Why not split up the Reserve Bank?

That is, deliberately, a slightly ambiguous title for the post.   I favour splitting the Reserve Bank in two, setting up a new New Zealand Prudential Regulatory Authority to pick up the regulatory and supervisory responsibilities the Bank currently has (subject to various refinement).   I’ve made the case here and here, and late last week highlighted my former colleague Geof Mortlock’s new article articulating the case for change.   We know that the Reserve Bank’s conduct of those functions isn’t highly regarded and the minutes (and OIA releases) suggest that the Reserve Bank’s Board –  paid to hold the Governor to account in the public interest –  were asleep at the wheel while this situation was developing.  All told, I reckon there is a pretty clear-cut case – in principle, and based on the actual track record – for structural reform.  Structural reform is never a panacea, but with the right will, and the political determination that things will be done better in future, it can play a part, making a demonstrable break with the past and establishing a new and better institutional culture.

But, of course, there are counter-arguments.  The Reserve Bank is likely to be making them forcefully at present.  I suspect the Governor’s populist participation in the current attack on the banks – not grounded in any of his own legislation, not related to any systemic soundness threats – may be a part of that effort, wanting to appear “useful” to the government and somehow “in touch” with some sort of “public mood”.  (If so, that in itself should be grounds enough for structural and personnel changes.)

If I were the Bank I would probably be trying to arrange a visit from (Sir) Paul Tucker, former Deputy Governor of the Bank of England, who has just published a new book Unelected Power, on the delegation of economic powers to independent agencies, with a particular focus on central banks.  Despite his background on the markets side of the Bank of England, Tucker didn’t leave officialdom to make money in the financial sector, but instead turned to academe and has produced an impressive book.  I’ve already referred to it in a couple of recent posts, and expect to do so in a few more relevant to the current efforts to overhaul the Reserve Bank Act.    As one UK commentator recently described it

….it is mainly about central banking, and on this is it authoritative. It will be an essential read for everybody involved in monetary policy, or researching it.

Not by any means will everything he writes be music to the ears of our central bankers, but Tucker’s views on the structural separation issue will be.  Perhaps that isn’t too surprising, since he was Deputy Governor at a time when the British government was bringing the financial sector supervisory and regulatory functions back inside the Bank of England (albeit with separate government-appointed decisionmaking structures for the various functions).  The Bank of England model informed Iain Rennie’s report last year, and if a decision is finally made to leave all the existing Reserve Bank functions together a structure like that of the Bank of England (but slimmed down for our circumstances) would probably be the way to go.

Tucker doesn’t argue that the regulatory and supervisory functions have to be in the same institution as monetary policy but that, subject to certain important conditions, it is better if they are.    To my mind, one of the weaknesses of his book is that it is very focused on the US and the UK (with some discussion of ECB/Europe but the issues are very different there given the idiosyncratic relationship between the ECB –  set up by international treaty and not really very accountable to anyone –  other EU institutiuons, and the national states of the EU/eurosystem).  Thus he simply doesn’t engage with the experience of the many other advanced countries – in fact, most of those outside the euro –  where the primary role in prudential regulation/supervision is undertaken by an entity other than the central bank.   This was the list of such countries I ran the other day

Canada, Australia, Norway, Sweden, Korea, Japan, Poland, Chile, Turkey, Mexico, Switzerland, and Iceland

It is a mix of large and small, of countries with very big financial systems operating internationally and countries with mostly domestic banks, of countries where the split has been longstanding (eg Canada) and countries where it has been more recent (eg Australia), and countries that ran into financial crisis in 2008/09 and countries that did not.  I’d find Tucker’s claim for the superiority of his model (essentially the UK one) more compelling if he’d addressed the experience of some of these countries.

As I read his material on this issue, it seemed to me that Tucker had two main arguments for keeping prudential supervision/regulation and monetary policy together.

The first of these is about the central bank’s lender of last resort function (for which the relevant statutory provision in our legislation needs refinement).  A central bank is the only agency with the unquestioned ability to provide immediate liquidity to an individual bank, or to the system as a whole, when severe liquidity pressures arise –  whether it is a run to physical cash, or simply a freezing up of interbank markets leaving some plays unable to operate with external injections of liquidity.     Failure to respond to systemwide increases in liquidity demand will, all else equal, be likely to result in the central bank falling short of its inflation target (through the resulting financial crisis and economic shakeout).  Provision of liquidity, and a responsiveness to changes in demand, is an integral part of modern central banking (even though the stress events may not occur even as frequently as once a decade –  in the New Zealand case, Y2K and the liquidity pressures around 2008/09.

Liquidity provision usually involves either buying assets outright from a bank, or lending on the security of those assets (mainly repo agreements).  That is easy when it involves the outright purchase of a well-known widely-traded asset like a government bond.  But it gets trickier when it is a loan (in economic substance, if not legal form) and when the assets involved are pretty opaque and not generally traded at all, and when the general injunction –  enshrined in legislation in some countries –  is that central banks should only lend to solvent institutions (solvency here being a positive net assets test, not an “ability to meet payments as they fall due” test).    To caricature just slightly, Tucker argues that a PhD in macroeconomics won’t be much use in enabling a central bank to decide whether or not to provide funds to be bank that comes knocking.  You need, instead (or as well), detailed banking and credit expertise.  Specifically he notes

Even opponents of “broad central banking” generally accept that, as the lender of last resort, the central bank cannot avoid inspecting banks that want to borrow.

Going on to argue that

A central bank must be in a position to track the health of individual banks during peacetime if it is to be equipped to act as the liquidity cavalry.

I’m not persuaded, for a number of reasons.  First, of course, we don’t “inspect” banks in New Zealand at all (but that is trivial point).   Second, clearly many countries operate with exactly the sort of model Tucker deems impossible or inadmissible (presumably by relying on some peacetime exchange of information, and wartime written recommendations or assurances from the prudential regulatory agency).   Third, the central bank making the decision to lend is not the only option: it would be possible to envisage a model in which the central bank was simply the operational agent, but the credit risk from any crisis support to an individual institution was taken directly by the government, on the advice (and analysis) of the prudential regulatory agency.   And, finally, LLR powers aren’t the only relevant ones.  In the 2008/09 crisis, perhaps the most important single regulatory response was the deployment of the Minister of Finances’s extensive guarantee powers (under the Public Finance Act).    The Minister of Finance doesn’t do prudential supervision, and to the extent he needed comfort that institutions guaranteed were likely to be solvent, he had to rely on advice from officials.  In this case, it was primarily the Reserve Bank, but it could as easily have been advice from a Prudential Regulatory Authority.   The same gooes for choices (regrettable as they may be) to bail out individual institutions.

My claim isn’t that there can never be any advantages in having prudential supervision and central bank liquidity operations in the same entity, just that the case for them to be so isn’t generally compelling once set against the other arguments for structural separation.  In a crisis, it is typically all hands to the deck (including, for example, the Treasury and the Reserve Bank working together, even though they are separate institutions) and. more generally, there are numerous examples of interagency arrangements for information sharing.   It is undoubtedly important for all relevant agencies to coordinate closely, and have in place appropriate protocols and be prepared to run exercises to war-game the handling of crises.  But the functions don’t all need to be in the same agency, and there are likely to be costs in normal times to having them all together.

One point Tucker touches on elsewhere, but not here, is the importance of having people running functions believing in them.  It is, for example, dangerous to have financial supervision (or AML) in an institution where the chief executive doesn’t really believe in the importance or value of the function.  Reasonable people could argue that that –  rather than separation –  was part of the UK’s problem in 2007: the then Governor, Mervyn King, seemed quite averse to lender of last resort responsibilities, even though they were still an intrinsic part of the powers assigned to the Bank of England.

Tucker’s other main argument for keeping prudential and monetary policy functions together is one he lists under a heading “Harnessing the authority of the central bank”. noting that

If an economy’s central bank is already endowed with both authority and legitimacy, giving it responsibility for stability might be preferable to the uncertainties of starting afresh. In particular, the risks of industry capture might be reduced, as monetary policy makers’ standing in the community does not depend on bankers.

I’m not even persuaded by that final sentence – too many central bank policymakers have seen their post-central bank opportunities being among the bankers (just among Governors, Glenn Stevens, Ian Macfarlane, Ben Bernanke).     Tucker’s argument appears to be made in the UK context, where the prudential functions were split out of the Bank of England after 1997, into a new standalone prudential agency.  Perhaps the FSA never had the prestige of the Bank of England but it isn’t obvious that the problems the UK ran into were a reflection of lack of prestige or legacy legitimacy.  Political emphasis on promoting the financial sector was a significant part of the story.  And the Bank of England’s own analysis of the emerging macro risks didn’t exactly cover the institution in glory.

But, perhaps more importantly, there are other case studies.  The Reserve Bank of Australia had been around for decades when APRA was split out of it. It looks to have been a pretty successful split, and it isn’t at all obvious now that the RBA enjoys any greater authority or respect in its areas of responsibility than APRA does in its.   And, given the feedback on the Reserve Bank of New Zealand’s regulatory stewardship, I don’t suppose anyone would want to mount a serious argument here to leave the functions together because the Reserve Bank’s reputation and authority stands so high.

As it happens, Tucker isn’t too keen on New Zealand’s contribution to public sector management, including the notion (from the New Public Management literature of the 1980s) that one function per agency helps to enhance accountability.

In the UK, I suspect that NPM was a subtle (and baleful) influence on the 1997 decision to transfer prudential supervision away from the Bank of England.  That mattered beyond Britain’s shores. Given London’s position as a global financial centre and given that various other countries, including China and Korea, followed the UK, at least in some cases probably encouraged by the IMF, the UK contrived to put the world onto a false, even delusional, path,  Administrative fashions come unstuck eventually, and this one did spectacularly.

Methinks he doth protest too much  (even as he goes on to note that there are hazards in having the functions together).

He highlights one issue which I hadn’t given much thought to, arguing that if the central bank is to be responsible for both monetary policy and prudential supervision it needs to have the same degree of autonomy in each function.  He argues that if the central bank has less authority in supervisory areas than in monetary policy, it could provide a wedge through which politicians could exert pressure on the Bank over monetary policy.  I’m not so sure that is right or that, realistically, it is that important an issue especially if (as he insists) each function has its own statutory committee, and its own direct accountability.

But if there is something in what Tucker says, it would reinforce my doubts about keeping the functions together.  At present, in the New Zealand system the PTA is set every five years (and central bank budgets too), and beyond that there is no routine ministerial involvement in monetary policy.  On the regulatory side, plenty of powers are reserved to ministers (eg around failure management, AML, the rules of non-bank deposit takers, disclosure rules for banks) and as I’ve argued here before the Reserve Bank still has too much discretionary policymaking power over banks (LVR limits, with significant distributional consequences are a good example –  and one Tucker seems to have quite a lot of reservations about).  As I read him, he would favour delegating more policy power to the central bank in the supervisory/regulatory area.  Personally, I think there is a good case for giving the regulator less power, and for a clearer delineation between setting the rules of the game (politicians) and implementing them (independent agency).  And keeping the functions in separate institutions will make for stronger effective accountability –  a key theme of Tucker’s –  than two or three committees with the Governor and his Deputy sitting on all of them.  You can only fire – or not reappoint –  a Governor once.

One of reasons Tucker worries about differing degrees of independence is that it would not ‘be conducive to successful institution-building’, citing the way in which the Greenspn Fed looked down on supervisors as a “lower form of life”.   Again, I’m not sure I fully buy the argument –  it is more about the priorities and beliefs of the people at the top than about formal statutory remits –  but as both Geof Mortlock and I have argued in the New Zealand context, standalone agencies helps enable the creation of cultures of excellence in both institutions.   And even Tucker recognises that culture is one of the challenges to a multi-function central bank, even if both functions have equal statutory importance.

In many cases, these aren’t open and shut issues.  There are different models around the world, although on my reading in most advanced countries –  and especially most small ones –  structural separation is the route chosen.  It is far from obvious that the new British model is better than the old (only the next crisis is likely to test that), even if appropriately some issues have been clarified and powers refined out of the 2008/09 experience.  But in the New Zealand context, most of the arguments now line up pretty clearly in favour of structural separation, and the creation of a new standalone prudential regulatory agency, with powers, personnel, and governance/accountability structures specifically fit for purposes, rather than shoehorned into an institution designed primarily for a monetary policy role.