In response to the sanitary crisis, extraordinary operations are under way in most countries (Baldwin and Weder di Mauro 2020). Exceptionally large, often open-ended fiscal support programmes have been launched and they are being coupled with exceptionally large purchases of government bonds. In the UK, the Treasury and the Bank of England have just announced the temporary reactivation of a scheme that makes it possible for the central bank to finance public spending directly.
These developments have raised fears that large-scale monetisation will result in a major inflation episode. Yet, other commentators would like the central banks to do even more and embark on some form of ‘helicopter money’ (e.g. Galí 2020).
This column is an attempt to clarify what we see as a confused discussion.
Let us start with the easy part. Governments everywhere are channelling funds to companies and households to protect them from the fallout of the economic contraction. In a way, they are practicing what the proponents of helicopter money asked for – but in a much more targeted way than anything central banks could ever do. As a result (and because of the drop in government revenues), fiscal deficits are exploding. At the same time, central banks have initiated new, large-scale government bond purchase programmes. The question is no longer whether monetary institutions will embark on direct transfers, as supporters of helicopter money had asked for, but whether we are seeing in effect the equivalent – namely, large-scale monetisation of the deficits – and if so, what the future implications will be.
What is monetisation?
Monetisation is an ambiguous concept. Evidently, not all central bank purchases of government bonds qualify as such. In the US, the Fed buys and sells government bonds all the time so as to achieve an interest rate consistent with its mandate of low inflation and full employment. In the euro area, the traditional modus operandi of the ECB was to repo government bonds, which equally affects the market equilibrium. The influence of centrals banks on the government bond market has been magnified since they have embarked on quantitative easing (QE). Their aim has been to broaden the set of interest rates that they are able to influence and thereby to flatten the yield curve, even when the policy rate is at, or below, zero. Sustained, large-scale government bonds purchases have become part of the toolbox of central banks, irrespective of the fiscal policy stance.
So, worries cannot be about the principle of central banks buying government bonds. They must be about them buying too many of them and for the wrong reasons – what one might call excess monetisation, motivated by public finance sustainability objectives rather than price or macroeconomic stability objectives.
What would then be the consequences? To think about the potential effects of excess monetisation, it is useful to start with a simple proposition.
To a first approximation, when interest rates are equal to zero, the purchase of bonds by the central bank in exchange for money – that is, the degree to which public debt is monetised – does not affect public debt dynamics. The reason is simple: it just replaces a zero interest rate asset, called debt, by another one, called money. This is true whether none of the deficit, some of the deficit, or all the deficit is financed by issuing money.
If this were the end of the argument, it would be hard to see why central banks would ever embark on such monetisation. And, indeed, the proposition must be refined in at least three ways.
First, the eventual impact of central bank purchases of government bonds depends on what will happen in the future when economic activity and inflation are such that the central bank will want to increase interest rates. Monetisation today may affect expectations of what will happen then.
Second, when there is one central bank, many national treasuries, and different rates for different sovereign bonds (as in the case of the euro area), monetisation does affect the distribution of risk across countries.
Third, when markets becoming dysfunctional, or become potentially subject to multiple equilibria, monetisation – or even the threat of monetisation – can improve market functioning and avoid the convergence of expectations on ‘bad equilibria’.
Monetisation and future central bank behaviour
Our earlier proposition was that, so long as interest rates are close to zero, whether the liabilities of the consolidated government are debt or money does not matter. But what about the point in the future when economic activity warrants an increase in the monetary policy rate?1 The central bank then has two options.
The first option is to pay interest on money – the way, for example, the Fed did before this crisis by paying a positive interest on excess reserves held by banks. The consolidated government has now two types of debt: regular debt and interest-paying money. Neglecting the impact of term premia (i.e. the effects of QE if the central bank buys long maturity bonds), the total interest rate burden is the same whatever the composition of the debt is between the two.
The second option is to keep the interest rate at zero. If, however, the economic situation warrants a positive interest rate, keeping it at zero will lead to overheating, and eventually to higher inflation. One of the implications of higher inflation will be a decrease in the real value of nominal debt, alleviating the debt burden.
What matters therefore is what the central bank, which may have a large balance sheet by then, will do when it needs to increase interest rates to achieve its mandate. If monetisation today is a signal that it will keep a large balance and not pay interest rate, then indeed there are reasons to worry about inflation.
Should, then, current large-scale purchases by central banks of government securities indeed be interpreted as a signal that, when the time comes, they will not pay interest on the large money stock and thus allow for overheating, inflation, and a reduction in the real value of government debt? It is true that the larger the portfolio of government bonds held by the central bank, the stronger the effect of its policy on debt sustainability. Large purchases do increase the risk of fiscal dominance. None of the central banks has hinted, however, at such future behaviour,2 and past experience is reassuring. The Fed and the Bank of England, among others, paid interest on reserves when they increased their policy rates in 2017-2018. The ECB did not, but because of the persistently low level of inflation expectations, not because of its government bond holdings.
Should central banks be clearer, emphasise that monetary dominance will remain unchallenged and commit to not allow for inflation when the time comes? We think not. Central banks face a familiar trade-off. On the one hand, having the ability to decrease the real value of the debt if things are exceptionally bad is clearly a useful option to have. If the virus crisis lasts for long and imposes such a debt burden on governments that they cannot repay their debt, they will be bound to choose between inflation, debt restructuring, financial repression and wealth expropriation, and there is no a priori reason to pretend that they must rule out inflation. But, on the other hand, fuelling the anticipation by investors that the central bank may have recourse to inflation in the future will increase nominal rates on longer maturity bonds today, and increase the cost of debt finance today.
There is no simple answer as to whether the trade-off is favourable, and remaining silent about what will be done in the future may indeed be the best policy today.
Monetisation in the euro area: The basics
So far, we have assumed that there is only one government and one central bank. What about monetisation by the ECB, in a common currency zone where interest rates on sovereign bonds differ?
Again, we can think of monetisation in this case as governments sending checks to households, issuing bonds to finance them, and the bonds being purchased by the ECB in exchange for euros.
Assume that the euro area consists of just two countries: a low-debt country that issues safe debt, and a high-debt country whose bonds carry a positive premium, reflecting the perception by investors of a (small) probability of default. Assume also that the safe rate, the rate on the low debt is equal to zero, and the rate on the high debt is higher, and therefore positive.
Now suppose that both governments run deficits and issue bonds, and that the ECB buys the bonds in exchange for euros, thus increasing central bank money. From the point of view of the consolidated government of the euro area (that is, putting together all the treasuries and the ECB), this is just an internal transfer of risk from the holders of securities issued by the high debt country to the shareholders of the ECB – ultimately national governments – with no implication for the total debt held by the public. But now, there is an implicit transfer of risk across euro members. Thus, monetisation in this case has an effect: it leads to some risk sharing across euro members.3
Whether or not this is the best way to achieve some risk sharing among euro area members is questionable. Other mechanisms (expenditure sharing, a dedicated credit line) would help lower the burden on the ECB and alleviate fears of monetization. Mutualisation is a political choice and it is advisable to practice it in a transparent way.
Monetisation in the euro area: The good and bad equilibria
If monetisation has no obvious implication for debt dynamics, and risk sharing is not its main purpose, why has the ECB announced a large purchase programme, the “pandemic emergency purchase programme” (PEPP), with an envelope of €750 billion, which allows it to buy sovereign bonds without necessarily adhering to the capital key? The answer is multiple equilibria and disrupted markets.
Sovereign bond markets are potentially subject to multiple equilibria. At a low interest rate, the probability that the debt is sustainable is high, justifying the low rate. Think of this as the good equilibrium. But there may well be another one, in which investors get worried, ask for a higher premium, increase debt service, and in so doing make their worries self-fulfilling and make debt unsustainable. Call it the bad equilibrium. Multiple equilibria can emerge nearly at any time, but they are more likely in the current circumstances when investors are edgy.4
In this case, the central bank can play a crucial role: by committing to buy if the investors sell, it can eliminate the bad equilibrium. One way to do this is to do what the Bank of Japan has been doing, which is to commit to maintaining a given low interest rate, a strategy called yield curve control. The ECB mandate does not allow it to adopt such a strategy, but it has made clear that, were rates to increase beyond what is justified by fundamentals, it will intervene and buy the bonds that investors are selling. Standing ready to purchase bonds in this context is not an attempt to monetise the debt. Indeed, if the strategy is successful, it actually deters investors from selling, and may achieve its purpose with little or no intervention, little or no monetization, and little or no cost to the other governments.5 In this case, the insurance that it provides to the high-debt country has no cost to the low-debt country. It may even benefit it by preventing a debt crisis and its cross-border spillovers.
This role is not limited to the euro area or to government bonds. Markets everywhere can become dysfunctional. Some investors have to sell to get liquidity. Others may not have the liquidity to take the other side. Or there can be multiple equilibria. In recent years, and again in this crisis, we have seen examples of both. When markets become dysfunctional, the central bank can take the other side until investors return, or others come in.
So far, there is no evidence that central banks have given up, or are preparing to give up, on their price stability mandate. It may eventually happen, if the fiscal cost of the crisis proves to be unbearable, but the size of the current public bond purchases should not be regarded as indicative of future excess monetisation.
In the specific case of the euro area, the ECB’s bond-buying purchase programme can evidently serve as a channel for mutualising the cost of the crisis. This is in part by default: we see good reasons why part of the burden of fighting the pandemic should be mutualised among EU members, but it would be more appropriate to do so in a more transparent way through explicit budgetary and financial channels.
So far, no agreement has been reached on such schemes, and this is unfortunate. But this is no reason to interpret ECB actions as mainly distributional. The PEPP is not a hidden budgetary mechanism. At a time when investors are prone to nervousness, its main purpose is to prevent the convergence of expectations on a bad, self-fulfilling crisis equilibrium. Such action serves the interest of all the members of the eurozone.
In short, there are obviously some reasons to worry, but we see no reason to panic. The central banks are doing the right thing. Their actions are sustainable. And they have not tied their hands to the inflation mast.
Baldwin, R and B Weder di Mauro (2020), Mitigating the COVID Economic Crisis: Act Fast and Do Whatever It Takes, a VoxEU.org eBook, CEPR Press.
Galí, J (2020), "Helicopter money: The time is now", voxEU.org, 17 March.
1 This may be far in the future, but it will happen one day.
2 The joint press release by the BoE and HM Treasury of 9 April explicitly states that any use of the direct financing scheme will be short-term and temporary.
3. As the interest paid on the bonds held by the ECB is redistributed to its shareholders, it also involves a transfer from the high debt country to the low debt country, which can be regarded as a remuneration for the risk transfer.
4 We do not discuss here the risk of redenomination of the public debt following an exit from the euro area. It only strengthens the argument.
5 To be clear, this is not a strategy without risks. Distinguishing between the emergence of a bad equilibrium, and a justified increase in the rate in the good equilibrium is not easy, and the central bank may find itself taking credit risk.