The Big Secret: Banks are banks

I have a confession to make. During the financial crisis I secretly bailed out the Bank of Montreal. I was part of a vast conspiracy of millions of other Canadians doing the same thing. It's worse. My personal bailout program for the Bank of Montreal started long before the financial crisis. It started when I first arrived in Canada. It's even worser. The Bank of Montreal still has not repaid me what it owes, and my bailout continues to this day. Without this bailout program, the Bank of Montreal could never have continued in business; it would have gone bust centuries ago. This secret bailout program is far bigger than the total market value of Bank of Montreal shares, so the Bank of Montreal is deeply underwater! It would have been much cheaper for me and my fellow conspirators to have bought the Bank of Montreal outright. And we could have earned a far higher return on our investment if we had done that. The Bank of Montreal eagerly borrowed every single dollar we were prepared to lend it in our bailout program, and made a profit by lending those dollars to other borrowers. We should have just lent our dollars directly to those ultimate borrowers, and cut out the middleman.

Or, to say the same thing more simply: I have a chequing account at the Bank of Montreal, and the Bank of Montreal is a bank.

There are real questions that can be asked about the relationship between government and banks, and between central banks and banks. Should central banks act as lender of last resort to commercial banks? If so, in what circumstances and on what terms? How should monetary and fiscal policy be conducted during a recession and/or financial crisis, and how much and by what methods should central banks try to influence the behaviour of commercial banks in order to attain their monetary policy objectives?

I do not know the answers to those questions, but I think they are important. For what it's worth, my hunch is that the channel between central banks and commercial banks currently plays a far larger role in the monetary policy transmission mechanism than it should. And I am very uneasy about the lack of transparency that seems almost inherent in the role of lender of last resort; it's in stark contrast to the transparency of inflation targeting.

David Macdonald has collected the facts of Bank of Canada lending to the commercial banks during the financial crisis. No worries. But his interpretation of those facts is simply to restate those facts in the same way that I restate my facts in my opening paragraph above.

For example, David says (pages 8 and 9):

"In fact, several banks drew government support whose value exceeded the bank’s actual value. Canadian banks were in hot water during the crisis and the Canadian government has remained resolutely secretive about the details."

And on page 28:

"Total support for CIBC was worth almost one and a half times the value of all the company’s shares in March 2009. In fact, almost every day from mid-January 2009 and the end of April, CIBC was completely “underwater,” receiving support worth more than the value of the company."

All banks always borrow more than the value of the bank. (OK, I expect that's not strictly true, and if I searched I could find a counterexample, where a newly-formed bank had raised capital but hadn't yet got its act together to attract some deposits and make some loans.) That doesn't mean the bank is in "hot water", or "underwater". It means it's a bank. Hell, most homeowners with a mortgage worth more than 50% of the value of the property are "underwater" or in "hot water" and being "bailed out" by the mortgage under that definition. Banks typically borrow more than 90% of what they lend. They always live "underwater" and in very "hot water" indeed. That's where banks live.

Banks borrow and lend. Banks' balance sheets are much bigger than the market value of their shares. That's what banks do. Banks are banks.

(Oh, and the one really interesting and important thing about banks, which is what sets them apart from you and me and other financial intermediaries, is that some of banks' liabilities are used as media of exchange. That is the really important fact about banks.)

What we really need is a counterfactual conditional. What would banks have done instead if the Bank of Canada had not provided those loans? Would they have gone bust? Or would they have just had smaller balance sheets? And that analysis is exactly what I didn't find in reading David's Report.

If I hadn't got the mortgage, what would I have done? Maybe I would have gone bust, or maybe I just wouldn't have bought the house? You can't answer that question just by noting that the mortgage was more than my equity in the house.

[Update: let me make the central point even more strongly. When a recession appears more likely, the central bank may offer loans on easier terms. That's what central banks are supposed to do. And commercial banks may accept those loans. That's what banks are supposed to do. We wanted the banks to accept that "bailout" money, whether they needed it or not. The time to get worried is not when banks act like banks and accept that "bailout". The time to get worried is when banks stop acting like banks and don't accept that "bailout". (Or, like many US banks, they just park the "bailout" money back at the same Fed which lent it out in the first place). It's when banks stop acting like banks that we should start to wonder what use they are in the big scheme of things. We wanted the Canadian banks to borrow the "bailout" money. We wanted them to act like banks. The very last thing we should be doing is criticising them for doing this. The very worst result of this Report would be if it made banks more reluctant to borrow and lend in any future recession. That's the last thing we need.)

I own some BNS shares. The Carleton University pension plan almost certainly holds Canadian banks' shares. I have a chequing account at BMO, as mentioned above. I'm probably biased.

Comments

[Deus-DJ]

Much agreed, good post. However, I don't like this:

"I do not know the answers to those questions, but I think they are important. For what it's worth, my hunch is that the channel between central banks and commercial banks currently plays a far larger role in the monetary policy transmission mechanism than it should."

Why do you need a hunch when MMT has been on your blog for the past 2-3 years explaining these operations back and forth? Can't you give credit where credit is due (no pun intended)? and how else could the channel operate if the central is changing the level of bank reserves (oops I do forget you're referring to Canada with no RR). Anyway what would bother you about this relationship anyhow?

[Trevor]

Excellent post.

[Nick Rowe]

DDJ: Wow! Thanks! That must be my first post you've said you liked!

Good questions. I can't articulate my answers very well. Short version: because a bank, almost by definition, is an accident waiting to happen. And because I think monetary policy is very important. So I worry that such accident-prone organisations are related to such an important policy. And we can imagine a radically different form of monetary policy that is less reliant on the relationship between banks and central banks via reserves.

[Frances Woolley]

Nick - I was trying to think of some witty remark about your closet socialist or crony capitalist leanings, but realized nothing could match the awesomeness of this post so gave up.

[JP Koning]

Nick, good points, I agree. The CCPA report is a bit shrill. That being said, I do agree that the lack of transparency in LOLR services is a problem - it breeds resentment which in turn spawns reports like the CCPA's. That's why I lean towards market-provided LOLR products such as traded liquidity insurance/liquidity options.

[Mike Sproul]

"what sets them apart from you and me and other financial intermediaries, is that some of banks' liabilities are used as media of exchange."

What about gift certificates, Disney dollars, WalMart dollars, and iphone dollars? What about the time I paid my mechanic with my personal IOU for $20, only to have the mechanic's employee pay me for some tutoring with the same IOU?

We're all bankers now!

[Jeremy]

Great Post! This report was a farce, news coverage was worse!

[David]

Thanks Nick -- this definitely needed saying. Can't repeat this enough.

On your comment above "Short version: because a bank, almost by definition, is an accident waiting to happen."

I presume what you mean by this is that a good chunk of a typical bank's liabilities are made demandable (these days, convertible into government cash, in the old days, convertible into specie). In fact, any agency that issues short term debt to finance longer term (read: illiquid) assets is potentially an accident waiting to happen. The question for theorists is: what role does short-term debt of this nature play in the resource allocation mechanism? I suspect that it is a second-best solution to an agency problem. If it is, then government interventions can easily make matters worse, than better. (And by worse, I include the outcome of perfect financial stability--as exists in the state of financial autarky).

[Bob Gratton]

What would banks have done instead if the Bank of Canada had not provided those loans?

Well, you can ask yourself why did they get them in the first place?

If you can't free up capital, you can't lend (i.e. create money) because you need an amount say 6% or whatnot in reserves.

So if the value of your liabilities shrink, you need more capital to cover your new loans - you raise capital. Can't sell stock, interest rates in the private market were too high, if available at all - here comes the lender of last resort.

and yes, without that lender, you would go bust - simply because you need to raise a capital amount to maintain solvability in an environment that is dry, dry, dry of liquidity for a short period.

Banks don't "borrow what they lend"....they need capital to create money. That's it. its called banking. And what they do is discount bills. Discount bad bills, go bust.

[Antonio de Sousa]

Timely since in the Toronto Star they are running a piece from the Centre for Policy Alternatives about how the Harper government secretly bailed out all the Canadian banks in the latest financial crises.

Haven't read the report myself, but frothing commentary on the Star site point to another example of Harper's dictatorially and secretively saving the banks on the backs of the Canadian tax payer.

Perhaps someone can comment, what does "bail-out" mean?, is the role of the Central Bank not to "bail-out" banks? How can taxes collected by the government be used to bail-out banks, what is the mechanism?

I have to say that I don't understand banking, the financial interactions and Toronto Star stories about them. It just confuses me. Are we or aren't we supposed to bail out banks if by not bailing them out we threaten the normal working of the real economy?

[Ryan]

Nick, FWIW, people like myself in occasional blog comments, and people like Steve Horwitz in occasional guest lectures, provide exactly the kind of counter-factual you're looking for; but every time we do, the response we get is that you liken a free market money system to a barter system. There's a big disconnect here. You need not agree with the counter-factual, of course, but if you want to answer the kinds of questions you list above, then that requires stepping back and second-guessing some fundamental assumptions about money and credit.

I think Mises' analysis of money and credit in Human Action and elsewhere (my personal favorite is "Causes of the Economic Crisis") provide a profound and deeply persuasive synopsis of the "counter-factual."

I think a lot of people like me are waiting for mainstream economists to address Mises' arguments head-on. And what I mean by that is, it would be nice if many mainstream economists actually and seriously read Mises' essays and commented directly on them, as opposed to simply responding to isolated Hayek quotes out of context or the heavily-summarized views of the LVMI crowd.

In my opinion, there is a big, important thing to be addressed there, and it can be kind of disappointing to see such sparse treatment of it by the mainstream economists.

[Jacques René Giguère]

Agree. But
On this blog we more or less understand what a bank is and what it does and why. We have some understanding of what oa CB is, what it does and why. We may have vaque inklings of what money is etc. We may even have some vague concepts of what monetary policy is etc.

But in the public mind, banks "received free money". Noboody except banks receive "free money".
Nobody except banks receives "free money" and them serves you morality lessons.

A banker leaves the BoC in his limo after getting his bailout. He sees a panhandler and order his chauffeur to stop. He lowers the window and shout:"Get yourself a job, you lazy bum! "
Only bank receives "free money" and then tell you to get a job.
Or at least this is how the public sees it.

[dwb]

What we really need is a counterfactual conditional. What would banks have done instead if the Bank of Canada had not provided those loans? Would they have gone bust? Or would they have just had smaller balance sheets? And that analysis is exactly what I didn't find in reading David's Report.

wrong question though. we live in a probabilistic universe. its not a certainty either might have happened, its a question of what the probabilities were (and whether there are multiple equilibria, potentially some with critical points whereby the outcome becomes self-reinforcing as we move closer).

Liquidity crises (which is ultimately what topples a bank) appear to be such cases (which is why rumors can destabilize a bank).

i am not a big fan of armchair quarterbacking for that reason (whats the equivalent hockey metaphor?).

I was in an accident once, a car was coming down the road to me head on. The car hit one ahead of me head-on, and the next car in line swerved and sideswiped a tree (leaving a clear path between me and the still-moving car).

Insurance company said to the owner of the car who hit the tree: its your fault (!). But he was swerving to avoid a worse collision!

the moral of the story is: a) one might have to take responsibility for a lesser evil to avoid a much bigger one; b) sometimes the best you can do is ensure sufficient preventative measuers are in place. In the accident case, seatbelts and airbags. in the bank case: liquidity and capital.

[Phil Koop]

Great post, except that you forgot to add that you "have remained resolutely secretive about the details." OMG!

[ianlee]

Nick - superb post

David (author of the CCPA "report") is one with Moliere's Monsieur Jourdain's "discovery that he has "been speaking prose all my life".

I did read the CCPA report and sprinkled throughout the report, are quotes from Gov Carney and Finance Minister Flaherty annoucements of the "secret" liquidity support. Not clear how it is secret - when he is quoting the Gov and Finance Minister annoucements as evidence of the support. Support empirically verified yes - secret - no.

e.g. “It was a good thing we didn’t press pause when we provided over $30 billion
of liquidity to the Canadian banking system. It was a good thing the government
of Canada didn’t press pause when it provided...very timely and effective
term liquidity to the Canadian banking system.” Mark Carney, pg. 10, CCPA Report

David failed to note that if the aggregate support had been broken down by bank, it could have caused problems for some of the perceived weaker banks ie those that received more support than other banks. Indeed, Carney has spoken about the importance of confidence in the banks on various occasions (as did Geithner frequently re TARP).

David also used the word "bailout" at least 15 times in the report - even while repeatedly discussing the banks' record profits in Overall Bank Utilization pg 20-30.

David also compared the $114 B in liquidity support provided to the Cdn banks to the market cap of the banks - instead of expressing it as a percentage of total bank assets in Canada ($3.4 trillion) - which is an approariate way to normalize and comp;are the data - and which would have revealed that GoC and BoC support approximated 2.5% of total Cdn bank assets.

Moroever, he failed to compare it to level of support as a percentage of GDP, provided by BoE, BoJ, ECB.

Finally, he characterized the banks' SALE of CMHC insured mortgages to the GoC - as a bailout - when it was a sale of an asset from one entity to another entity.

Or as Freud allegedly said, sometimes a cigar is simply a cigar.

[Kevin Milligan]

Hi Ian,

I don't think much of the CCPA report for many reasons, but I do think they have a point to make about the transparency of the allocation of the intervention. In particular, you say:

"it could have caused problems for some of the perceived weaker banks ie those that received more support than other banks. "

I say that is exactly the kind of information that market participants ought to have--if not at the time of crisis, then at least now; ex post. If it causes 'problems' for weak banks to be exposed as weak banks, I think that is called 'the market doing its job'.

[Determinant]

Interesting that you mention the Bank of Montreal, Nick. I've banked there since I was six. It will celebrate its 200th Anniversary in 2017. Expect the requisite show. But it has survived through thick and thin in those two centuries and its relevant to put in perspective just how much trouble the bank was in in 2008.

My sense is that it was uncomfortable but not dangerous, no intervention would have meant curtailed lending but not endangered the life of the bank.

[ianlee]

Kevin - yes to transparency at the aggregated level. I was in the Green Room at CTV News Channel this morning speaking with David (he went first and then I followed). I pointed out that the GoC DID disclose the aggregated level of liquidity support - but did NOT provide disaggregated data. David agreed completely with this statement.

Thus, the issue is not about secrecy vs transparency of full reporting for budgetary or accountability reasons.

Rather, the debate concerns whether the public should know about the particular levels of support for specific banks e.g. TD vs CIBC.

This morning David told me that he spoke with BoC re filing an access request and the BoC advised him they would deny it. Presumably due to commercial confidentiality or perhaps my argument, to ensure stability of financial instituions and not create a run on a bank.
As every economist and banker knows, "confidence" in a firm or a bank is intangible and elusive. But when it collapses, well ... - remember Nortel.

BTW, today at Economic Club Carney spoke and stated unequivocally that liquidity support is NOT a bailout.

I made the same argument this morning:

David: link here NR
My response: link here NR

[Reverend Moon]

link to Interfluidity post here NR

[Patrick]

It appears that somebody didn't do the Bagehot reading assignment ...

[Simon van Norden]

Nick: Great post. Wish I could write intros like yours.

But reading through the post and the comments, I'm at a loss to know whether govt support of the Canadian banking industry (broadly defined to include the BoC's actions) cost anything. Or when we might find out, if it did. Like you, I don't agree with MacDonald's call to "release the full details of how much support each Canadian bank received, when they received it, and what they put up as collateral." I'd be more interested in knowing whether
- all support has been repaid (with interest at least equal to the government's opportunity cost of funds)
- what assets (if any) the BoC and Govt. acquired from troubled institutions, at what price, and their liquidation values.
I suspect the answers to several of these questions might already be in the public domain. Perhaps you or other readers can help me find the answers.

[Nick Rowe]

Thanks Trevor and Frances!

JP: "That being said, I do agree that the lack of transparency in LOLR services is a problem - it breeds resentment which in turn spawns reports like the CCPA's."

Agreed. But I'm not sure that this case was a LOLR action for the banks themselves (as opposed to the rest of the economy).

"That's why I lean towards market-provided LOLR products such as traded liquidity insurance/liquidity options."

I'm not sure whether or not that could work, in a system-wide liquidity crisis, if it wasn't backed by the ultimate creator of money.

Mike: "What about the time I paid my mechanic with my personal IOU for $20, only to have the mechanic's employee pay me for some tutoring with the same IOU?"

The exception that proves the rule! The fact that you remembered that particular case shows it's rare. Most Wicksellian triangles have many more than three sides, and the corners don't all know each other.

Jeremy: thanks! I stumbled across the news this morning, thought "WTF?", had a look at the Report, and wrote this.

David: Thanks! "I presume what you mean by this is..."

Yep. That's pretty much what I meant. Banks borrow short, safe, and simple, and lend long, risky, and complex. Presumably they do change the allocation of resources, otherwise they wouldn't need to exist. Whether or not the government (central bank) is needed to help them do this, by acting as LOLR, I still see as an open question. Banks as we know them aren't the only way to do this job. But second-best interventions in an agency problem can sometimes help. Depends on the intervention.

Bob: "What would banks have done instead if the Bank of Canada had not provided those loans?

Well, you can ask yourself why did they get them in the first place?"

Because the Bank of Canada wanted them to increase lending to other borrowers, and rather than do that itself, the Bank of Canada bought some of the existing assets off the banks?

Antonio: "Perhaps someone can comment, what does "bail-out" mean?"

Good question, when applied to banks. But one not answered, or even addressed, by this Report. I would say there are two conditions: that the bank(s) would have been in trouble without the bailout; and that the bailout was costly to the government. Neither of which I could see being argued for in the Report.

"Are we or aren't we supposed to bail out banks if by not bailing them out we threaten the normal working of the real economy?"

Most economists would say that we need to bail out banks (in the full sense of "bailout") under some circumstances, because the costs of not bailing them out could be even worse. But many of us find that answer unsatisfactory, even if we accept it under present arrangements. It's an open question.

Ryan: "You need not agree with the counter-factual, of course, but if you want to answer the kinds of questions you list above, then that requires stepping back and second-guessing some fundamental assumptions about money and credit."

Yep. It's hard to address those questions without going deep. I can't remember HA and TMC well enough to say for sure, but I don't remember a big "Aha!" moment that totally changed my views on everything concerning money and credit.

"In my opinion, there is a big, important thing to be addressed there, and it can be kind of disappointing to see such sparse treatment of it by the mainstream economists."

Maybe it's not that we're not trying. Maybe some of these questions are just hard. Haven't you seen me desperately banging my head against some of them? Superman leaps tall buildings in a single bound. The rest of us mortals try to leave a good set of scratches reasonably high up the side of the building. Then argue about whose scratch marks are highest, and whether we've got the right building.

Jacques Rene: "But in the public mind, banks "received free money". Noboody except banks receive "free money".
Nobody except banks receives "free money" and them serves you morality lessons."

Yep. It's even worse when you think that banks can create free money, which they both can and can't. And then charge interest on it! And where does interest come from anyway?? Money and banking is fertile ground for conspiracy theories, of both left and right. Social Credit, for example.

dwb: "wrong question though. we live in a probabilistic universe. its not a certainty either might have happened, its a question of what the probabilities were (and whether there are multiple equilibria, potentially some with critical points whereby the outcome becomes self-reinforcing as we move closer)."

I would say that your question is a better version of my question. If he had first answered my ex post question, we could then have asked your ex ante version. We never got anywhere near step one.

Phil: thanks! Yep. And I'm still keeping mum about how much exactly I bailed out BMO!

Ian: thanks! I wondered if you would have gotten involved in this story. Good additional points and facts.

Kevin and Ian: yep. That's why I said that lack of transparency seems to be almost inherent in the role of LOLR. Not just secrecy. But if we could formalise the rules, and put Bagehot into a computer program, I wonder if we would even need a LOLR in the first place?

Determinant: "My sense is that it was uncomfortable but not dangerous, no intervention would have meant curtailed lending but not endangered the life of the bank."

That is my sense too, but i don't know for sure. Part of the problem is that we need to look at this not just from the perspective of the single banks, or even of the financial system as a whole. We need to look at the whole macroeconomy. If NGDP had fallen far enough, if the BoC had acted differently, not even the safest bank would have been safe (though they did survive the Depression).

Rev Moon. Yep. Worth a re-read.

Patrick. I don't think we need to go as far as reading Bagehot, in this case. Just remember that a bank is a firm whose primary business is both to borrow and lend. Most firms say "yes" when you offer them more of the business they normally do. It wouldn't be surprising if you or I turned down an offer of a loan on easier terms. Because we are not banks.

[Nick Rowe]

Thanks Simon! The intro was what took all the thought. That's what makes me an econoslut! The rest was just ECON1000, plus recycling half-remembered lines from other bloggers (probably Brad DeLong or Willen Buiter or someone).

The ex-post accounting question of the government's profit or loss would be worth knowing. But only as a start. The ex ante counterfactual conditional is the biggie.

[Bob Smith]

So much for secrecy, the CCPA could have saved itself a lot of trouble uncovering the "secret" $114 billion "bail-out" by citing this chapter of the government's Second Report on the "Canadian Action Plan" (thank you Michael Ignatieff) to the effect that: "Over $115 billion in financing support has been provided, all of it on a commercial basis to protect the taxpayer"

"I'm at a loss to know whether govt support of the Canadian banking industry (broadly defined to include the BoC's actions) cost anything."

The government claims that the terms were "commercial", but who knows. I do have a vague recollection of department of Finance press releasees crowing about how favourable the terms were, but I'm too lazy to look them.

[Bob Smith]

Finance did prepare this report, seemingly for parliament, which suggests that the Insured Mortgage Purchase Program was structured in such a way as to ensure a spread between the government's cost of finance and the return on the mortages being purchased (at minimal addition risk to the government - there was some counterparty risk on interest rate swaps for variable mortgages).

There's this press release from Finance to the effect that the IMPP program was producing a "positive return" , though I'd take anything out of Finance with a hefty grain of salt.
[edited to embed links NR]

[Stephen Gordon]

I do know that a Senior Official with whom I had a corridor chat about it was very satisfied with how it went. "We made money on it" is how he put it.

BTW, I add my voice to the chorus of praise for this post. A gem.

[Alex]

Nick,

"When a recession appears more likely, the central bank may offer loans on easier terms. That's what central banks are supposed to do."

Should ordinary people get bank loans on easier terms in those circumstances too? Preferential treatment of banks creates hostility between population and bankers.

[Gregor Bush]

Great job Nic,

I remember reading a Bloomberg article about Fed lending at the discount window being larger than TARP and the AIG bailouts combined but that none of the Fed loans had any political oversight or some such nonsense. The article was written as if the author had uncovered some great scandal.

MacDonald and the author of the Bloomberg article don’t seem realize what a central bank is saying when it announces a target for the overnight rate: that the central bank has pledged to lend in unlimited quantities to prevent the rate from rising above its stated target. The central bank is ALWAYS promising to lend massive amounts of money if need be – but it’s only times of financial stress that banks take them up on the offer. In normal times they mostly borrow and lend from each other. But in a crisis banks are unwilling to lend to one another and would prefer to borrow from the discount window.

Progressives 100 years ago understood all of this very well. Indeed, they were the driving force behind the creation of the Fed which they demanded perform precisely this function to avoid a repeat of the Panic of 1907. But today’s progressives seem to not grasp this at all.

[Bob Smith]

What's really frustrating is that had the government said something along the lines of "we'll let the market take care of itself" and a bank had failed, the CCPA would have (quite fairly) blasted them for sacrificing the Canadian economy on the alter of neo-liberal free-market ideology (did I get the jargon right?). Intead, the government came up a clever way of providing the banks with liquidity without bailing them out or taking on any significant risk (and making a buck or two in the process) and they get blasted for bailing out the banks.

Anyone remember the old Mad Magazine cartoon about how you can't win with a bigot...

[Kevin Milligan]

"But today’s progressives seem to not grasp this at all."

I'm bemused that the 'progressive' position appears to be to just let banks fail; liquidate them in a crisis. That seems more like what the internet-trained Austrian economists advocate.

I wonder what internet blog conspiracy the CCPA will take up next. Uncovering the 'secret true' inflation rate? Will they start spelling things internet comment-style too? You're a looser! That's rediculous!

[Nick Rowe]

Stephen: thanks! And thanks for tweeting it. Just wish I had written this yesterday. Seems like the newspapers have now moved onto other things.

Alex: "Should ordinary people get bank loans on easier terms in those circumstances too?"

Again, the whole point of banks is to both borrow and lend. That's what banks do. They don't just borrow. And the whole point of the "bailout", AFAIK, was to increase the supply of bank lending, so that by offering more and cheaper loans to banks, banks would in turn offer more and cheaper loans to their customers.

Now it's true, the Bank of Canada could have made loans directly to ordinary people and businesses. And we could discuss the merits of whether or not the central bank should get directly involved in retail banking. But again, that is not normally what central banks do, and it would have been difficult for the Bank of Canada to have hired a lot of loan officers and set up a whole branch banking system very quickly when the recession appeared imminent. Instead, it stuck to its normal business of being banker to banks and the government only. It doesn't accept deposits from you and me (except when we hold currency).

Though, from what I've heard, the government did get Export Development Canada (which was already in the retail business and (presumably) knows what it is doing) to expand its business and make more business loans.

Gregor: thanks!

"The central bank is ALWAYS promising to lend massive amounts of money if need be – but it’s only times of financial stress that banks take them up on the offer. In normal times they mostly borrow and lend from each other."

and, I would add, from normal people and firms.

This is how I think of it. In normal times, if the Bank of Canada wants to loosen monetary policy to prevent inflation falling below target, it lowers the overnight rate target, and banks in aggregate automatically start to create loans and deposits simultaneously, lending and borrowing from ordinary people and firms, which creates a monetary hot potato, and Aggregate Demand expands, until the Bank of Canada reckons they have done enough and raises the overnight rate back up again to stop inflation rising above target. But yes, in the financial crisis the Bank of Canada thought this wasn't enough and decided to make loans available on a longer term basis than just overnight. And the banks, as the BoC hoped, took up the offer. (And in hindsight, and perhaps foresight too, the Bank of Canada can be faulted for not having been aggressive enough in doing this, and preventing NGDP falling relative to trend, but now I am going off topic, and will restrain myself).

"Progressives 100 years ago understood all of this very well. Indeed, they were the driving force behind the creation of the Fed which they demanded perform precisely this function to avoid a repeat of the Panic of 1907. But today’s progressives seem to not grasp this at all."

Bingo! Did Ron Paul manage to infiltrate the CCPA?

[Bob Smith]

Kevin: "internet-trained Austrian economists"

Awesome!

[Bob Smith]

Curiously, the CCPA has less to say about ACTUAL corporate bailout, notably of GM and Chrysler (not that there weren't reasonable arguments for intervention in those cases - but those were bailouts). I wonder why that is...

In unrelated news, I see that a number of CAW employees(Jim Stanford, Bill Murnighan) write for the CCPA and the CAW is listed as a supporting/participating organization as part of their Trade and Investment Research Project.

[Alex]

Nick,

"Now it's true, the Bank of Canada could have made loans directly to ordinary people and businesses."

Nope, there are many other options: subsidies, guarantees, etc.

[Determinant]

Should ordinary people get bank loans on easier terms in those circumstances too? Preferential treatment of banks creates hostility between population and bankers.

At the height of the crisis, some bankers in the US balked at the terms and didn't want to re-open (extravagant) compensation contracts, some resorted to Fifth Amendment arguments. To which I thought: when you come begging before the throne, it behooves you to be humble. (Extravagant) Bonuses have no place when the taxpayer is saving your firm from bankruptcy and you and your fellow employees from unemployment.

[Cornelius]

Of course all banks borrow more than the value of the bank, but we're talking about taxpayer money here. If a bank is receiving such money, then there need to be strict terms of borrowing - which there have not been, historically. We live in a democracy, and people deserve to have full disclosure over how public money is being used. Political economy matters much more than some economists think.

[Nick Rowe]

Cornelius: "If a bank is receiving such [taxpayer] money, then there need to be strict terms of borrowing - which there have not been, historically."

If we are talking about repoing things like CMHC-backed mortgages, which are already guaranteed by the Federal Government, as in this particular case, that isn't really an issue. But yes, if we are talking about genuine LOLR actions, it is an issue. Unfortunately, I find it hard to reconcile the sort of LOLR actions that might sometimes be needed with strict rules of borrowing. Valuing the sort of assets that might serve as colateral might in practice be a judgement call, that needs to be done quickly and sometimes secretly. The LOLR role is a nasty one. Maybe there's a better way.

[K]

Nick: "The LOLR role is a nasty one. Maybe there's a better way."

There's always 100% reserves, with loans backed by 100% capital. As Mankiw once pointed out, there is just no reason to believe that there are any associated efficiency costs. And, of course, lots of other even better solutions as we've discussed here many times.

Did you get a chance to consider my really excellent reply to you in the Short vs Long Run Natural Rate of Interest post? :-)

[Declan]

"I'm bemused that the 'progressive' position appears to be to just let banks fail; liquidate them in a crisis. That seems more like what the internet-trained Austrian economists advocate."

I think the point is more to make people recognize that much of the shareholder value in a bank is contingent upon government support (but don't call it a bailout!) in tough times, so when bankers try to argue (not that they do much in Canada, they tend to keep a low profile) that they are entitled to large profits and big bonuses because they earned them in a free market and sailed through the recession without needing a bailout, people realize that the truth is a bit more complex and that commercial banks are quasi-public bodies for the reasons that Nick outlines in his post.

For the most part, impending meltdown in the mortgage market notwithstanding, I think we manage this fairly well in Canada - the government tolerates a cozy oligopolistic banking sector and in return the banks pay their taxes, keep things steady and don't go overboard competing or bonus paying or 'innovating' themselves into bankruptcy. Given the tenor of the times (despite the recent meltdown, still best exemplified by the incident where regulators staged a photo-op of themselves taking a chainsaw to a stack of financial regulations) it probably takes a little pushing from the left to keep this system in place, so from an end-justifies-the-means perspective, the CCPA report is probably helpful, regardless of the exaggerated nature of some of the claims made.

[Stephen Gordon]

much of the shareholder value in a bank is contingent upon government support

No, you're missing the point. Shareholder equity was not really in danger; those mortgage assets were safe. The scenario that the IMPP was trying to avoid was that the banks would simply stop lending and be forced to start calling in existing loans.

Here's how I put it on twitter:

Scenario 1: You go to the ATM to withdraw $1000 and are told you have insufficient funds. A friend gives you $1000.

Scenario 2: You go to the ATM to withdraw $1000 and find that although you have plenty of money in your account, there's a technical malfunction and you can't get access to it. Your friend gives you $1000 cash in exchange for a cheque for $1100. Your friend deposits the cheque without incident (you aren't insolvent) and you use the $1000 to keep body and soul together, although you may be grumbling about how your friend was taking advantage of your misfortune. When the ATM is finally repaired, you go back to using it.

Scenario 1 played out in the US; *that* was a bailout. We had Scenario 2. You can call it a bailout if you want, but as far as I'm concerned, doing so is a willful misrepresentation of what happened.

[HJC]

Stephen, you're right, shareholder equity was not in danger, but in reality it's not easy for banks to "stop lending" or "to start calling in existing loans" - pre-committed, irrevocable lines of credit can be drawn at the discretion of the borrower. Banks would be more worried about the inter-bank market, this is where the CBs need to provide funds otherwise it would freeze up. To call it a bail-out is not correct. It sounds like the Canadian banks were not insolvent, but had liquidity problems instead.

[Nick Rowe]

Declan: If that is how this is perceived, then in the interest of transparency the Bank of Canada should clearly state:

"Look, we really really want you to accept this loan, whether you need it or not, and increase your own lending in turn. But we must warn you that this loan comes with strings attached. If you accept it then the Canadian Centre for Policy Alternatives will call it a bailout, and say that you are in hot water and deeply underwater, and most journalists will pass on that accusation without comment. Furthermore, we will consider you a quasi-public body from now on, and so the government will want a say in how much profit you earn and what bonuses you pay. So think very carefully before accepting this loan."

[HJC]

Nick: As we have seen recently there are plenty of people who have a fundamental misunderstanding of how banks work. Plus, grumbling about how banks operate in a constant state of insolvency since they cannot pay bank their debts on demand, so only exist due an implicit government backing. But governments regulate capital requirements and control benchmark interest rates so have a lot to say about bank profitability. Finally, since banks are licensed credit-money issuers perhaps it is right to consider them quasi-public bodies!

[Nick Rowe]

Actually, the more I think about it, the more pissed off I get. This wasn't some blog post, banged out at midnight, where stuff happens. They must have done a lot of work getting all that data. And it was reviewed by some very good economists, who cannot of course be held responsible for the content, but who must have told them that the "analysis" simply wasn't there to justify the interpretation, and that banks are banks. And they put a lot of work into the presentation, and publicising the Report, and doing interviews.

This was a very carefully planned hatchet job. The thought that should have gone into analysis and interpretation went instead into political spin.

[Nick Rowe]

HJC: Banks historically have (sometimes) existed without any LOLR. Whether that is a good or bad thing is another question. It is not obvious to me that the fact that banks' liabilities are used as media of exchange makes them automatically quasi-public bodies. (Some say that government should get out of the money-business altogether). Yes, governments regulate banks, but the argument "we regulate you, therefore we have a right to have a say in everything you do" seems to be weak.

But if the government offers you a no-strings loan, that it wants you to accept, and you accept it, and then the government afterwards turns around and says: "Aha! You took our money, and that means you belong to us and we can do anything we like!" is a bad argument, both on deontological and consequentialist grounds.

I am suddenly reminded of my late colleague TK Rymes. TK used to argue that Carleton should refuse to accept any government money, because it would inevitably lead to full government control of what we taught and how we taught it and who taught it.

"You want a business loan, or mortgage? Are you in the government's good books?"

[Nick Rowe]

HJC @6:42am. That is (the beginning of) exactly the sort of analysis that should have been done in this Report, but wasn't. The distinction between liquidity and solvency, and an attempt to assess liquidity of existing assets and liabilities. Were banks calling in good loans and lines of credit at an unusually high rate, or dumping lots of assets at firesale prices, etc.?

[Nick Rowe]

K: I had a read through that Mankiw link on 100% reserves. I think that what he misses is the principal-agent problem. One of the explicit assumptions of the Mogliani-Miller theorem, IIRC, is no agency problems. Banks make loans whose quality is costly to verify. So the bank itself becomes the residual claimant so it has the full incentive to monitor those loans properly, and the rest of us who hold deposits at the bank don't have to go over their loans and check their quality. But of course, this only works up to a certain point, because if enough loans are bad enough then the depositors are on the hook too. The MM theorem says the debt/equity ratio is indeterminate. If we add agency problems we can get an interior solution.

Sorry. I couldn't think of anything useful to say on your Short vs Long comment. Maybe it will slowly percolate in the back of my mind, and reappear in a later post ;-)

[HJC]

Banks are licensed to create credit-money and their liabilities are now a (the most?) common form of payment. Their lending increases the broad money supply. Perhaps not technically quasi-public but the government is only other body that has this power.

I suspect that the the periods without LOLR correspond with a high incidence of bank runs.

As I said before, it's not easy to call in good loans (or bad!) most are committed to be used at the discretion of the borrower. It is more likely that lines of credit are being drawn at a high rate. During a crisis is precisely when bank assets are most illiquid and unpredictable. This is when Bagehot's famous dictum is most relevant.

It's true that regulators haven't used their powers too much recently but they are flexing their muscles now and they are having a lot of say in how things are and will be done!

[Patrick]

Based on conversations with friends & family, I'd say many people don't understand that illiquidity can be infectious among banks precisely because they are banks; they don't have giant piles of cash sitting around. They all rely on each other, and if the system takes a bit enough hit it can fall over unless there is a LOLR (and if it falls over unemployment can and does increase very very rapidly).

FWIW, I suspect things like the CCPA report get traction in the media because people tend to try to understand financial 'news' by analogy, and in most people's lives there's no meaningful difference between insolvent and illiquid, so when they read stuff like the CCPA report they predictably get all hot and bothered. Most of us don't live in reality where we have a big stack of mortgages and no cash.

[Bob Smith]

Cornelius: Of course all banks borrow more than the value of the bank, but we're talking about taxpayer money here. If a bank is receiving such money, then there need to be strict terms of borrowing - which there have not been, historically.

Just to be clear, the big chunk of the program didn't entail banks borrowing public money. The IMPP involved the banks SELLING their government guaranteed assets (insured mortgages) to the government for cash.

And for what it's worth, the Montreal Gazette's cites a spokesman for Jim Flaherty to the effect that by 2015, the government will have netted %2.5 billion dollars from the program - hey, 10% of the way to paying for new fighter jets.

[Nick Rowe]

Bob: was it a simple sale, or was it a repo?

[Stephen Gordon]

My understanding is that it was a sale: insured mortages for CMHC paper. And since CMHC paper trades at a discount form T-bills (even though they are also guaranteed by the govt), that meant the banks lost money on the exchange. (That's why my scenario 2 above involved writing a cheque for $1100 to get $1000 cash.)

[Joseph Laliberté]

Bob Smith:
If the government paid the market price for these MBS in the context of the IMPP (as I supposed he did), would you say that the government's action provided "market support" for these instruments so to avoid a debt-deflation spiral? If you respond yes, would you reconsider your position on "providing the banks with liquidity without bailing them out"?

[JP Koning]

Bob, the government may have netted $2.5 billion dollars from the program. But that's not the right way to think about it. Perhaps the government mispriced the liquidity services it was offering the banking system and it should have netted $2.6 billion. Maybe it should have made $10 billion. Maybe just $1.0 billion. We'll never know if it properly priced the program.

[ianlee]

I agree completely with Stephen's comments and his judgment that "it was a willful misrepresentation to characterize the liquidity support as a bailout".

In further support, 70% of Cdn bank needs are provided by depositors - the safest, most reliable and cheapest form of finance for a bank (i.e. for CDN banks). And there were no runs on Cdn banks.

AND, Cdn banks have always had much low leverage ratios (required by OSFI - and Cdn banks exceed the OSFI standard) - much lower leverage than US banks and much much lower leverage ratios than the European banks that seem to operate as fonctionnnaires of EU countries' finance ministries. Cdn banks mostly exceeded the standards of Basel 3 - before Basel 3 existed or was agreed to.

(when I was in the bank in the 1970s and 80s, we characterized ourselves as "prudent balance sheet lenders" while we characterized the American banks as "go-go bankers" for they would lend against cash flow projections (i.e. fairy dust). We never discussed European banks but if we had, would likely have called them Gosbanks - USSR).

Only 30% of Cdn bank needs are borrowed by Cdn banks from short term commercial markets (CBA data and BoC data). And as Carney and others noted tirelessly, it was the commercial money markets that "froze up". BTW, the credit crunch hit and hurt the shadow banks far far more than the Cdn banks - e.g. leasing stopped almost overnight and some firms exited from Canada back to US.

The GoC and BOC liquidity support was transparently designed to end the credit crunch to get Cdn banks lending again to SMEs which had almost stopped cold - and to announce to the capital, credit, deposit markets to: "calm down, everything is OK, the Bank of Canada and the Govt of Canada are standing behind the Cdn financial system. There is not and will not be a financial meltdown - now - go watch the Maple Leafs lose to whomever" (sometimes a cigar is simply a cigar).

I agree with Nick that this was a pure hatchet job. Good empirical research but interpretations that are simply not sustainable - presumably in pursuit of a larger agenda e.g. "tax the rich greedy banks".

[Nick Rowe]

Stephen: OK, and a sale is even simpler than a repo, or loan.

JP: If a big buyer makes a big purchase, that will move the market price. But we normally say that exploiting that monopsony power to the full, and buying less than you wanted to because you want to keep the price below the competitive equilibrium, is not a good thing. Not sure how well that analysis applies here.

[JP Koning]

Nick: "I'm not sure whether or not that could work, in a system-wide liquidity crisis, if it wasn't backed by the ultimate creator of money."

The CMHC/IMPP plugged much of the Canadian side of the crisis in 2008 by buying NHA-MBS, and the CMHC/IMPP aren't ultimate creators of money but taxpayer funded.

A counterfactual is a world in which NHA-MBS liquidity options are sold by private actors to holders of NHA-MBS. These options allow NHA-MBS holders to sell all MBS back to the option writer at any time at a liquidity-protected price (some favourable point in the bid-ask spread). In a liquidity crisis bid-ask spreads increase, so the value of these options would quickly rise. The CMHC/IMPP provided MBS holders with a liquidity option, but we'll never know if they required MBS holders to pay the market price for this option.

[Stephen Gordon]

Yeah, the goal here wasn't to maximise govt revenues, it was to provide an incentive to the banks to stop using govt-provided liquidity when markets started functioning again. As it was, only $69b of the $125b made available was taken up. If really were free money, they would have snapped it all up.

[Bob Smith]

Bob, the government may have netted $2.5 billion dollars from the program. But that's not the right way to think about it.

"Perhaps the government mispriced the liquidity services it was offering the banking system and it should have netted $2.6 billion. Maybe it should have made $10 billion. Maybe just $1.0 billion. We'll never know if it properly priced the program."

True, although that's a criticism of anything the government does - how do we know that civil servants wages are properly priced? It's particularly problematic in this context, because there is no market competitor for the liquidity services that can be offered by the government - private parties can't print money.

In any event, the terms weren't so favourable that the banks were hopping to make make use of the IMPP, since the government only ended up buying $69 billion of the $125 billion it had offered to puchased. Not conclusive, I admit, but sugestive that the liquidity services weren't significantly underpriced. An of course, if the CCPA wants to argue that there was a bailout because the the liquidity services were underprised, that might be a reasonable argument, but it'll require a lot more evidence that they've accumulated.

[Bob Smith]

"Would you say that the government's action provided "market support" for these instruments so to avoid a debt-deflation spiral"?

You'll have to clarify, what do you mean by "market support".

[ianlee]

Again, agree with Stephen.

The banks were never in jeopardy of failing themselves - they simply stopped lending in order to hoard cash when the crunch hit.

(The small and mid-sized firms account for 75% of employment in Canada).

This caused Catherine Swift (MA, Economics, Carleton) Pres of CFIB and the SMEs put unbelievable pressure on Flaherty to “do something” - to get the banks lending again.
To restate the obvious, the credit crunch was passed on by the Cdn banks to SMEs (and was also passed on to shadow banks who had commercial lines of credit with banks in addition to using the commercial credit markets).

The SMEs responded by shutting down their supply chains i.e. stopped ordering stuff from the mfgs. And the SMEs threatened to start laying off large numbers of employees – a disaster for the Finance Minister if that were to pass.

The shadow banks responded by inter alia, stopping leasing of cars, trucks, industrial equipment and credit to SMEs as well.

These actions would have caused huge increases in unemployment – and would have caused the CAW, CCPA et al to hurl abuse at the GoC for not intervening. And it would have been a disaster for Flaherty.

This is why the CCPA interpretation is “nonsense on stilts”.

[Joseph Laliberté]

Bob Smith:
What I mean by "market support" is, let say, by running the counterfactual simulation of no purchase of MBS whatsoever by CHMC, one find that the price of MBS would have been actually much lower. Would you then still be comfortable saying that IMPP was just about providing liquidity to banks and that it had no "bail out" component?

[Bob Smith]

"one find that the price of MBS would have been actually much lower"

Would it have been? Certainly the offering price for such securities, from financial institutions facing liquidity contraints of their own, would have been lower, but would financial institutions have been willing to sell at those offering prices? Not so clear. They might just have preferred to hoard their cash and not lend, rather than selling safe assets at a hefty discount to free up cash. In that scenario it would not be true to say that the price of MBS would have been lower (or not), there just wouldn't have been a price. In fact, that's precisely the scenario that the government was trying to avoid.

You could describe the government's role here as a market maker, but it's hard to conclude (at least without a lot more evidence) that they were providing "market support".

[dwb]

i was remiss.
one cannot mention banks without mentioning the balance shet mechanism that makes recessions worse.

which means a big mechanism torward preventing bailouts is... ngdp targeting.

every post should have a tie in. lol. my bad.

[Nick Rowe]

dwb: "which means a big mechanism torward preventing bailouts is... ngdp targeting."

Bingo!

Looks like Evans at the Fed is now on board too, so we are slowly winning!

[Joseph Laliberté]

Bob Smith:
first, banks do not have to "free up" cash to lend. They might have to free up capital if they are capital constraint, but this is entirely different to freeing up cash.

second, CMHC auction results that yields on MBS were subtantially higher than the risk free rate when the IMPP got started in late 2008, and this yields differential progressively came down thereafter. This suggest that IMPP was quite sucessfull in brining yields on MBS closer to the risk free yield. If anything, this suggest that there was indeed a "bail out" component to the IMPP to the extent that it proped up MBS price, and improved banks' capital position.

[JP Koning]

"In any event, the terms weren't so favourable that the banks were hopping to make make use of the IMPP, since the government only ended up buying $69 billion of the $125 billion it had offered to puchased. Not conclusive, I admit, but sugestive that the liquidity services weren't significantly underpriced."

One could just as easily say that the terms were very favourable since the banks ended up selling $69 billion to the IMPP rather than $0. I'll stay mum on the issue.

"...because there is no market competitor for the liquidity services that can be offered by the government - private parties can't print money."

Private clearinghouses have been money printers. But you don't need to be a "printer" to provide liquidity services. Commercial banks can create deposits with a mouse-click, these deposits serving as back up lines of credit to smaller banks that are liquidity providing. A formal liquidity options market doesn't exist but it might at some point in time. The more that liquidity services are priced in competitive markets the better.

[Simon van Norden]

So the CCPA analysis that got Nick so upset has been called "nonsense on stilts" and "a hatchet job" and generally trashed so hard that the contrarian in me is trying to find something in it to like.

So let me ask, is there any reader of this blog that thinks the govt. and the BoC have done a good job in accounting for the costs (and profits!) of their extraordinary interventions of 2008 and 2009? Do you think that we've had a credible accounting by program? Do you think that such a public accounting is part of the BoC's mandate? Do you think the precedent of clear accounting helps maintain the central bank's and govt.'s credibility and popular support in times of crisis?

Personally, I'd stop well short of advocating the institution-by-institution breakdown that the CCPA calls for (for reasons others have mentioned above.) But I also can't shake the feeling that a transparent accounting could have headed off this media "controversy."

[Nick Rowe]

Simon: good question. But I think the answer - what that accounting would look like -- might be really weird.

For any normal business, the accounting might be hard to do in practice (if you are dealing with assets of uncertain value) but would be possible in principle. And the answers to that accounting (did the firm benefit? was there a cost to the government?) could tell you the amount of subsidy.

Suppose the Bank of Canada really screwed up, and engineered a recession and financial crisis. And then took absolutely minimal actions to keep the banks just afloat, by overpaying for junk assets. That would meet the accounting definition of a bailout.

[Bob Smith]

"CMHC auction results that yields on MBS were subtantially higher than the risk free rate when the IMPP got started in late 2008, and this yields differential progressively came down thereafter. This suggest that IMPP was quite sucessfull in brining yields on MBS closer to the risk free yield. If anything, this suggest that there was indeed a "bail out" component to the IMPP to the extent that it proped up MBS price, and improved banks' capital position."

Let's deconstruct that. "The yields on MBS were substantially higher than the risk free rate". An odd result, don't you think, since the MBS in question were risk free - i.e., they were insured mortgages backed by the full faith and credit of the government of Canada. The banks who used to IMPP weren't dumping bad assets on the government, they were paying a premium for liquidity - as they should. It doesn't follow that there was a bailout component to the IMPP. Indeed, the fact that the spread betwween the yields on MBS tendered to the government under the IMPP and the risk-free rate came down as the program extended into 2009 suggests that, as liquidity concerns loosened, the banks were no longer willing to pay a material premium for liquidity.

"One could just as easily say that the terms were very favourable since the banks ended up selling $69 billion to the IMPP rather than $0. I'll stay mum on the issue."

Fair enough. Still, if you think of it in terms of the government making $2.5 billion dollars for, basically, shuffling its balance sheet (i.e., issuing debt to purchase government guaranteed debts), that's a pretty shrewd investment.

[Stephen Gordon]

I can't get over this quote:

"CMHC was not providing loans that needed to be paid back, as was the case with the other two aid programs. CMHC was buying mortgages and, as such, the banks did not need to pay this money back. The CMHC program was thus a straight cash infusion for Canada’s banks."

I just came from giving a straight cash infusion for one of Canada's grocery chains. Someone should inform the media.

[Bob Smith]

Yikes!

Nick, remember when you said: " it was reviewed by some very good economists"? I hope you're wrong. It sounds like it was prepared by the local OPIRG chapter. By an intern.

[Stephen Gordon]

I don't understand why demanding a larger haircut on MBS would have been a good idea. The goal of the program was to make sure banks could keep lending during a financial crisis, ideally in a way that didn't produce moral hazard problems. That's what the IMPP did. If anything, there's a stronger case for having a smaller haircut: why would we want banks to restrict lending because they couldn't get short-term liquidity?

[JP Koning]

"Fair enough. Still, if you think of it in terms of the government making $2.5 billion dollars for, basically, shuffling its balance sheet (i.e., issuing debt to purchase government guaranteed debts), that's a pretty shrewd investment."

I though we already agreed that $2.5 billion isn't a sign of shrewdness, since we don't know if the return the government should have earned providing liquidity was $1 billion, $2.5 billion, or $10 billion.

[Stephen Gordon]

What would be the criteria for making that decision?

[Bob Smith]

"I though we already agreed that $2.5 billion isn't a sign of shrewdness, since we don't know if the return the government should have earned providing liquidity was $1 billion, $2.5 billion, or $10 billion."

Making anything for rejuggling your balance sheet is a sign of shrewdness, whether it's $1billion, $2.5 billion or $10 billion.

[Stephen Gordon]

Well, guess what everyone? CBC radio's The Current will be talking about this tomorrow morning, and I've been drafted for the show. First segment, 8:30 local time. I promise to be as boring as possible, with heavy use of jargon and unexplained acronyms.

[Nick Rowe]

Well done Stephen. It's good you are doing this.

JP and Bob: Milton Friedman said (Optimum Quantity of Money) that the government should earn precisely $0 from providing liquidity, since the marginal cost of the government providing it was $0, and Marginal Cost pricing maximises total surplus. (That's the 0% nominal interest rate solution, with negative inflation equal to the natural rate of interest.) I would have to think a bit before translating that into a world of a 2% inflation target.

Just shows that the accounting goes all weird once you get into this sort of monetary question.

[Joseph Laliberté]

Bob Smith:
Ever held an illiquid asset? Ever refrained from selling it because you are afraid that putting it on the market will be enough to make the clearing price collapse? Would you not find nice in such situation if the government would give you a helping hand by offering to purchase your asset at a given maximum price through an auction process? Would you not find it nice also if the government progressively increase its maximum price offer in each passing weekly auction (so that minimum rate offered decrease from 3.9% to 3.4% in a matter of three weeks)?

[Kevin Milligan]

Hi Joseph,

you need to make your illiquid asset 100% govt guaranteed for your example to be analogous.

[Stephen Gordon]

Also, why is it a good idea for the govt to be extracting rents from banks during financial crises?

[Bob Smith]

NR: "Milton Friedman said (Optimum Quantity of Money) that the government should earn precisely $0 from providing liquidity, since the marginal cost of the government providing it was $0, and Marginal Cost pricing maximises total surplus"

That was my intuition, but it's nice to know that there's some basis for it.

SG: "I promise to be as boring as possible"

You'll fit in perfectly on the Current.

JL: "Would you not find it nice also if the government progressively increase its maximum price offer in each passing weekly auction (so that minimum rate offered decrease from 3.9% to 3.4% in a matter of three weeks".

Keep in mind that the offered rate reflected the government's underlying cost of borrowing (plus a spread). Since the MBS are government guaranteed bonds, it would make sense for their price to track the price of direct government bonds.

[Joseph Laliberté]

Kevin Mulligan:
Got your point. However, my broader point is that banks do not need liquidity to lend; they need capital. The paradox in this whole discussion is that the guys arguing that IMPP helped banks extend loans are also the ones arguing that IMPP only helped banks on the liquidity front (not on the capital front).

[JP Koning]

"What would be the criteria for making that decision?"

Dunno. I'm just taking an agnostic stance between the CCPA's story of a banking subsidy/tax-payer penalty and its mirror image (Bob Smith's point) that the government's canniness led to a tax-payer subsidy/banking penalty. Both arguments are wrong because without a market price to use for comparison, it's impossible to know if the "liquidity services" provided were subsidized or not. We'd only know if we could compare the price the government set for the IMPP liquidity-option to a competitively-priced liquidity-option on NHA-MBS, but those options don't exist. Unhelpful, I know.

[Bob Smith]

"However, my broader point is that banks do not need liquidity to lend; they need capital."

It's not so cut and dry, you need both. A bank facing liquidity concerns isn't going to lend, or isn't going to lend as much. It'll hoard cash. That was the concern in 2008.

[Bob Smith]

"Both arguments are wrong because without a market price to use for comparison"

Don't we have a market price for the typed of government guaranteed debt being purchased by the government? We have government debt being issued by the government on the market (subject to the caveat, I suppose, that the BoC is a player in that market). And the price being offered by the government here was linked to its underlying financing cost (I.e., the price of government debt). There's a spread - which I suppose contradicts Friedman's prescription, but which hardly constitutes a bail out.

[Stephen Gordon]

Does anyone know if the identities of purchasers of T-bills at the Bank of Canada's auctions are made public, along with the quantities purchased? That information doesn't appear on the Bank's site:

http://www.bankofcanada.ca/markets/government-securities-auctions/calls-for-tenders-and-results/

[Ritwik]

Nick

Friedman was forced to conclude that because he didn't envision a world with non-zero interest on reserves. With interest on reserves, free liquidity only means that the IoR = interbank rate = discount window rate. It does not translate into interbank rate = 0.

Also, the whole price of liquidity = marginal cost = 0 is a classic partial equilibrium with no strategic benefits or agency problems considered. In the Perry Mehrling worldview (the 'Money view') it is the central bank giving up the discipline part of its function, trying to make things 'perfectly elastic'. The Friedman rule is by default inflationary for asset prices, at all levels of IoR.

[Nick Rowe]

Ritwik: Fair points. Or, Friedman assumed 0% interest on currency. And yes, if the Friedman rule were followed, it would mean that the central bank would take over (pretty much) the whole economy, and buy up all the assets. (It's a commie plot!) Because if you can get the same return on holding cash as you could on any other asset, you would rather hold cash.

[Ritwik]

Nick : Interesting. But wouldn't that happen only in an economy with zero risk appetite and a central bank that accepts all forms of collateral - a weird Friedman-Bagehot mongrel central bank(incidentally a reasonably accurate descriptor of the pre crisis Greenspan-Bernanke Fed!)

Friedman rule only makes the price of liquidity zero, not the price of default or the price of the term structure. And when risk appetite does crash to zero, the central bank of course has the option of making the price of liquidity non-zero (if it is indeed accepting all forms of collateral).

Though, Mehrling's point is also that a roughly Friedman rule following central bank (which is incidentally also a very Tobin-esque central bank)encourages a downward trend in default premia as well and is over-sensitive to declines in asset prices. All this is of course tangential to the original Friedman rule, which is only supposed to apply to a central bank that does not accept default risk or term structure risk of any kind on its balance sheet.

A Bagehotian central bank (liquidity manager) is conceptually very different from a Friedmanite central bank (macroeconomic manager) and in the money view, one of the fundamental reasons of the fuelling of a financial crisis (or the lack of prevention of a financial crisis)is the mongrel hybrids that developed economies have, esp. in the US.

[Nick Rowe]

Ritwik. My old thoughts here.

[Ritwik]

Ah, Nick. But the trick in that post was not deflation, which was a veil. The trick in that post was increasing the real rate of interest on money while implicitly assuming that the real rate of interest on other assets would remain 'constant' over the long run, in the extreme case assuming that it was greater than the real rate on all assets in the economy. It was to take away the 'moneyness' of money. Obviously you ended up with freaky results when the most liquid and most short term asset was assumed to be yielding the highest premium.

If this kind of a scenario were to ever happen, you would have to drop the assumption of constant convertibility. You cannot assume high deflation and zero interest bearing money at the system. That's just assuming the conclusion. Your thought experiment was overdetermined.

[dwb]

banks are banks but canadian banks are strong:

http://www.bloomberg.com/news/2012-05-02/canadians-dominate-world-s-10-strongest-banks.html

[Ritwik]

The Friedman rule takes the desired rate of interest as a given and then wants inflation to be the negative of that. In the Friedman rule framework of your thought experiment, deflation would merely the sign of a very high risk free rate of interest. The zero interest on money would put a floor on the interest rate on all other assets and the size of the central bank would be indeterminate or a matter of other policy goals.

[Steve Roth]

Hey Nick if you're casting about for something to write about, I'd love to hear your thoughts on this:

http://www.concertedaction.com/2012/04/12/kaldors-reflux-mechanism/

[Peter N]

Lavoie is interesting, and the circuit matrix approach has been around for a while, but the model seems to assume what it seeks to prove. There are only firms and consumers. Banks are just a link. Everything borrowed is spent in production. All proceeds from production go to consumers as deposits. Consumers keep some of this as residual deposits and spend some on consumption goods, which closes all the circuits.

This matrix model is exactly the type of model Keen uses to reach quite different conclusions, so these conclusions clearly aren't inherent in the use of a matrix model (hardly surprising, the matrix is just a tabular representation of a set of equations. Use different entries, and you get a different model). Lavoie says:

"The fact that these mechanisms appear to be totally independent has led some authors to claim that there could be a discrepancy between the amount of loans supplied by banks to firms and the amount of bank deposits demanded by households. This view of the money creation process is however erroneous. It omits the fact that while the credit supply process and the money-holding process are apparently independent, they actually are not, due to the constraints of coherent macroeconomic accounting. In other words, the decision by households to hold on to more or less money balances has an equivalent compensatory impact on the loans that remain outstanding."

This doesn't in any way prove that the model is a model of the real economy, and I'm afraid an appeal to "the constraints of coherent macroeconomic accounting" doesn't persuade me. In particular money flows through production, whereas in the real economy a great deal of activity has nothing to do with production. Much of it goes into asset transfers.

Banks create money and use it to purchase claims on collateral - mortgages. These claims can be securitized and sold to homeowners, who may pay by using as collateral the increased value that mortgage enabled purchases have created. This may just look a bit familiar. It doesn't however appear to be representable in the model.

All the complications of the relationship between credit and money are ignored, yet the supply of broad money is vastly larger than that of base money.

BTW it looks like here as in other places the term GDP is used to mean GDP after I've adjusted it for my model (GDP has something like $3 trillion of imputed production and ignores at least as much other economic activity. There's plenty of motive for adjustment). This is quite common and has caused a lot of confusion.

It'll be interesting to here what other people here have to say about this.