The Bank of Canada is important precisely because it’s not a bank

Nick Rowe has a new post that explains what’s special about central “banks:”

What makes a central bank special?

The Bank of Canada can borrow and lend. So can the Bank of Montreal. So can I. Nothing special there.

The Bank of Canada can set any rate of interest it likes when it lends. So can the Bank of Montreal. So can I. Nothing special there. “If you want to borrow from me, you have to pay x% interest.” We can all say that. (Whether anyone will want to borrow from us at x% interest is another question.)

The Bank of Canada can set any rate of interest it likes when it borrows. So can the Bank of Montreal. So can I. Nothing special there. “If you want to lend to me, you have to accept y% interest.” We can all say that. (Whether anyone will want to lend to us at y% interest is another question.)

And yet there’s this utterly bizarre belief among many economists that it is the Bank of Canada that has the power to set Canadian interest rates, just by borrowing and lending. And that the Bank of Montreal, and ordinary people like me, somehow lack this special power. Even though we can borrow and lend too.

Now it’s true that the Bank of Canada is a lot richer than me. But what if I were a Canadian Bill Gates or Warren Buffet? And is the Bank of Canada richer than the Bank of Montreal? That can’t be the difference.

Now it’s true that the Bank of Canada can create money at the stroke of a pen, and I can’t. But the Bank of Montreal can. What makes the Bank of Canada special, compared to the other banks? What power does the Bank of Canada have that the Bank of Montreal lacks?

I’m going to give the same answer I gave nearly three years ago. And then I’m going to expand on it.

The fundamental difference between the Bank of Canada and the Bank of Montreal is asymmetric redeemability. The Bank of Montreal promises to redeem its monetary liabilities in Bank of Canada monetary liabilities. The Bank of Canada does not promise to redeem its monetary liabilities in Bank of Montreal monetary liabilities.

And then he ends the post as follows:

There is something very seriously wrong with any approach to monetary theory which says we can assume central banks set interest rates and ignore currency. It is precisely those irredeemable monetary liabilities of the central bank (whether they take the physical form of paper, coin, electrons, does not matter) that give central banks their special power. That’s what makes central banks central.

Now let’s consider a different monetary system.  Imagine a gold standard and a monopoly producer of gold.  The gold mine company would reduce short term interest rates by increasing the supply of gold.  Like currency, gold is irredeemable.  But no one would call this gold mining company a “bank” because it possesses no bank-like qualities.  Banks don’t create irredeemable assets, gold mines do.

Central banks may have some bank-like qualities, but what makes them special is their ability to produce currency–i.e. paper gold.  And that has no relationship to banking at all.

For years I’ve dealt with commenters who wanted to turn the discussion to banking:

“How do you know negative IOR will increase lending?”  I don’t care if it does, because banking has nothing to do with monetary policy.

“The Fed can’t cut rates any lower–how are they supposed to boost the economy?”  It doesn’t matter whether they can cut rates, because rates aren’t the transmission mechanism.

The Fed affects NGDP by changing the current supply and demand for the medium of account, and also the expected future path of the supply and demand.

“Monetary policy is already quite easy.”  No; credit is easy, monetary policy is ultra-tight.

In the comment section Nick says:

I just remembered. Back on the I=S post, IIRC, some people were complaining I didn’t talk about banks enough. OK, here’s a post on banks.

I’d say it’s a post on why central banks aren’t banks.


Tags:

 
 
 

37 Responses to “The Bank of Canada is important precisely because it’s not a bank”

  1. Gravatar of dilletaunted dilletaunted
    11. September 2011 at 17:24

    “I don’t care if it does, because banking has nothing to do with monetary policy.”

    why not just learn about banks, scott? just give it 30 min a day

  2. Gravatar of W. Peden W. Peden
    11. September 2011 at 17:37

    “because banking has nothing to do with monetary policy”

    I’m not sure if I agree (it depends how you cash this sentence out). For example, take old monetary policy tools like lending ceilings, special deposits, supplementary deposits, reserve requirements and minimum lending standards, as well as the still-current capital requirements. In what sense can these be understood without a fairly intimate understanding of banking?

    (Interest rate policies obviously require a much less intimate understanding of banking.)

    Now, there are some monetary policy tools that work in a bankless economy (asset purchases and overfunding/underfunding; some would argue that overfunding/underfunding doesn’t constitute monetary policy, though when it was practiced it was regarded as monetary policy by its practitioners and everyone else at the time) but of course we don’t live in such an economy.

  3. Gravatar of Joe Joe
    11. September 2011 at 17:50

    You said…. “How do you know negative IOR will increase lending?” I don’t care if it does, because banking has nothing to do with monetary policy.”

    I don’t understand. Then what exactly is the purpose of a reduced IOR? A reduced IOR will of course reduce demand for reserve money by banks. But banks will do this by… lending out the reserves and create demand deposits or simply spending them on stuff.

    “Central banks may have some bank-like qualities, but what makes them special is their ability to produce currency-i.e. paper gold. And that has no relationship to banking at all.”

    But banks DO do that. They’re called Checking and Savings accounts. In fact, were it not for regulations, commercial banks would also make paper currency.

  4. Gravatar of Nick Rowe Nick Rowe
    11. September 2011 at 18:01

    Thanks Scott!

    Another way of looking at it:

    A regular bank has assets on one side of its balance sheet, and liabilities on the other. The central bank has assets on one side of its balance sheet and irredeemable “liabilities” on the other.

    It is a serious distortion of language to call an irredeemable liability a “liability” at all. It’s not a promise to pay anything.

    A central bank that torched all its assets could still be a functioning central bank. True, it couldn’t do an open market sale any more, but it could still decide on any positive amount of money creation.

    A regular bank that torched all its assets would be an ex-bank.

    Joe: “But banks DO do that. They’re called Checking and Savings accounts. In fact, were it not for regulations, commercial banks would also make paper currency.”

    Whether banks create money in the form of paper or electrons doesn’t matter. Whether that bank currency would be *redeemable* in central bank currency is what matters.

  5. Gravatar of W. Peden W. Peden
    11. September 2011 at 18:05

    Joe,

    “But banks will do this by… lending out the reserves and create demand deposits or simply spending them on stuff.”

    Strictly speaking, they won’t lend out the reserves. What they will do is try to replace an asset that is hurting their profits (cash reserves) with one that earns a profit, like a corporate bond. However, they can only do this at the level of the entire financial system by buying new securities and thereby increasing the quantity of money.

    (The Fed could create a similar effect without dealing directly with banks at all, by buying securities from insurance companies, pension funds & other holders of securites, which would create a hot potato effect with the new cash and thereby bid up corporate equity, allowing for companies to invest by selling new corporate bonds to banks at the new market value. That would increase the broadly defined money supply.)

  6. Gravatar of OhMy OhMy
    11. September 2011 at 18:37

    “The Fed affects NGDP by changing the current supply and demand for the medium of account”

    Wrong. Again. The Fed does not change the supply of money. MMT explained it here 100x times.

  7. Gravatar of dilletaunted dilletaunted
    11. September 2011 at 18:47

    “Wrong. Again. The Fed does not change the supply of money. MMT explained it here 100x times.”

    take this from a fellow mmt’er, OhMy: please step your game up

  8. Gravatar of OhMy OhMy
    11. September 2011 at 19:01

    dilletaunted,

    The Fed doesn’t influence the (credit) money supply other than by changing rates. Disagree?

  9. Gravatar of Nick Rowe Nick Rowe
    11. September 2011 at 19:17

    dilletaunted’s credibility went up in my eyes. (And by extension, MMT’s too, if just a little bit).

    OhMy: first understand the distinction between: a supply curve (or supply function); quantity supplied; and actual quantity sold. Then the distinction between money and credit.

  10. Gravatar of OhMy OhMy
    11. September 2011 at 19:53

    Nick,

    Do *you* understand the distinction between money and credit? Most of the transactions today are done with credit money. And no, the Fed doesn’t “supply” it.

  11. Gravatar of David Pearson David Pearson
    11. September 2011 at 20:58

    W. Peden,

    The “hot potato” effect implies the private sector is left holding less portfolio risk than it desires. However, the Fed absorbs this risk. The Fed passes the risk onto the Treasury in the form of potentially lower remittances. Treasury passes it back to taxpayers and/or bondholders in the form of potentially higher taxes and/or higher Treasury issuance. Some private sector actors might react to Fed purchases by increasing duration, others (seeing the possibility that their taxes will rise along with s.t. interest rates) by reducing it. The net effect is zero, as the Fed does not “cancel” the private sector’s exposure to risk, only shuffles it around.

  12. Gravatar of David Pearson David Pearson
    11. September 2011 at 21:03

    Nick,

    As I asked on your blog, if irredeemable liabilities pay interest, couldn’t the Fed lose money by “torching” its assets? If it decided to issue yet more bank reserves, wouldn’t it just lose more money? When marginal demand for bank reserves is zero, how does the Fed control the mix between ER’s, RR’s and currency?

  13. Gravatar of David Pearson David Pearson
    11. September 2011 at 21:06

    The above assumes IOR>0. Of course, the Fed controls the IOR, so in theory you are right. As long as the Fed leaves the IOR at zero, it can never go bankrupt. Can it do this and also control the price level?

  14. Gravatar of Jon Jon
    11. September 2011 at 22:07

    Nick writes:

    Whether banks create money in the form of paper or electrons doesn’t matter. Whether that bank currency would be *redeemable* in central bank currency is what matters.

    Suppose a credit card company got merchants to accept whatever payments were made, not an absolute claim on the ‘CB liability’ of the original transaction?

    Would that fundamentally change how the system works? It seems to me that the CC processor would not have liabilities redeemable in CB currency then and then there wouldn’t be a central currency.

    I wonder, does that make such an arrangement illegal given the injunction against private currency?

  15. Gravatar of Morgan Warstler Morgan Warstler
    11. September 2011 at 22:23

    What matters is that OTHER banks are kept from doing what the Fed does.

    David Pearson,

    When the Fed goes to Goldman and buys a bunch of T-Bills, the correct buyer (not the Fed), doesn’t get to buy those.

    Right?

    So ultimately everyone down the line is taking on more risk than they want to.

    The way to get rid of risk is to screw the sellers, wait around until hey are bankrupt, and then buy their stuff at such low prices you can’t possibly lose money on it.

    Interesting question: say BofA buys a pile of shit, and in the process gets a guarantee that NO OTHER bailouts for anyone will be forthcoming, not even IOR.

    At that point, does their risk exposure go up or down?

    I mean yes, prices on assets will fall, affecting their ability to liquidate / clean up Countrywide, BUT since everyone else will be dying, BofA gets to choose to pull the trigger, and play last man standing.

    Is it worth it to them? I mean they thought: A) countrywide was worth more B) they were stronger than others.

    MAYBE they were stronger than others UNTIL everybody else got continued Fed support.

  16. Gravatar of cato cato
    11. September 2011 at 23:24

    precisely. banks can care about interest rates. the monetary authority should care about the supply and demand for money.

    its sortof like the chinese cultural revolution, we’ve had eighty years of blindness worrying about interest rates rather than looking at supply and demand of money – duh!

  17. Gravatar of Martin Martin
    11. September 2011 at 23:27

    “Do *you* understand the distinction between money and credit? Most of the transactions today are done with credit money. And no, the Fed doesn’t “supply” it.”

    OhMy, as I understand it, credit is trust, the Fed supplies that trust. If there is too much credit relative to currency, credit contracts until the Fed injects currency. The amount of currency in circulation is a negative check on the amount of credit. What worth is a promise in dollars when there are no dollars?

  18. Gravatar of Martin Martin
    11. September 2011 at 23:37

    Nick could also promise you an antique chair to sit on for five minutes. When he only has 1 chair, 1 hour and 24 people, Nick will have to tell some people to stand.

    Unless Nick promises people something else – a stainless steel chair for example – the output of the antique chair factory will determine the number of people Nick can credibly promise to sit on a chair for five minutes within one hour.

    In the present day the government has forbidden stainless steel chairs, out of a safety concern because they’re afraid it might crack and people might find themselves sitting on the floor, so the antique chair factory has a monopoly.

    Even in the case that financial technology, let’s say I put a piece of wood across the chair that can hold two people, the amount of currency/antique chairs still determines how many people can sit for five minutes within the hour. You need less chairs, but you still need chairs.

  19. Gravatar of Wadolowski Wadolowski
    12. September 2011 at 02:49

    (it’s not a comment, rather astonishment)

    How is it, that “everybody” says about higher inflation target. Truely, almost everybody:

    Blanchard:
    http://www.imf.org/external/pubs/ft/spn/2010/spn1003.pdf

    Rogoff:
    http://www.project-syndicate.org/commentary/rogoff51/English
    http://www.project-syndicate.org/commentary/rogoff83/English

    Mankiw:
    http://www.nytimes.com/2009/04/19/business/economy/19view.html

    DeLong:
    http://www.economist.com/debate/days/view/697#pro_statement_anchor

    Stiglitz:
    http://www.mediatheque.lindau-nobel.org/#/Video?id=622

    even Diamond a little bit:
    http://www.washingtonpost.com/blogs/ezra-klein/post/peter-diamond-on-the-job-crisis-the-deficit-and-what-congress-and-the-fed-can-do/2011/07/11/gIQA3SQy8I_blog.html

    and Krugman all the time.

    Why the Fed is waiting? Who else have to support the idea: Barro, Lucas, Presscot…???

  20. Gravatar of bill woolsey bill woolsey
    12. September 2011 at 04:15

    Nick:

    Unless the central bank is targeting nothing, or the quantity of the monetary liabilities it issues, then it is useful to think of the monetary instruments it issues as liabilities.

    Think about a shift from a gold standard with redeemability to a gold-price targeting regime. They are the same.

    Shift to a multiple standard. Sure, it is hard to store the stuff, but the monetary instruments are liabilities. Then shift to a CPI target. It is the same.

    What about indirect redeemability? It is all about schemes for enforcing the rules on the monetary authority.

    Scott:

    What happens if the private banks issue the currency, and the central bank only issues deposit balances?

    What happens then? (I advocate that system, and I don’t think it makes the central bank any less of a monetary authority. Well, “think” is too weak. It is still a monetary authority.)

  21. Gravatar of W. Peden W. Peden
    12. September 2011 at 04:49

    David Pearson,

    Interesting. What does that imply for bond yields during QE?

  22. Gravatar of W. Peden W. Peden
    12. September 2011 at 04:52

    Wadolowski,

    One thing that puzzles me is that the Fed doesn’t even have a formal inflation target right now. There’s an implicit target of 2%, but that’s all.

    If they were to introduce a 3% inflation target and commit to it for (say) an eight year period, it wouldn’t technically be a “higher” inflation target at all- it would be the Fed’s first real inflation target.

  23. Gravatar of W. Peden W. Peden
    12. September 2011 at 04:53

    (Though I would prefer a wide range of alternatives- price level targeting, NGDP targeting, or even some sophisticated sort of aggregate targeting, before I supported an inflation targeting regime.)

  24. Gravatar of Scott Sumner Scott Sumner
    12. September 2011 at 05:09

    dilletaunted. I’m all ears–tell me why banks are important for monetary policy.

    W. Peden, I probably should have said banking has something to do with monetary policy in precisely the same way that drug dealing has something to do with monetary policy–they both are demanders of base money.

    Joe, Checking accounts are not irredeemable.

    Yes, banks would react to IOR by reducing their demand for base money. That was my point. But it makes no difference whether they reduce that demand by lending, or by buying assets.

    Nick, I’ve also argued that bank liabilities aren’t really liabilities. But be careful that Mike Sproul doesn’t hear you say that.

    OhMy, I define money as the monetary base. All my statements should be interpreted that way. The Fed controls the monetary base, and influences the broader aggregates.

    They need not change interest rates to affect the base, but changing the base will often (not always) affect rates.

    David; You said;

    “The above assumes IOR>0. Of course, the Fed controls the IOR, so in theory you are right. As long as the Fed leaves the IOR at zero, it can never go bankrupt. Can it do this and also control the price level?”

    IOR didn’t exist in the US until 2008, so it’s obviously not an integral part of monetary policy. It can be ended. The Fed cannot control the price level if the demand for money falls and it lacks assets to sell.

    Jon, I don’t understand your question, could you be more specific–with an example? If the merchants were paid in say bonds, they’d probably just cash in the bonds.

    Morgan, That’s right, they’d be arrested for counterfeiting if they tried.

    Martin, You said;

    “Do *you* understand the distinction between money and credit? Most of the transactions today are done with credit money. And no, the Fed doesn’t “supply” it.”

    Yes I do, and I agree with both assertions.

    Wadolowski—Good question.

    Bill, I agree that if the Fed is inflation targeting it makes sense to think of them having liabilities. It’s very similar to Social Security. Whether they have liabilities depends on whether you think of their vague promises to pay future benefits as liabilities.

    I agree that banks might issue currency redeemable into bank reserves. In that case only bank reserves would be irredeemable, and private currency would be redeemable. The monopoly producer of bank reserves would still be the monetary authority.

  25. Gravatar of W. Peden W. Peden
    12. September 2011 at 05:16

    Scott Sumner,

    “I probably should have said banking has something to do with monetary policy in precisely the same way that drug dealing has something to do with monetary policy-they both are demanders of base money.”

    In that sense I very much agree. The fundamentals of monetary economics- portfolio preferences and hot potatos- are the same for banks as they are for drug dealers.

  26. Gravatar of Martin Martin
    12. September 2011 at 05:17

    Scott, I was replying to ‘OhMy’ and the part you quoted is the OhMy’s post I replied to.

  27. Gravatar of David Pearson David Pearson
    12. September 2011 at 05:43

    W. Peden,

    I think the mistake is to think of private sector portfolios as holding “securities” instead of generic “claims”. If the duration risk of private sector claims is unaffected by QE, then so should long term interest rates. The term premium is merely the aggregate of compensation demanded for liquidity and duration risk held by the private sector. If you don’t affect duration risk, you don’t affect the term premium.

    Liquidity risk is a special case: the Fed can extinguish private sector liquidity risk through QE. In normal times, the liquidity premium is a small part of expected portfolio returns.

    Of course, you could build a thesis about how the private sector is unaware that the Fed shuffles duration risk back to it via the Treasury. Due to information costs, QE might be able to impact the term premium (temporarily, I would think).

  28. Gravatar of W. Peden W. Peden
    12. September 2011 at 06:14

    David Pearson,

    I don’t think people think of the risk effects of QE on the Treasury at all. What I wonder is why… Why do we end up with QE apparentely having an impact on asset prices that is more than temporary?

  29. Gravatar of David Pearson David Pearson
    12. September 2011 at 06:25

    “Why do we end up with QE apparentely having an impact on asset prices that is more than temporary?”

    Its not clear that we do. To the extent that QE does have that impact, it could be because of: 1) information costs of analyzing fiscal projections; 2) successful “signalling” effect of QE. The problem with the “signalling” concept is, if something theoretically does not affect private sector portfolio risk today, why promising to do it in the future influence behavior today?

    On the information cost question, I know bond analysts make projections of fiscal deficits under different interest rate scenarios. In fact, these projections are fairly easy to carry out. Even the rating agencies are smart enough to do them!

  30. Gravatar of Morgan Warstler Morgan Warstler
    12. September 2011 at 06:44

    Ya know, it is so damn funny that you guys can have a conversation about the Fed is not a real bank.

    And it gets discussed.

    But when I say the Fed and thus monetary policy itself, like government itself, is executed, not for everyone, but for people who OWN STUFF, CREATE STUFF, TRADE STUFF – people who matter.

    You are completely unable to rationally discuss it.

    My argument is to the Politics of money, what Scott / Nick’s argument is to the Agency of it.

    You can’t pretend there is an expertise to Monetary Theory, but not to Political Economy.

    Scott missed the Tea Party.

    IF DeKrugman KNEW the stimulus wasn’t going to be enough, he should have screamed DON’T DO IT!

    But he didn’t he thought, it wasn’t going to be enough, but hey we could do more later, so he supported it.

    The point here is the same one I make over and over and over, ALL POLICY must pass through the interests of hegemony and not some “upside later” thing – direct appeals to the immediate short term interest of hegemony.

    And being really inventive, truly INVENTING shit, means getting down into the rules as they are, and STILL solving for problems.

    If Scott for the get go, made pure appeals to the Tea Party, and attacked viciously DeKrugman, never finding common ground, questioning his motives at every turn…

    We’d be closer to the goal line.

  31. Gravatar of W. Peden W. Peden
    12. September 2011 at 06:59

    “if something theoretically does not affect private sector portfolio risk today, why promising to do it in the future influence behavior today?”

    Explain. Do you mean “why do present promises to do things in the future effect present actions”? Because that seems a very odd question.

    I think the evidence that QE does have a permanent impact on asset prices is as good as one could possibly expect given the amount of data we have-

    http://www.ecb.int/events/conferences/shared/pdf/macroecimp/Woods.pdf?2c1a6dea2966a4dd9cca169e38da09da

    http://www.moneyscience.com/pg/blog/Admin/read/87034/the-financial-market-impact-of-quantitative-easing-in-the-united-kingdom-bank-of-england

    http://www.rbs.com/investors/economic-insight/transcript/transcript-210110.ashx

    http://www.bankofengland.co.uk/publications/speeches/2009/speech405.pdf

    http://visinomics.com/?p=1387

    By-the-by, I’ve noted that many of my recent comments on QE (as well as everything I’ve ever read on QE written by a chartalist) has ignored the fact that it is primary dealers rather than traditional banks which are the main partners in QE. Not quite back to the drawing board, but it does make many of the complexities of banking irrelevant and makes me lose a lot interest in the “all the new money will get stuck in the financial system” arguments.

  32. Gravatar of Benjamin Cole Benjamin Cole
    12. September 2011 at 08:47

    Excellent post–somehow we need to amplify the point that monetary policy is actually tight.

  33. Gravatar of David Pearson David Pearson
    12. September 2011 at 09:16

    W. Peden,

    The “signalling” aspect of QE supposedly moves inflation expectations by communicating commitments about the path of future policy. However, if QE has no identifiable mechanism for affecting the price level in the short term, how could promising more QE in the future affect the price level now?

    I think we still need a theoretical construct for how QE works. There are three: 1) the interest rate channel; 2) the portfolio balances effect; and 3) signaling the path of future policy. Scott takes issue with the first. I take issue with the secon — the “hot potato” effect. I think it looks at only private sector securities holdings, and not its aggregate portfolio of claims. I also take issue with the “signaling” effect for the reason cited above (signaling requires a method that will work credibly in the future).

    In any case, if you can show me how the Fed can reduce the private sector’s aggregate duration or earnings volatility risk, then I would be glad to stand corrected.

  34. Gravatar of Scott Sumner Scott Sumner
    12. September 2011 at 17:18

    Martin, Sorry for mixing that up.

  35. Gravatar of Martin Martin
    13. September 2011 at 00:46

    Scott no prob, I am still amazed that you’re able to reply to all that’s thrown at you every day.

  36. Gravatar of W. Peden W. Peden
    13. September 2011 at 10:06

    David Pearson,

    “However, if QE has no identifiable mechanism for affecting the price level in the short term, how could promising more QE in the future affect the price level now?”

    Nope, I’m still reading you as being mind-numingly hydraulic, which I’m sure you aren’t. Assuming that (a) QE has hydraulic effects on the price level via boosting NGDP (with a very short lag, if any) and (b) people are not automatons, it’s no mystery why promising more QE would change people’s current actions. Just like promising to do the dishes in the future can influence other people’s behaviour to you in the present. So I think I’m just missing your point or you’re just stuck in a muddle caused by false premises (the denial of premise (a)).

    Your objection to the hot potato effect is based on private sector fears of fiscal risk from QE having an offsetting effect on the hot potato factor, right? Well, I think we both agree that there’s no evidence of that, but it’s an interesting hypothesis. Would it be fair to say that it assumes a government without monetary sovereignty and no capacity to raise debt whatsoever, i.e. a government that has to raise taxes at some point in order to finance any risk to the central bank’s portfolio?

  37. Gravatar of Guest post: Market Montarism and Financial Crisis (by Vaidas Urba) « The Market Monetarist Guest post: Market Montarism and Financial Crisis (by Vaidas Urba) « The Market Monetarist
    22. December 2012 at 00:27

    […] true, and accordingly, it is not important what assets do central banks hold. Sometimes it is even argued that the central bank is not a […]

Leave a Reply