the pure liquidity approach, which is fact is more commonly looked at than the second alternative I noted. I think the best way to illustrate the liquidity approach in relation to solvency is to visualize a hypothetical bank that has zero capital, and makes no money. Not a great operation. But can it “survive” for a period of time? In theory, this bank could service its cash flow liabilities provided the realized cash inflows from assets arrived before cash outflows required to pay off liabilities. It would have to be “liquid” in the sense that its assets were shorter duration (in a cash flow sense) than its liabilities.

So solvency can include all three aspects.

E.g. from Calculated Risk:

“The banks are facing huge additional losses for these legacy assets, and these losses will make some banks "balance sheet" insolvent (liabilities will be great than assets). However, the bank is not insolvent in the business sense, because the bank can still pay their debts as they come due - at least for now. “

He’s referring to the current capital condition when he says “balance sheet insolvent”. And he’s referring to the pure liquidity approach when he says “in the business sense”.


“These "??????" assets are either future retained earnings or additional money from the government. Although the bank is balance sheet insolvent, the bank will never be business insolvent because the government will continue to provide money to cover losses.”

This is interesting because he refers to “future retained earnings”, which is more along the lines of the second alternative I described, as opposed to the pure liquidity approach.

The “cash flow” approach to solvency is potentially confusing terminology. It could refer to the incorporation of future expected profits or it could refer to the pure liquidity approach. Those are two different ideas. The first idea extends solvency from current balance sheet value to expected balance sheet value. The second idea interprets solvency not from a capital or value perspective, but from a principal cash flow perspective. I remember discussing the first idea with Nick when looking at central banks, but maybe we discussed the pure liquidity approach as well at some point.

The fact is the terminology is what you define it to be. It’s best to be very specific about what is being analyzed, rather than get hung up on particular language for definitions.


My personal preference is to keep capital, expected earnings, and liquidity all analytically distinct. If talking solvency, I start (and sometimes end) with capital, but work out from there, qualifying it if necessary with future earnings and liquidity prospects.

Yes Nick, the comments you have excerpted seem accurate from my experience. Keep in mind that banking is far more concentrated in Canada than in the U.S., especially historically, which sets the tone for a different competitive (or not) environment which goes along with a different relationship between regulator and regulated.

JKH: Your "3 perspectives on solvency" struck an immediate chord in my head. Bernie Madoff was liquid right up to the very end, but had always been insolvent from both the balance sheet and expected future values of cash flow perspectives.

Agreed, Nick. Ponzi schemes are liquid, up until the moment they're not.

I think I finally get the zombie metaphor. You guys aren't making me feel any better. I think I'll go read about swine flu to cheer up a little ;)

"I'm still not fully clear on the rules based vs principles based distinction"

I am wondering whether it might be analogous to the distinction one finds between competition law (e.g., thou shalt not engage in anti-competitive activity or predatory pricing) vs. sector-specific regulation (thou shalt not price below some measure of cost), the former being principles-based and the latter being rules-based.

Rules-based regulation is more prescriptive, permits less flexibility (at least with respect to proscribed behaviour) and is clearer ex ante in the sense of determining whether or not one is off-side. Principles-based "regulation", at least in the competition law case, is less prescriptive and more flexible in that it permits a wider range of behaviour but, as one approaches the boundaries of what might be permitted by the regulations, is less clear ex ante. The principle-based rules are open to interpretation. I suppose that that uncertainty under principles-based regulation may prompt risk aversion on the part of the regulated company, i.e., in order to be well inside the rules.

Re: Principle based vs. Rules, Felix Salmon has a food post from a while back. Drawing a comparison to how football vs. soccer is officiated:


Hi Alan. Thanks for the link. I am not sure that the soccer/football analogy quite does it for me - fewer rules and less precisely enforced rules (i.e., soccer) is not the same as principles-based regulation, I wouldn't have thought (although the rule against "obstruction" in soccer might be more in the nature of a principles-based regulation?). I do agree with his point that "a bad or biased ref can ruin a soccer match in the way that no bad ref can ruin a football match". Principles-based regulators are likely going to have more discretion in applying regulations and determining whether compliance has taken place.

The summary of the different styles of regulation is correct, but the survival of the Canadian banking system is due to three other factors. First, under the Bank Act, Canadian banks cannot give a mortgage for more than 75% of the value of a house or condominium, so a decline in real estate values is less damaging.

Second, Canada does not have the same tax incentives for borrowing.

Third, and perhaps most important, the average Canadian bank is much larger than the average US bank and has a much more diversified mortgage book. US banks, which are on average much smaller and much less geographically diversified, have a much greater need to securitize mortgages, which breaks the link between the person who decides to extend the loan and the person who bears the consequences if the borrower defaults. In Canada, a loan officer who made a lot of NINJA loans would be looking for a new job because they would all stay on the bank's books.

Sone of the differences in regulatory approach may also be due to cultural differences between the US & Canada. I read somewhere that Canadians were generally more risk-averse than Americans partly because in the US if you fail you can move to another part of the country and re-invent yourself. In Canada, with 1/10 the population, we cannot hide quite so easily.

Hey Nick. You're sort of famous:


Thanks David!

And Colby Cosh gets it exactly right:

"...Nick Rowe,... , spitballed a list of bullet points about Canadian finance in a comment thread at a monetary-policy Weblog. It’s the kind of thing a scholar might more typically put on a bar napkin — it wasn’t even an actual Weblog post, and it was prefaced with: “I don’t really know why Canada’s banks seem to have done well” and “Maybe we just got lucky” — but analysts seized upon it hungrily; it even ended up appearing, without attribution, on the Web site of the Atlantic magazine."

Yep, it took me about 5 minutes to write, and I confessed I didn't really know what I was talking about. I was amazed it went semi-viral.

It just shows the power of getting the timing right, and writing a list of points.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad