Recent Posts

Sunday, June 14, 2015

How Should The Bank Of Canada Operate?

The Bank of Canada has asked for comments on potential changes on its operating procedures. Many of the potential changes appear to only matter to the central bank and its bank counter-parties, however others raise some philosophical issues about how central banks operate. I do not think that the Bank of Canada wishes to get into debates about the philosophy of central banking, but the reality is that when a crisis hits, we need a coherent view of what a central bank is supposed to be doing.

What This Consultation Paper Is About

The overview paper provided by the Bank of Canada explains the purpose of the public consultations.
The Bank believes that its operating framework has generally been effective in achieving the Bank’s objectives, but that it would be prudent to make some enhancements in various areas. In some cases, changes are being proposed to routine operations. In others, the proposals relate to introducing new tools to manage exceptional circumstances. The underlying objective of the enhancements is to increase the effectiveness of the overall framework in response to the ongoing and expected changes in the external environment, while managing the Bank’s market presence in an appropriate manner. No inferences should be drawn from these proposals about the current or future stance of monetary policy.
In interests of space, I will not go further into the background here. The overview paper offers a discussion of the role of the central bank as a lender-of-last-resort.

I found out about this consultation paper when asked about it by another participant at the CEA Conference. The deadline for responses is July 4th, which is somewhat unfortunate from the point of view of my writing here. It would have been preferable to publish this after other articles that introduce central bank operations, but doing so in that order would make this response too late.

Overnight Repo

The paper states:
While repurchase and repo operations are only occasionally required to reinforce the target for the overnight rate, their effectiveness seems to have declined somewhat in recent years. This appears, at least in part, to be because the market’s capacity to channel funds to entities that are in need of liquidity has become less effective, since direct counterparties of the Bank seem to have become less willing to borrow extra liquidity and redistribute it to other counterparties. Rather, decisions by institutions on whether or not to participate in these transactions when the Bank offers them seem to be mostly based on whether they themselves need the liquidity. [Emphasis added - BR] The Bank is therefore proposing to change some of the parameters of its SPRA and SRA operations with the intent to increase the effectiveness of these operations when they are required, while still limiting the Bank’s role in the intermediation of the overnight funding market. 
Questions (note SPRA/SRA are overnight repo operations):
SPRA/SRA Parameters
1. What are the potential effects of changing the pricing format of SPRA/SRA operations to a competitive and discriminatory rate auction format?
2. What implications, if any, of the proposed structural interventions in the SPRA/SRA overnight markets do you foresee for the behaviour of your firm or of other firms and for the overnight funding market more generally?
[Note: "repo" is the shortened version of "repurchase agreement", which is a form of lending against security collateral.]

This is an area that I had no direct contact with, but it does seem that moving over to a more flexible pricing process would allow the system to function better. One of the side effects of the crisis was that the circular flows in the inter-bank lending market were disrupted, and the central bank ended up acting more as intermediary. I think the best assumption going forward is that the inter-bank lending market should not be relied upon.

(There was a question about the LVTS system; since I was never involved with bank treasury operations, I have no comments.)

Term Repo

Against this background, the Bank is proposing to conduct term repos as part of its regular market operations. It is envisaged that such operations would help the Bank’s monitoring of liquidity conditions in term funding markets and encourage the further development of and liquidity in the longer-term repo market in Canada. 
In addition, conducting longer-term repos would help reduce the need for the Bank to acquire Government of Canada securities outright for balance-sheet-management purposes. Currently, the Bank’s assets consist almost entirely of Government of Canada nominal bonds and treasury bills acquired on a regular basis through purchases, on a non-competitive basis, at the government securities auctions conducted by the Bank as fiscal agent for the Government of Canada. The distribution of the Bank’s portfolio broadly reflects the composition of the federal government’s stock of domestic marketable nominal debt in terms of maturity distribution and the split between treasury bills and bonds. 
(Note: a term repo is a repurchase agreement with longer holding periods, such as 1 or 3 months.) Two out of the three questions in the consultation paper were quite specific, but I have a short response to the following.
What would be the likely impact of the Bank conducting regular longer-term repos on activity in the overnight and longer-term repo markets in Canada?
I think the effect of introducing term repos would match the objectives set out in the paragraphs above. By continuously operating in the term repo market, the bank will develop market intelligence in this market. Term repo financing should be encouraged, as it moves market players away from relying too much on overnight funding. Obviously, 1-month financing still has to be rolled (roughly) every twenty trading days, but it reduces the notional amounts that must be rolled.

The only question that term repo operations raise is that the pricing of a term repo depends upon the expected path of the overnight rate. The Bank hopes to avoid signalling what it thinks about the path of the policy rate by having the rate set in auction. That said, it still needs to determine a minimum bid rate for the auction. Moreover, the size of the auctions is going to be somewhat discretionary.

Under normal circumstances, the issue of price signalling will not matter. However, in some form of a crisis (either financial or fiscal), the Bank will have to decide what levels of term repo rates are unacceptable. In my view, the Bank should be doing this - the belief that the central bank only sets the overnight rate is a fiction. In reality, it should have a strong opinion about the level of 3-month rates, or even longer. However, not everyone will agree with that assessment.

Secondary Market Purchases Of Bonds

The Bank is considering lowering its purchases of Government of Canada bonds and bills at auction (the "primary market") from its currently fixed 20% of all auctioned bonds. (This is different than some other central banks, which are barred from buying at auction.) The Bank needs to hold assets against its liabilities (mainly notes and coin, but also the Federal government's deposit - the Consolidated Revenue Fund of Canada), and so the missing purchases would be made up by term repos (as discussed above) and secondary market purchases of bonds. The purchases would be concentrated in "off-the-run" bonds that were issued years before.


Secondary Market Purchases of Government of Canada Securities
1. What would be the impact on the benchmark and non-benchmark segments of the Government of Canada securities market of the proposed reduction in the Bank’s participation at nominal bond auctions and proposed secondary market purchases?
2. What measures or parameters should be included to minimize any potential negative effects of the secondary market purchases on market functioning? 
The effects of such a change is mixed.

  1. During the crisis, liquidity in off-the-run GoC bonds dried up. Regular operations by the Bank might help prevent a repeat (assuming that the primary dealers do not hoard the liquidity).
  2. It would tighten off-the-run spreads versus benchmarks, which technically might raise the interest rate faced by the Federal Government by several basis points.
  3. It means that the private sector should have a greater net supply of long-dated bonds, which are needed for liability-matching purposes. (Canadian institutions have embraced liability matching much more seriously than American institutions.) Pension funds need 30-year paper, but it is unclear who has a need for 15-year bonds.
  4. Unfortunately, such a change raises the very low odds of a crisis associated with "rollover risk". With a smaller percentage of guaranteed purchases by the central bank, the Ministry of Finance has greater exposure to whims of discretionary decisions made by the Bank of Canada.

Given the track record of the ECB with regards to its interactions with the treasuries of the euro area, I do not think this is a beneficial a change from the perspective of a Canadian taxpayer. I would prefer a system where the Bank only buys 10% of bonds with maturities 10-years and up, but still buys 20% of shorter-dated paper.

Securities-Lending Programme

The Bank is considering expanding its security-lending programme in order to reduce tightness in the repo market. (Incidentally, the Bank spells the word as "program", which does not seem to be the usual convention in Government of Canada publications.)

The previous intervention threshold was a level that is 50% of the Bank's target rate, which breaks down in a low interest rate environment. The suggest intervention thresholds seem to make more sense.

Contingent Term Repo Facility

In addition to the facilities already at its disposal, the Bank is proposing to create a Contingent Term Repo Facility (CTRF) to enhance its ability to respond to any future episodes of intense market-wide stress. Activation and deactivation of the CTRF would be at the Bank’s sole discretion, as conditions warrant, and would be intended to be used only to deal with severe market-wide liquidity stress. 
This is the most interesting part of the proposal, from my point of view. The outlined facility would give the Bank broad powers to open up the CTRF to financial entities beyond its usual set of counterparties (primary dealers). Moreover, the range of collateral would be broad - "such as securities issued or guaranteed by the Government of Canada and provincial governments; municipal securities; and certain private sector securities, including corporate bonds, bankers’ acceptances and asset-backed securities."

Such a facility would have proved to be extremely useful during the Financial Crisis. By opening up the range of potential counterparties, the Bank is recognising that the non-bank financial sector is large and important, and needs access to lender-of-last-resort facilities.

The expansion of the types of collateral raises the most questions. It needs to be done, as otherwise the facility would be largely useless. That said, it may be that too much emphasis is being made upon the last crisis, and the possibility that a future crisis may look different may not be taken into account.

If and when the Canadian housing market rolls over, the impact on the Canadian labour market will be severe. The banks have been protected by the CHMC's rather generous provision of default insurance to mortgage lenders. However, provincial governments will face severe budgetary shortfalls. The Canadian provinces have large economic footprints, in particular relative to American states. It takes little amount imagination to see provincial finances under attack if Canada faces a rapid downturn, and the Federal government decides to follow its instincts towards balancing the budget.

The analogy between the Canadian provinces and euro area states has been made for a long time. Political antagonism between the different players creates a toxic decision-making environment. We see this now in Greece, and we saw it in Alberta during the 1930s. By lending against provincial bonds, the Bank of Canada would end up in the middle of these political struggles, if not acting as a major political player itself. Putting into place such a programme without any firm rules of engagement is not a sign that government finances are truly understood.

In my view, the problem with this proposal is not that it open-ended, rather that it is contingent. The trend that was encouraged by Monetarism that central banks should only have a technocratic focus on the amount of government securities it owns (as well as the overnight rate) was a serious mistake. The central bank is a bank, and should act like one. It should operate continuously in wider funding markets, in order to make it possible to monitor developments, as well as steer the private sector away from too-brittle funding arrangements. Otherwise, the central bank is flying blind, and ends up having to lend against collateral it does not understand when a crisis hits (as was the case for the Fed in 2008).

I discussed this further in Appendix A.2 of Understanding Government Finance.
Moving to such a system [BR - of central banks lending against private sector assets] in the United States was a reform advocated by Hyman Minsky in Chapter 13 of Stabilizing an Unstable Economy. His concern was that the financial system has an innate tendency to drift from safe (“hedge”) financing schemes towards “speculative” or “Ponzi” financing. If the U.S. Federal Reserve was deeply involved in determining which assets it was willing to lend against, it could act as a countervailing force against this tendency for excessive risk. If it thought a type of lending was unsafe, it could make it ineligible for rediscounting at the central bank. Some specialist lenders may take their chances with such lending, but they would be outside the safety net of the lender-of-last-resort operations. Minsky argued: “Central banks have to steer the evolution of the financial structure.”*

I agree with Minsky that such a reform would be one of the few mechanisms that would bring some stability to financing arrangements. The catch is that I doubt that such muscular interventions into the banking system fit current political trends. Progressives would be horrified to see the government actively involved in providing financing for bankers, while free market advocates would be horrified by the scope of the government interventions. It might require the financial system to blow itself up in an even more impressive fashion to force deep reforms of this nature.

* Page 359 of Stabilizing an Unstable Economy, by Hyman Minsky. Published by McGraw Hill, 2008.

(c) Brian Romanchuk 2015


  1. Brian, how about we go back to thinking that "the government always spends twice". That is, when the Treasury keyboards my pension into my Bank Account, it is the "matter" part of "out-of-thin-air" fiat spending process. The "anti-matter" part is kept by my Bank as an asset to balance the liability the government has just given it; my pension. The anti-matter part, only exists in the virtual dimension of the Central Bank and its clients.

    In order to try and bust the neo-liberal myth, than sovereign fiat currency "issuing" governments, have the same restrictions as fiat currency "using" households or firms, Canada should adopt the "Reserves" concept of interbank settlement. Your Mr Carney, has not indicated, so far, that he wishes the Bank Of England to adopt the Bond/Bills settlement system in preference to a Reserves model.

    PS. Congratulations on "Understanding Government Finance", an excellent read. I was not aware that Canada did not use the Reserves system.

    1. Thanks!

      The use of bonds/bills is a simplified explanation I use. They need to cover net spending, since I assume that settlement balances remain at zero. In practice, central banks will use lending operations (repo) to adjust balances, but a repo is just a matched buy/sell operation. (In my book, this lending is introduced in Chapter 5, after I explain the "no reserves" system.) It's more convenient to use lending to adjust settlement balances, but it is not necessary.

      Switching over to use "reserves" (settlement balances) instead of bonds/bills would eliminate the need for bond issuance, but it would cause other problems. I got involved in a discussion on Twitter this morning about that topic. I guess I will try to address it "soon", assuming that I do not need to say anything about the Greek situation.


Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts. I get no warning about this, and only go through my "spambox" infrequently. The best bet it to keep comments short, and if you think the spam filter struck, let me know with a short comment.