Paying for It: Part III - Central Banks & Debt Cancellation

There is a seemingly attractive way of financing the expected massive expansion of government debt: get the central bank to buy it, and subsequently simply cancel it.

The first thing to acknowledge is that this looks like an attractive, and possible, option. After all, around the world central banks no longer have qualms about quantitative easing as a response to economic or financial discomfort, they already own a large amount of government debt, and the fears of excessive money creation generating inflation have so far proved groundless. Since there are no 'lessons' to be learned from pandemics about the economic and financial systems prior to the virus' arrival, residual worries about moral hazard aside, why not grasp at this 'easy way out'.

But let's look a little more closely at the size and implications central bank holdings of public debt prior to Covid 19, and ask whether a large new development along similar lines is desirable.

The world's main central banks already hold an unhealthy proportion of government debt.

• In the US, the Federal Reserve holds $3.34tr of public debt , which is equivalent to just under 20% of the $16.89tr in publicly-held marketable holdings.
• In the Eurozone, the ECB currently holds Eu2.925tr worth of public debt instruments, out of a total outstanding of Eu8.305tr. In other words, it holds about 35% of the outstanding issue.
• In Japan, the BOJ holds Y462tr of government debt out of a total Yn887.2tr in outstanding public bonds - an amount equivalent to 46.8% of the total.

This raises the question: how much is too much? By now it is a distant memory, but until relatively recently, central banks intervened to influence only the very short end of the curve. They did this not only, and not principally, because it was possible and cheap, but also because they realized that it was necessary for the market to set the 'risk free rate' at the longer end of the curve in order to price other risk assets. The assumption was that the market's ability to price risk was the sine qua non of a functioning capital market.

At what point does the central bank's dominant position in the government bond marketamount to a de facto financial coup against the capital market? ('We had to destroy the market to save it'). Certainly in Japan, it seems quite clear that the Bank of Japan has decided that, in the absence of a properly functioning financial system, there is little limit on its role in allocating capital - hence its willingness not only to corner the JGB market, but also to extend its rule to corporate bond issues and ETFs too.

The ECB has not yet fully embraced BOJ's de facto financial ambitions, but with 35% of the public debt paper on its balance sheet, and its long-term refinancings operations for various categories of commercial bank debt, not to mention its direct purchase of corporate bonds and commercial paper, the direction of travel is very clearly Japanese.

And so to the US, where in late March, the Federal Reserve reported it would buy an unlimited amount of Treasureis, commercial paper, corporate bonds, munis, agency bonds, and mortgage backed securivies in unlimited amounts in order to ward of a credit crunch.

Broad-based asset purchases by the Federal Reserves Acting to alleviate immediate liquidity crunches may be unavoidable in times of crisis. But the lesson from Japan and now the Eurozone is that those temporarly 'emergency' measures have tended to metastatize into central bank control of broad financial markets, in which previously functioning and misfunctioning parts of the financial system find themselves replaced by de facto public sector financial administration. There is a description of that state of affairs which does not include the words 'free market'. In practice the absence of a free market has tended towards economic torpor.

I conclude that prior to coronavirus both central banks in both Japan and the Eurozone already held an unhealthy dominance in public debt markets, and that in both cases, since that dominance had undermined the rational pricing of other financial assets, the central banks were being lured into ever-wider swathes of capital allocation. This had not yet happened in the US. However, the dramatic fiscal impact of coronavirus will surely result in an even greater concentration of holdings of public sector debt in central banks throughout the world, which will necessarily further exaccerbate all the problems already visible.

Debt Cancellation

The strategic need to avoid yet further central bank dominance of government debt markets is a further reason why cancellation of government debt held by central banks looks appealling. It is not without its problems, however.

This is not immediate obvious. After all, if the central bank is seen as a part of the government, then if the government issues debt which is 'bought' by the central bank, and if the govenrment pays interest on those bonds, which the central bank subsequently returns to government by way of a dividend, then what's the problem?

The problem, broadly state (for reasons which will become obvious) is that the interactions between the government and the central bank are not the only one which are involved: at some point the rest of the financial system will necessarily get involved. Let's take even the most direct financing route possible, in which the government issues bonds which the central bank buys by issuing credits to its member banks, those member banks deposit those credits back to the central bank which then uses those increased deposits to buy the bonds.

The point to realize is that there are at least three relevant moving parts to this transaction:

1. The government gets cash from the sale of the bonds

2. The central bank gets an asset on its balance sheet (the bonds).

3. The central bank also gets new liabilities on its balance sheet (the increased 'excess deposits' from banks).

(The cashflows and balance sheet implications will get more complicated depending on precisely how the bonds are issued).

Notice that the central banks balance sheet has expanded in both assets and liabilities. Two things follow from this: first, the inflationary consequences of this sort of public debt monetization are negligible precisely because the banking system is unable or unwilling to lend out those new deposits. In the early stages of Japan's quantitative easing, this non-inflationary aspect was made absolutely explicit when BOJ targetted the level of commercial bank deposits in the central bank as a measurement of the success of quantitative easing! (And then seemed surprised that inflation did not emerge.)

If, however, these conditions did not apply - for example, if commercial banks found some genuine end-use for these 'excess' reserves - then the quantitative easing might indeed prove inflationary. Just as there are times when Ricardian equivalence genuinely does kick in to offset the stimulative impact of a sharp rise in fiscal deficits, so there are also times when financial dysfunction and/or collapsed confidence means that massive monetary easing is not inflationary. But neither Ricardian equivalence or the lack of inflationary consequences of very steep monetary easing represents the usual state of affairs.

The second thing that follows is that in these circumstances, simply cancelling the central bank's holdings of government debt would immediately leave a very large hole in the asset side of central bank's balance sheet. Several things could happen at that stage. First - nothing could happen, with the financial system conspiring to pretend that the central bank was not in fact running with a ruined balance sheet. This would require a consensual hallucination to be maintained over decades - possible but hazardous. Second, the central bank could quite literally print money to inject itself with assets. Third, the central bank could decide it would 'work its way out' of its balance sheet problems by widening the spread it offered between loans and deposits. This is what bankrupt central banks have done elsewhere (for example,the Philippines). And fourth, it could accept a similar-sized shrinkage of its liabilities, which in practice could be achieved only by a fall in the reserves of commercial banks which are deposited with the central bank. The problem with that, of course, is that it would genuinely represent a very severe tightening of monetary conditions which would ramify through the banking system.

The balance sheet problem may perhaps be avoided. A smarter economist than myself has suggested that, rather than debt cancellation, the government could buy back the central bank's holdings of its debt by issuing a similarly-sized IOU, but carrying a zero interest rate and with no maturity date. This could be used to paper over the hole in the central banks' balance sheet, but only at the cost of ensuring that, since the IOU could never be called, realized or sold, there would be a permanent and irrevocable expansion in the central banks' overall balance sheet. Does this matter? I don't know.

There is one other immediate point which perhaps ought to have been made right at the start: the real problem with cancellation of government debt is that . . . in a way it is a default.