Macroeconomic Policy Outlook Q3 2021

With the economic outlook improving, there has been much debate about when the Bank of England should start tightening policy. But the Bank also faces a key decision about how to tighten. This is important because a successful strategy for withdrawing support is key to delivering a strong recovery. So in this edition of our regular ‘Macro Policy Outlook’ we focus on how the Bank should tighten, and how that should be coordinated with fiscal policy. Here there are two key questions. First, what is the right tool for active tightening – policy rates or reducing Quantitative Easing? And second, should that tightening come before or after fiscal tightening?  

On the first question, historically the Bank has said it would tighten by first raising its policy rate to around 1.5 per cent, and then do further tightening by unwinding QE (i.e. selling assets). But, since the arrival of Andrew Bailey, it has become clear that the Bank are revisiting that strategy, with the new Governor dropping heavy hints that he would like to see asset sales begin long before interest rates return to 1.5 per cent. Our view is that revising the Bank’s approach to tightening is desirable and should include earlier asset sales than previously envisaged. A renewed approach however should combine a gradual and predictable asset sales programme with clarity that changes in policy rates are the key marginal tool for tightening policy. This is because the impact of policy rates is less uncertain and more nimble if circumstances change. Large short-term changes in asset sales also risk undermining the crucial role that the Bank is playing in supporting fiscal policy. Given these sales would be designed to impart little or no discernible tightening, they should start before rate rises, but only when it is clear that a sustained recovery is in progress.  

The second question is important because the Chancellor has already made it clear that he also has an eye on tightening policy. Coordination is crucial in order to avoid withdrawing collective support too quickly as happened in the aftermath of the financial crisis. Prior to that crisis, fiscal policy tended to tighten before monetary policy; safe in the knowledge that rate cuts could be used to respond if needed. But with policy rates mired at the lower bound, it is important that monetary policy tightens first, building policy space in case the outlook worsens.

With a recovery from the Covid-19 crisis seemingly well under way, attention has shifted towards tightening macroeconomic policy. These flames have been fanned by rising inflation and recent hints that some on the Bank of England’s Monetary Policy Committee (MPC) may be close to following Andy Haldane in voting to curtail Quantitative Easing (QE). So, is the time right for the Bank to start tightening? As we have set out in our previous work, the obvious answer is ‘no’. Here our analysis points to the rise in inflation proving temporary, and the labour market recovery still having much further to run. On top of that, the Delta variant seems to have led the recovery to slow down.

But there is a second question that has received less attention than when monetary policy tightening should take place: how it should be implemented. That question is the focus of this edition of our regular Macro Policy Outlook, along with how tightening should be coordinated with fiscal policy. These are important issues because a successful exit strategy will be key in delivering a strong recovery and building policy space ahead of the next downturn.

The Bank’s choice of how to tighten boils down to a decision about which of its main policy tools should be used for active tightening

The key choice facing the Bank in deciding its approach to tightening is how it should use (and crucially sequence) its two main policy levers: raising its main policy rate (Bank Rate) and reducing the stock of assets built up via QE.

Historically, the Bank’s answer to this has been that it will raise rates first before selling assets. Back when Mervyn King was Governor, the Bank committed to raising Bank Rate until it reached a level from which it could be ‘cut materially’ and then sell assets (referred to as Quantitative Tightening, or QT). Under Mark Carney, this approach was refined, and the threshold for selling assets was defined as Bank Rate at ‘around 1.5 per cent’.

But under Andrew Bailey the thinking appears to have changed radically. The Governor is on record as saying he thought that “it may be better to consider adjusting the level of [QE] first without waiting to raise interest rates on a sustained basis”. This has prompted a review of this approach by Bank of England staff (see the final page of February’s MPC minutes) which is set to report back imminently. So below we weigh up the arguments for moving towards tightening through QT.

There is a strong case for selling assets

Andrew Bailey’s apparent discomfort with the amount of QE is not hard to understand. The expansion in the Bank’s balance sheet has been massive – with an £895 billion in asset purchases since QE started in 2009. The policy of waiting for significant rate rises before commencing asset sales has meant no progress in reducing that stock between the financial crisis ending and the pandemic arriving. As well as creating political pressure, QE can also increase the chances of financial instability, for example by pushing financial firms to take more risk.

Partly reflecting these economic considerations, which should be front of mind for the Bank, there has also been increased political debate about the scale of QE. Buying up large swathes of government debt has led some to accuse the Bank of covertly funding the Government by printing money, an accusation not helped by the coincidence in the size of QE purchases and the increase in gilt issuance (Figure 1). This was evident in our own survey of MPs which found a low level of support for expanding QE in 2018. And, more recently a House of Lords report called for the Bank to unwind QE.

Figure 1: Since the start of the pandemic the Bank of England’s gilt purchases have been a similar size to issuance

But Bank Rate should remain the primary marginal tool for tightening

While there are definitely reasons for thinking that it is a good idea to revise the Bank’s approach to tightening so that asset sales begin earlier, that is not the same as saying that it’s the right tool for actively tightening policy.

There are two key advantages of using Bank Rate. First, changing Bank Rate is quicker and easier than adjusting the stock of assets. It takes months for a programme of asset sales to be planned and executed, whereas Bank Rate can be changed almost instantaneously. Second, a gradual programme of asset sales, rather than big short-term changes, is desirable to avoid excessive tightening. In the US, this risk materialised in 2013 when the Fed announced that it would start to reduce (or taper) the pace of its QE purchases. This gave rise to a very sharp rise in market interest rates – the ‘taper tantrum’ – with US Treasury yields increasing by around 1 percentage point. With the benefit of hindsight, it is clear that markets interpreted the announcement as a signal that increases in policy rates were likely to come sooner than previously thought. By contrast, in 2017-2018 the Fed very gradually reduced the stock of QE by about a sixth without a discernible impact on rates (in fact yields fell – Figure 2).

The key takeaway from these episodes is that small, gradual changes in the stock of QE, that are clearly communicated, can have little or no impact on financial conditions. By contrast, abrupt changes in policy that can be interpreted as holding information about future monetary policy can lead to a larger-than-expected tightening.

Figure 2: The Fed reduced QE with little or no discernible impact on longer-term interest rates

The impact on fiscal policy should be considered

With interest rates at their effective lower bound, and given the unique nature of the pandemic, monetary policy has played a supporting role during this crisis. Framed in this way, QE has been key to facilitating fiscal policy makers to deliver unprecedented levels of support by contributing to continued low borrowing costs. If the monetary policy tightening forced fiscal policy to follow suit – for example because borrowing costs rose sharply – that could lead to excessive tightening overall and lasting economic weakness.

One argument that may be used in favour of QT in this context is that the Bank’s purchases of Government debt have made overall public sector borrowing costs more sensitive to Bank Rate. This is because the expansion of QE effectively swaps a long-term fixed-rate liability (gilts) for one that moves with Bank Rate (central bank reserves). This change is illustrated in Figure 3 which shows OBR estimates of the sensitivity of debt interest payments to changes in interest rates and inflation.

Figure 3: OBR estimates show that mechanically the public finances are currently more sensitive to changes in Bank Rate

There is a significant debate about whether the current arrangements for remunerating reserves should remain in place (the way the Bank implements the MPC’s decisions about Bank Rate in financial markets is by remunerating reserves at that rate, but other approaches have been taken by the Bank in the past and are currently in operation by other central banks). But even leaving those decisions to one side, the impact on fiscal policy of significant QT should not be discounted by the focus on the immediate effect of a rate rise.

Large changes in asset sale volumes could make it harder for the Government to issue debt. Here the concern is that, because asset sales could change the balance of supply and demand in the gilt market, the Government would find it harder to issue debt. Here it is obviously difficult to assess the likelihood that market conditions change significantly. But the ‘taper tantrum’ does point to the need for a gradual and telegraphed approach to changing QE policies.

Our view is that the Bank should change its approach to tightening and adopt a ‘Fed-like’ strategy in which marginal tightening is achieved through changes in Bank Rate preceded by gradual asset sales

Overall, then, our view is that there is a strong case for using Bank Rate as the key marginal tool of tightening. It is quicker to implement – so faster to reverse if conditions change – and the impact is less uncertain. It is also likely that only small changes will be needed. This is because the forces pushing down on longer-term interest rates (including: demographics, inequality and debt) have intensified during the pandemic (for a discussion see Gertjan Vlieghe’s recent speech). This means that even small increases in Bank Rate should be sufficient to deliver the desired tightening.

There are, however, compelling reasons for wanting to reduce QE over time. Here the experience of the Fed suggests that accompanying small increases in Bank Rate with gradual, telegraphed reductions in the stock of QE should have little or no discernible impact on financial conditions. For such gradual reductions to significantly reduce the stock of assets held by the Bank, and so create policy space ahead of the next downturn, there is an argument for starting asset sales prior to raising Bank Rate. Such a strategy would be very similar to that adopted by the Federal Reserve in 2017.

The Bank’s policy must be coordinated with fiscal policy as both seek to tighten

With the Chancellor clearly prioritising tightening – having already announced £29 billion of tax rises at the Budget – how should fiscal and monetary policy coordinate to make sure enough support is provided to the economy?

Following past recessions, fiscal policy has tended to tighten first, leaving monetary policy to adjust as necessary given the outlook (Figure 4). Prior to the financial crisis, such an approach made sense: if additional support for the economy was needed as fiscal policy tightened, policy rates could be cut to deliver that. But, since policy rates hit their effective lower bound, that is no longer possible. This means that tightening fiscal policy risks weakening the economy without an equilibrating force to offset it. Following Covid-19, then, it is vital that monetary policy tightens first.

Figure 4: In the aftermath of recent recessions, fiscal policy has tightened first

Overall, the scale of the risk that policy tightens too quickly and derails the recovery means that a sensible exit strategy from the Bank is not sufficient, and there is a case for coordination to be reinforced by changes to the macroeconomic policy framework. A strategy under which the Bank combines small increases in Bank Rate preceded by gradual and telegraphed asset sales should help minimise these risks, as should waiting until the Bank has started raising rates before tightening fiscal policy. But one way to enhance this strategy would be to make supporting changes to the framework for policy. The most obvious step is a set of fiscal rules that make it clear tightening will not start until a full recovery is secured (in line with our previous proposal). This could be reinforced by steps to increase the inflation target slightly, either permanently (as we have previously called for) or temporarily (as we have seen in the US and euro area).


1 Useful conversations with Tony Yates are gratefully acknowledged (although the views in this note are the responsibility of the authors), as is input from Torsten Bell and Adam Corlett.