The world’s central banks led by the four heavyweights of the US Federal Reserve, the European Central Bank, the Bank of England and the Bank of Japan have done their utmost to support economies battered by Covid-19. They cut policy rates (close) to (or kept them below) zero, relaunched quantitative easing programs to support huge fiscal easing, and set up sizable lending programs for banks and even businesses.
As a result, the balance sheets of many central banks have ballooned to historic highs. But as vaccinations have brightened the public health outlook, financial markets’ attention has now shifted to the process of exiting (some of) these stimulus measures.
While the Bank of England and Bank of Canada have already announced reductions in the pace of their QE purchases, and the Reserve Bank of New Zealand even stopped buying bonds in July, other central banks like the Fed and ECB have not yet specified how or when they will end their QE-induced balance sheet expansion.
Central bank assets as a percentage of their relevant nation’s GDP

Source: Bloomberg, central bank websites, Robeco, August 2021.
GFC offers a clue
The Fed’s normalization process after the Global Financial Crisis offers a template for what could soon be in store. In December 2013 it started to taper its USD 85 billion monthly purchases by USD 10 billion following each Fed meeting, concluding the process in October 2014 before starting to raise policy rates in 2015. It did not start reducing the size of its balance sheet until 2018.
Currently, the Fed is adding USD 80 billion in Treasuries and USD 40 billion in mortgage-backed securities to its balance sheet every month. If the Fed were to initiate a reduction in its monthly purchases of USD 15 billion in November and make a similar reduction at each of its subsequent meetings, tapering could be concluded in October 2022.
The ECB tapered its monthly net asset purchases, which were part of its 2015 Asset Purchase Program (APP) to zero between 2017-18. But due to the economic outlook worsening, in Q3 2019 it reintroduced net APP purchases, which are currently supposed to run until “shortly before” the start of rate hikes. Note that its net asset purchases under the pandemic emergency purchase program (PEPP), which was launched in March 2020, are expected to be phased out by the middle of 2022.

Expected Returns 2023-2027
Our latest 5-year outlook
Read all articlesTotal targets vs. monthly amounts
The Bank of England’s purchase programs target a total amount of QE stock to be held. This is in contrast to the weekly or monthly target amounts without an end date set at other central banks. The current target for the BoE’s QE asset portfolio is GBP 895 billion. Our base case is for this target to be reached by the end of this year, which means that net monthly purchases – currently GBP 14 billion a month – would end in December 2021.
The Bank of Japan’s QE program started as a reflationary and a foreign exchange policy tool in 2010 but has since transformed into a supplementary tool to the yield curve control policy (YCC) framework. The bonds it has purchased have mainly had maturities of under 10 years, making the policy less distortionary for long-end bonds. In our base case, we expect the Bank of Japan to continue with its current policy mix of YCC and QE over the coming years.
Much like the Federal Reserve and Bank of England, most other G10 central banks that have also engaged in sovereign QE since the start of the pandemic are expected to stop increasing their holdings in the course of 2022. The exception may be the Reserve Bank of Australia given the role of QE in its own YCC policy.
No imminent rate rises among the big four
Regarding rate rises, the markets are generally not pricing in a first rate hike until H2 2022 for the Federal Reserve and Bank of England, late 2023 for the ECB, and not at all within the next five years for the Bank of Japan. Among these central banks, the Bank of England will probably be the first to hike rates. Our base case assumes an initial rate hike by the BoE in Q4 2022. The Fed is expected to follow in Q2 2023.
Perhaps more importantly for bond yields is where policy rates are going to peak in any upcoming rate normalization cycle. Two things matter in this respect. The first is the cyclical inflation outlook and the associated need for policy changes. The second is the perceived ’neutral’ rate based on inflation expectations and real neutral rate estimates implied by markets or calculated by official institutions. These are shown in the table below.
The range of neutral policy rate estimates for selected central banks

Source: Bloomberg, IMF, Fed, ECB, BoJ, BoE, August 2021.
In our central scenario, policy rate peaks in the upcoming normalization cycles are likely to be in the lower part of the range of neutral policy rate estimates for the Federal Reserve and Bank of England and even below the estimated range for the ECB. In this scenario, consumer price inflation pressures do not spiral out of control, so do not warrant a contractionary policy rate stance, where actual policy rates need to exceed the nominal neutral rate.
Any new crisis
So, what might central banks do if a new crisis were to emerge over the next five years? Forward guidance, large-scale asset purchases, generous lending programs and macroprudential adjustments have all became ’normal’ components of the central bank toolbox.
In any new recession, those that were able to hike rates in the run-up to it will quickly cut them back to zero. Some, including the Reserve Bank of New Zealand and Bank of England, will probably follow the examples of the ECB, Bank of Japan and Swiss National Bank and cut policy rates (slightly) below zero.
And governments will yet again resort to running large budget deficits, supported by central bank bond purchase programs. We would also expect renewed and/or bolder private sector debt purchases – perhaps with broader adoption of equity ETF purchases like the Bank of Japan undertook – and even larger and more generous loan programs to banks and (via banks) to the non-financial private sector
It’s time to unwind
In conclusion, our central scenario is that most central banks will aim to unwind many of the policies they enacted during the pandemic, but that they will struggle to do so in full.
Assuming that fears of an inflationary regime change recede, our base case is that QE tapering will end next year (except for the Bank of Japan and ECB), most central banks start hiking rates in 2023. For most central banks, reducing their balance sheet holdings will only be on the agenda from late 2023 onwards, if at all.
