Thursday, November 7, 2013

Rates or quantitites or both

Roaming around the econ blogosphere, I often come across what seems to be a sharp divide between those who think monetary policy is all about the manipulation of interest rates and those who think it comes down to varying the quantity of base money. Either side get touchy when the other accuses its favored monetary policy tool, either rates or quantities, of being irrelevant. From my perch, I'll take the middle road between the two camps and say that both are more-or-less right. Either rates, or quantities, or both at the same time, are sufficient instruments of monetary policy. Actual central banks will typically use some combination of rates and quantities to hit their targets, although this hasn't always been the case.

Just to refresh, central banks carry out monetary policy by manipulating the total return that they offer on deposit balances. This return can be broken down into a pecuniary component and a non-pecuniary component. By varying either the pecuniary return, the non-pecuniary return, or both, a central bank is able to create a disequilibrium, as Steve Waldman calls it, which can only be re-equilibrated by a rise or fall in the price level. If the net return on balances is sweetened, banks will flee assets for balances, causing a deflationary fall in prices. If the return is diminished, banks will flock to assets from balances, pushing prices higher and causing inflation.

The pecuniary return on central bank balances is usually provided in the form of a promise to pay interest, or interest on reserves.

The non-pecuniary return, or convenience yield, is a bit more complicated. I've talked about it before. In short, it's sorta like a consumption return. Because central bank balances are useful in settling large payments, and they are rare, banks find it convenient to hold a small quantity of them as a precaution against uncertain events. This unique convenience provided by scarce balances is consumed over time, much like a fire extinguisher's property as a fire-hedge is consumed though never actually mobilized. By increasing or decreasing the quantity of rare balances, a central banker can decrease or increase the value that banks ascribe to this non-pecuniary return.

Now some examples.

The best example of a central bank resorting solely to the quantity tool in order to execute monetary policy is the pre-2008 Federal Reserve. Before 2008, the Fed was not permitted to pay interest on reserves (IOR). This meant that the only return that Fed balances could offer to banks was a non-pecuniary convenience yield, a point that I described here. By adding to or subtracting from the quantity of balances outstanding the Fed could alter their marginal convenience, either rendering them less convenient so as to drive prices up, or more convenient so as to push prices down.

The Bank of Canada is a good example of a central bank that uses both a quantity tool AND an interest rate tool, though not always both at the same time. Since 1991, according to Mark Sadowski, the BoC has paid interest to anyone who holds overnight balances. This is IOR, although in Canada we refer to it as the deposit rate. In addition to paying this pecuniary return, BoC balances also yield a non-pecuniary return. Banks who hold balances enjoy a stream of consumptive returns, or a convenience yield, that stems from both the rarity of BoC balances and their exceptional liquidity.

The best way to "see" how these two returns might be decomposed is by looking at the short term rental market for Bank of Canada balances, or the overnight market. In Canada, this rate is called CORRA, or the Canadian overnight repo rate. A bank will only part with BoC balances overnight if a prospective borrower promises to sufficiently compensate the lending bank for foregone returns. Assuming that the Bank of Canada's deposit rate is 2%, a potential lender will need to be compensated with a pecuniary return of at least 2% in order to dissuade them from socking away balances at the BoC's deposit facility.

The lender will also need to be compensated for doing without the non-pecuniary return on balances. If the overnight lending rate, CORRA, is 2.25%, then we can back out the rate that a lender expects to earn for renting out the non-pecuniary services provided by balances. Since the lender of balances receives the overnight rate of 2.25% from the borrowing bank, and 2% of this can be considered as compensation for foregoing the 2% pecuniary return on balances, that leaves the remaining 0.25% as compensation to the lender for the loss of the non-pecuniary return.

So in our example, the pecuniary and non-pecuniary returns on BoC balances are 2% and 0.25% respectively, for a total return of 2.25%.

The Bank of Canada meets each six weeks, as Nick Rowe points out, upon which it promises to provide banks with a given return on settlement balances, say 2.25%, for the ensuing six week period. When it next meets, the Bank will  introduce whatever changes to this return that are considered necessary for it to hit its monetary policy targets. The BoC can modify the return by changing either the pecuniary component of the total return, the non-pecuniary component, or some combination of both.

Say it modifies only the non-pecuniary component while leaving their pecuniary return untouched. For instance, with the overnight rate trading at 2.25%, the BoC might announce that it will conduct some open market purchases in order to increase the quantity of balances outstanding, while keeping the deposit rate fixed at 2%. By rendering balances less rare, purchases effectively reduce the non-pecuniary return on balances. As a reflection of this shrinking return, the overnight rate may fall a few basis point, or it may fall all the way to 2%. Whatever the case, the rate at which banks now expect to be compensated for foregoing the non-pecuniary return on balances has been diminished. Banks will collectively try to flee out of overpriced clearing balances into assets, pushing up the economy's price level until balances once again provide a competitive return. This sort of pure quantity effect is the story that the quantities camp likes to emphasize.

The story told by the quantities camp is exactly how the BoC loosened policy between April 2009 and May 2010. At the time, the BoC injected $3 billion in balances *without* a corresponding decrease in the deposit rate. The overnight rate fell from 0.5% until it rubbed up against the 0.25% deposit rate. The lack of a gap between the overnight rate and the deposit rate indicated that the injection had reduced the overnight non-pecuniary return on balances to 0%. After all, if lenders still expected to be compensated for forgoing the non-pecuniary return on balances, they would have required that the overnight rate be above the deposit rate.

The BoC's decision to reduce the overnight non-pecuniary return on balances to 0% would have generated a hot potato effect as banks sold off lower-yielding BoC balances for higher-yielding assets, thus pushing prices higher. A change in quantities, not rates, was responsible for the April 2009 to May 2010 loosening.

Likewise, in June 2010, the BoC tightened by using quantities, not rates. Open market sales sucked the $3 billion in excess balances back in, thereby increasing the marginal convenience yield on central bank balances. The deposit rate remained moored at 0.25%, but the overnight rate jumped back to 0.5%, indicating that the overnight non-pecuniary return on balances had increased from 0% to 0.25%. This sweetening in the return on balances would have inspired a portfolio adjustment away from low-yielding assets into high-yielding central bank balances, a process that would have continued until asset prices had fallen far enough to render investors indifferent once again along the margin between BoC deposits and assets. Once again quantities, not rates, did all the hard work.

While the BoC chose to tighten in June 2010 by changing quantities, it could just as easily have tightened by changing rates. For instance, if it had increased the deposit rate to 0.5% while keeping quantities constant, then the net return on balances would have risen to 0.5%, the same return that was generated in the last paragraph's quantities-only scenario. This sweetening in the return on balances would have caused the exact same chain of portfolio adjustments and falling asset prices that the quantities-only scenario caused.

Alternatively, the BoC could have tightened through some combination of quantities AND rates. It might have increased the deposit rate from 0.25% to 0.4%, and then conducted just enough open market sales to increase the non-pecuniary return on balances from 0% to 0.1%, for a total combined return of 0.5%. The ensuing adjustments would have been no different than if tightening had been accomplished by quantities-only or rates-only.

Putting aside the period between April 2009 and June 2010, does the Bank of Canada normally execute monetary policy via rates or quantities? A bit of both, I'd say. At the end of a six week period, say that the Bank wishes to tighten. It typically tightens by announcing a 0.25% rise in its target for the overnight rate combined with a simultaneous 0.25% rise in the deposit rate. The overnight rate, or the rental rate on clearing balances, will typically rise immediately by 0.25%, reflecting the sweetened return on balances.

Did rates or quantities do the heavy lifting in pushing up the return on balances? Put differently, was it the threat that open market sales might increase the convenience yield on balances that tightened policy, or was it the improvement in the deposit rate? I'd argue that the immediate punch would have been delivered by the change in the deposit rate. CORRA, the rental rate on balances, jumped because overnight borrowers of BoC balances were suddenly required to compensate lenders for the higher pecuniary rate being offered by the BoC on its deposit facility. Quantities don't enter into the picture at all, at least not at first. The rates-only camps seems to be the winner.

However, as the ensuing six-week period plays out, market forces will push the rental rate on BoC balances (CORRA) above or below the Bank's target, indicating an improvement or diminution of the total return on balances. The BoC has typically avoided any incremental variation of the deposit rate to ensure that the rental rate, or return on balances, stays true to target over the six week period. Rather, it has always used quantity changes (or the threat thereof) to modify the non-pecuniary return on balances during that period, thereby steering the rental rate back towards target. First rates, and then quantities, conspire together to create Canadian monetary policy.

To sum up, the Bank of Canada's monetary policy is achieved, it would seem, through a complex combination of rate and quantity adjustments. The rates vs. quantities dichotomy that sometimes pops up on the blogosphere simplifies what is really a more nuanced story. Monetary policy can certainly be carried out by focusing on quantity adjustments to the exclusion of rate adjustments (as was the case with the pre-2008 Fed) or vice versa . However, modern central banks like the Bank of Canada use rates, quantities, and some combination of both, to achieve their targets.

Note: The elephant in the room is the zero-lower bound. But the zero lower bound needn't prevent rates or quantities from exerting an influence on prices. On the rates side of the equation, the adoption of a cash-penalizing mechanism along the lines of what Miles Kimball advocates would allow a central bank to safely push rates below zero. As for the quantity side of the equation,  the threat of Sumnerian permanent increases in the monetary base may not be able to reduce the overnight non-pecuniary return on balances once that rate has hit zero, as Steve Waldman points out... but they can certainly reduce the future non-pecuniary returns provided by balances. Reductions in future non-pecuniary returns should be capable of igniting a hot potato effect, albeit a diminishing one, out of balances and into assets.


  1. JP: Good analysis if we zoom in very close.

    Now let's zoom out. The Bank of Canada cannot set interest rates, if it wants to keep inflation at the 2% target. It is forced to adjust interest rates to match the market equilibrium compatible with that 2% target. Which means the market, plus the 2% inflation target, is what sets interest rates.

    Now let's zoom waaay out. Suppose the Bank of Canada wanted to permanently double the price level path. In the long run, all interest rates would stay the same along that new target path. Only nominal quantities would double.

    You can tell the story of monetary policy using quantities at all three levels of zoom. You can't tell the story with rates.

    The rates people can only see what's right in front of their noses. The quantity people can see the big picture too!. (Plus, the rates people have this terrible tendency to mistake a social construction of reality for reality!)

    1. Or, assume the Bank of Canada raised the inflation target to 3%. All rates would rise by (roughly) 1%. The rates people say that raising interest rates means tightening monetary policy. But in this case it means a looser monetary policy. The rates people are wrong again. The quantity people know that this will mean nominal quantities will be rising 1% faster than before.

    2. Plus, the rates people suffer from Wicksellian and Neo-Wicksellian indeterminacy. The former can't explain P; the latter can't explain anything, without just assuming, for no reason whatever, that the economy always converges to full employment, even though it's not in their model.

    3. I'm surprised you're not on board with this post since I see it as a somewhat droning rendition of your concise comment on SRW's blog.

      "It is forced to adjust interest rates to match the market equilibrium compatible with that 2% target. Which means the market, plus the 2% inflation target, is what sets interest rates."

      Hmmm. The way I see it, the BoC is forced to adjust the return on balances to match the market equilibrium compatible with the 2% target. The return can be adjusted by either changing the deposit rate ("the interest rate") and/or the convenience yield (by changing quantities). So the market, plus the 2% inflation target, are what ultimately set IOR and/or the quantity of balances necessary to set a certain convenience yield.

    4. So you're fine with the idea that when we see a tightening in BoC policy and the overnight immediately jumps by 0.25%, it is the deposit rate, and not open market operations, that has done the lion's share of the lifting?

    5. Nick,

      As a rates person, I see the price level in the distant future as the outcome of today's price level and the various inflation rates between now and then. The inflation rates are some function of the rates between now and then. Beyond that I don't care what the price level is in the future. It has no meaning as an absolute level, it only matters for its relative value.

      Furthermore, you seem to suggest that the price level in the distant future is a function of quantities only in the future. However, if real quantities are path dependent then you need to know all the nominal quantities from now till then in order to deduce the final price level. In which case, you're in no different position from the person who needs to know all the rates from now till then.

    6. JP: I'm very much on board with your post, if we want a zoom-in view of what the BoC does now. I just wanted to add the zoom-out perspectives.

      @4.31: I'm fine with the idea that, given the BoC's current operating procedure, the actual overnight rate cannot go (much) outside the 50 basis point band, regardless of what it does with quantities. But that is no different from saying that the price of gold cannot go outside the small band (created by transport and transactions costs) around the central bank's target price of gold. But I wouldn't use that to argue that therefore, both rates and quantities are irrelevant, and the central bank's (explicit or implicit) target price of gold is always and everywhere what does the heavy lifting.

      It's begging the question to say that since the overnight rate is largely determined by the discount rate and bank rate, rates do the heavy lifting in monetary policy, because it's just saying that rates largely determine rates.

  2. Good post. I think I am in the process of arriving to a similar conclusion as I go through and post about Bindseil's book, but I don't want to speak too soon. In any case, this is pretty exciting for me, since this is a huge area of debate, and I think I've found all the material I need to develop a robust opinion about this. I'm not sure if Bindseil's models incorporate your idea of a convenience yield, so I'll have to see / may have to unify.

    Then once I do all that, I'll blow things out to the monetary policy strategy level, as opposed to the implementation level, to give Nick Rowe's ideas a fair shot.

    1. Thanks. I've been following your Bindseil posts faithfully. The ones on rates vs quantities provided me with some fodder for this post.

      I'm not sure if the convenience yield is a necessary idea, but it sure helps me to understand what is going on.

    2. And thanks to you for reading. I regret them being so long. Perhaps I’ll summarize the summaries at some point.

      Here are my current thoughts on the connection between that Bindseil post you referred to, this post, and Nick Rowe’s comments – again, with the risk of speaking too soon. Basically, I don't think Bindseil falls plainly into the blogosphere rates camp you refer to here (not that you were necessarily suggesting he does). There are several layers here:

      1. What type of variable should the central bank's operational target be at the day-to-day implementation level?

      2. What can the central bank do to hit that operational target?

      3. What's the monetary policy strategy model that decides what the operational target should be at any given point in time?

      I read your post very early this morning, but if I recall correctly, the point you're making here is that central banks can and do use a mix of quantities and rates to accomplish (2). In another post or two, I anticipate I will indeed be showing that Bindseil's models support this idea - which is why I said I think I'm coming to a similar conclusion as you.

      However, I do get the sense that Bindseil is an ardent rates guy with respect to (1). I think he would reject the notion that central banks can reliably target monetary aggregates on a day-to-day basis at the implementation level. He would say that theory and evidence suggests that the operational target needs to be some form of interest rate.

      That said, and I think this is somewhat consistent with Nick Rowe, he would allow for the idea that monetary aggregates could be intermediate/final targets in the context of the monetary policy strategy model in (3). In other words, quantities would be the longer-term target, but the central bank would be steering the economy there by manipulating interest rates on a day-to-day basis. It would the quantity target, however, that would determine the interest rate (and in this sense, the CB would not be fixing rates). Or substitute anything else for the long-term target - such as inflation, employment, NGDP, etc.

    3. "...blogosphere rates camp you refer to here"

      I probably shouldn't have used the "two camps" trope since the classical battle is over whether a central bank should target an overnight rate or some quantity of base money. In this post I'm assuming the first, which is pretty much how things are done these days, and considering whether quantities or rates get you there. Which I realize is not really the classical rates vs quantities debate, and that might cause some confusion.

    4. Right, with the clarification that the target you're talking about is an operational target.

    5. Hi JP - see my posts here , particularly the one 'implications for steering the rate.' I provide a matrix of possibilities for how the central bank can steer rates under a certain model of the interbank market. I think I'm pretty much saying what you're saying here.

    6. And a quick clarification - the model I focus on there is not exactly the opportunity cost / convenience yield model you have in mind, but they're related through arbitrage. The individual shock model I review I think is more aligned with your thinking. But I think both get you to the same place at the end of the day. Under uncertainty, the CB can steer rates using rates and/or qty. However, see the section on certainty in the aggregate liquidity model post. I think here you could argue that rates mostly dominate...

  3. "Either side get touchy when the other accuses its favored monetary policy tool, either rates or quantities, of being irrelevant.From my perch, I'll take the middle road between the two camps and say that both are more-or-less right."

    I agree that both are more or less right. Relevance is the key word though. Targeting base or reserve interest rates is less relevant to the economy when compared to broad money therefore limiting the effectiveness of policy. This can be addressed by making base equal to broad money and allowing the broader public to directly transact in it and deposit it at the fed. Or the fed can target market rates of interest which arent at zero through crediting deposits at commercial banks of broad public.

  4. “Roaming around the econ blogosphere, I often come across what seems to be a sharp divide between those who think monetary policy is all about the manipulation of interest rates and those who think it comes down to varying the quantity of base money.”

    I think this is a fundamental point of perception, but any characterization of it becomes semantically and meaningfully challenging.

    I think you’ll find that a number of those on the “rate side” are very knowledgeable on how central banks use the quantity of reserves – and price various tiers of that quantity – in order to set and maintain interest rate targets. I think Scott Fullwiler among the MMT’ers was prominent insofar as early published blogosphere exposition is concerned, with a number of papers, well ahead of QE but accurately anticipating such potential effects on the monetary base in terms of both quantities and administered and market rates. Marc Lavoie covers it all perfectly and elegantly within ‘Monetary Economics’ as published in 2006. Cullen Roche certainly understands it. My own interest in the subject stems from hands on experience with the reserve management function.

    Conversely, those on the monetary base “side” seem determined to debase the importance of interest rates in the discussion.

    So at the front end of monetary operations, I think the distinction between the two groups is somewhat asymmetric, but that’s just my impression I suppose.

    It’s always about both, in some sense. There are definitely things called a target fed funds rate and (administered) interest rates paid on reserves, etc., and similar things in Canada and other places. And those interest rate targets are characterized by more fixedness (even if re-examined at monetary policy meetings) than are reserve quantities in the short run. The operational function is to hit set targets, or administer set targets, using active central bank reserve management – QE or no QE – zero bound or not. With QE, the fact that the central bank creates chronically excess reserve balances while still setting a rate target of sorts makes it about both as well, but in a different regime of reserve management.

    And the idea that those interest rate settings be readjusted periodically – for whatever reason - shouldn’t be controversial.

    ... cont'd


  5. This piece by the Fed is the best I’ve seen on reserve management systems:

    Relative to that framework, the normal Bank of Canada system is a symmetric channel system, as generically described in the paper (where the objective implies an intended positive convenience yield according to your definition). The normal (pre-crisis) Fed system is a kind of asymmetric channel with a lower rate of zero. And QE implies a floor system (but not necessarily vice versa).

    The title of the Fed paper – “Divorcing Money from Monetary Policy” - is based on the availability of a floor system as an option, but it’s the chosen quantity of reserves that requires floor pricing, so it’s still “about both” in that sense. Generalized phrasing regarding one or the other should probably be avoided.

    What has happened during the crisis (the way you describe it in your post) is a kind of meandering by the Bank of Canada between the normal symmetric channel and a quasi-floor system. The convenience yield in a floor system is structurally zero (when convenience is defined as the spread between an administered central bank rate and a corresponding market interbank rate. Regarding that, I don’t know why some are bringing treasury bills and bonds into the picture. Those kinds of spreads are an entirely separable issue of credit quality and collateral pricing in my view.)

    The Fed paper discusses the subject matter entirely in terms of opportunity cost. To the degree that you define convenience yield as a sort of “offset” to opportunity cost, it seems redundant. It may be a nice way of thinking about things that I haven’t fully wrapped my head around yet, but it does seem analytically repetitive when juxtaposed against the idea of opportunity cost. As I said in a previous comment, it seems to me that the idea of convenience yield is necessarily isomorphic to opportunity cost.

    The referenced Fed paper is very operational in nature, as it should be, and as the subject is. The idea of the central bank aiming for these various rate settings at an operational level in my view should be cleanly separated from (but connected to) the more strategic upstream idea of how the central bank comes up with its short operational interest rate targets on the basis of monetary theoretic or other cognitive inputs. The upstream idea doesn’t supersede the downstream one. Both are necessary. And along those lines, the idea of the central bank manipulating reserve balances at an operational level to achieve interest rate and/or quantity targets (including general strategic QE objectives) should be made distinct from the idea of the broader monetary base being targeted in some strategic sense according to particular monetary theories. But I think those monetary theories are often not congruent with what amounts to the constraint of what the implementation options actually are at the operational end of things. And this frequent conflict is well covered by people like Fullwiler and Lavoie.

    1. Thanks for the link, will check it out.

      "Regarding that, I don’t know why some are bringing treasury bills and bonds into the picture. Those kinds of spreads are an entirely separable issue of credit quality and collateral pricing in my view."

      Yes, that confused me too. The key spread is the overnight to deposit rate spread.

      On the "two camps" trope that I used in my post, that may have not been the best approach. See my comment to ATR a few minutes ago.

      "To the degree that you define convenience yield as a sort of “offset” to opportunity cost, it seems redundant. It may be a nice way of thinking about things that I haven’t fully wrapped my head around yet, but it does seem analytically repetitive when juxtaposed against the idea of opportunity cost."

      I use the term convenience yield to refer in general to the opportunity cost of foregone monetary services on any asset. (The other opportunity cost in my post is foregone interest-on-reserves). In Canada the full opportunity cost on balances is the sum of the foregone interest and foregone monetary services -- the convenience yield is the specifically monetary component (the distance between the deposit rate and the overnight rate).

      I like the term convenience yield because it is already used in finance. Monetary policy is only one of the many subtopics in which the notion of a convenience yield arises. I often use "moneyness" interchangeably with convenience yield. Assets may provide a range of convenience yields -- the one that I focus on in this blog is the monetary convenience yield.

    2. I don't understand.

      Reserves left overnight earn the lower corridor rate - the deposit rate.

      Where's the opportunity cost associated with that?

    3. From the perspective of a potential lender, they will forgo the deposit rate if they lend reserves. So they need to be compensated by a return that is at least as good as the deposit rate if they are to lend.

    4. JKH - thanks for posting that paper. Perfect timing for me. I had bookmarked it at one point, finding it essential reading but not 100% understanding it. Now that I'm reading Bindseil, I think I understand it much better. Bindseil follows Poole (1968), as they do, which is basically what underlies the shape of those graphs. My upcoming post(s) on Bindseil will be explaining what’s behind this, with more detail and, in my opinion, more clarity than Keister et al. But they do a nice job, particularly considering space limitations and the graphs, which Bindseil doesn’t incldue. BTW, they refer to a “liquidity effect” in the Fed paper – which has been a point of contention of Fullwiler’s and Lavoie’s. I’ll want to explore this too.

      Also completely agree with your last paragraph.

  6. Thanks, JP.

    Finally, that explains my confusion in trying to understand this.

    That is not how I would interpret or define opportunity cost.

    If I’m a reserve manager in the pre-crisis Fed regime, and the interbank rate is 2.25 per cent and the rate earned on excess reserves is zero, then the opportunity cost of the excess reserve balance I end up with at the end of a reserve averaging period is 2.25 per cent on that balance.

    On the other hand, if I’m operating in a corridor system where the interbank rate is 2.25 per cent and the rate paid on deposits at the CB is 2 per cent, the opportunity cost on those same excess balances is .25 per cent.

    The opportunity cost is the cost as defined by the differential return between the two outcomes in either of those cases.

    1. I see. I'm using opportunity cost in the sense of the next best alternative. In the end, no major difference here. We can still have a conversation where you call it the opportunity cost and I call it the convenience yield ;)

    2. Agreed. That’s excellent.

      My world has resumed its normal axis rotation, and I’m no longer in danger of being hurled into space forever.

  7. Hi JP. Regards the BoC, aren't they clearly targeting the overnight rate? It looks like the BoC decided to target the lower band when the settlement bal increase to $3B ( They didn't have a choice but to adjust quantity because the overnight was *above* their target. I think the distinction between interest rate and quantity is artificial in the interbank market because interest rate <-> price <-> supply/demand quantity. From a practical perspective, BoC clearly says they implement monetary policy through commercial interest rates and the exchange rate ( I assume they mean all commercial interest rates and FX forward rates ripple out from the overnight rate). From a practical perspective, for citizens, unless the BoC controls the auction of retail credit (e.g. mortgages), price (interest rates) is the only implementation (i.e. commercial banks don't operate on the principle of quantity, other than via capital requirements).

    1. BTW The quantity theory aspect of monetary policy is something that I just don't get (for a modern economy). For base money, the distinction of quantity and rates seems artificial (above). For the wider economy, where the "payments system" is otherwise known as "working capital" (think corporations), I don't see how base quantity controls 'working capital" quantity other than through the interest rate channel. Of course this is just a rephrasing of the irrelevance of velocity, but more important, as a plea to monetary economists, what is the transmission channel from base quantity to Apple's "working capital" other than through interest rates?

    2. "They didn't have a choice but to adjust quantity because the overnight was *above* their target."

      The BoC chose quantities, but it could also have chosen rates. They could have lowered the bank rate, the rate that defines the top of the BoC's operating band, downwards while keeping the deposit rate fixed. That would have driven the overnight towards the deposit rate, creating the very same effect that the clearing balance injection did.

  8. jt26 "I don't see how base quantity controls 'working capital" quantity other than through the interest rate channel"

    I'm also puzzled by the claim that it does. Thinking about how the gold standard actually operationally constrained the price level might help I guess. Did all of the "constraining" occur on those ocassions when crises caused mass defaults and firesales. My impression is that the monetary base would constrain prices only if, when push came to shove, more monetary base were not supplied and instead forced asset sales, debt write downs and foreclosures caused prices to fall back down. Since modern central banks are not going to let that happen, the quantity of monetary base is never going to be an issue other than being one of several possible tools for adjusting interest rates????

    The other thing that strikes me is that asset prices are not really the same as prices for goods and services are they? The link between interest rates and NGDP requires more than just looking at asset prices. It has to take into account how interest rates influence consumer spending by influencing job security and household debt burdens.

  9. The other big factor I guess is how interest rates influence "hot money" flows between countries and so shift exchange rates and in that way influence prices and unemployment. Falling interest rates help to induce rising asset prices and that entices in "hot money" inflows and makes the currency stronger and so makes imports cheaper and frightens workers off making wage demands.

  10. Could the "base quantity doesn't set prices" argument be summed up as saying that nowadays more monetary base always gets supplied in reaction to a demand for it and before it becomes a binding constraint. So whether just enough gets supplied only at the last minute or whether far more than enough gets supplied decades before it is needed is neither here nor there.
    A bit like how foot binding caused C19th Chinese women to have small feet but today we wont get bigger feet by wearing shoes three sizes too big.

  11. Further to my last comment, it would be mistaken to claim that nowadays "shoe size determines foot size in the long run".

    Shoe size would only determine foot size in the counterfactual scenario where a child was forced to permanently wear footware that she had outgrown as in C19th footbinding.