Foreword by Andrew Bailey
Research is crucial to what any central bank does. It forms a critical part of the analytical apparatus that guides policymakers’ decisions. This document sets out the key new directions for research at the Bank of England in the coming years. The Bank has monetary policy, macroprudential policy and microprudential policy under one roof. We welcome further research to inform our decisions in all three of these policy areas.
This agenda builds on our original 2015 One Bank Research Agenda. The world and the associated research challenges have changed appreciably, but our approach here is evolution rather than revolution. We seek to build on our previous research and strengthen several key themes within it, both in our philosophy towards research and the set of questions laid out.
By making our research agenda public, we want to facilitate collaboration with external researchers. We want to tap the rich source of expertise that lies beyond the walls of the Bank of England in academia and the broader research community. We especially welcome collaboration with and insights from fields outside of economics, including psychology, data science, epidemiology, computer science, physics and many more besides.
Our approach to research is strongly pluralistic. A common thread running through this document is that research agendas are guided not just by the evolution of the economic and financial sphere, but also by a belief that traditional tools cannot always answer every question. Climate change and the recent Covid-19 pandemic provide powerful examples of new forces that reshape the research agenda. New, heterodox and traditional approaches are all key parts of our analytical toolkit.
The issues laid out here are not the sum total of all research that will take place in the Bank. Here we focus on the key questions which we think define the new directions for research for central banks. These sit alongside the more traditional central banking research avenues, which continue to be important. This agenda shapes not only the research that we do but also the data we seek to acquire and the intellectual capital we seek to invest in.
Whilst the questions here are those of the Bank as a whole, the answers provided by individual working papers are those of the individual authors. Our commitment to independence, openness and diversity of thought remains as strong as ever. Our research papers will continue to provide challenge to internal and external orthodoxies, as well as to each other in the years to come.
The Monetary Toolkit
Keywords: central bank balance sheets, unconventional monetary policy, transmission mechanism, reserve demand, zero lower bound, neutral rates, heterogeneity, communication, expectations formation.
Contemporary monetary policy goes well beyond simply moving an interest rate up and down. Policymakers now have a much wider array of levers at their disposal. This throws up important questions about how these new instruments work, how they interact with each other and the policy rate, and which ones should be deployed when. What is the appropriate perimeter for central bank asset purchases? Do central banks have the right set of tools to provide liquidity support beyond the banking sector, and how do these shape the incentives of financial market participants? More analysis is needed to understand the efficacy, consequences and transmission mechanisms of each individual unconventional policy instrument.
Implementing unconventional policy efficiently requires research into the optimal structure of market operations, including collateral criteria, lending facilities and auction design. This calls for a better understanding of how economic conditions and market structures affect demand for reserves. Further work is needed on how policy design can support interest rate pass-through around the lower bound. More research is also needed into the operation of the repo market as the primary transmitter of monetary policy.
The determinants and sources of demand for central bank money underpin the monetary transmission mechanism. Research into that demand and the frictions that encourage banks to hold low yielding assets is needed, including how they interact with broader conditions.
The deep economic disruptions caused by Covid-19 have renewed focus on optimal policy co-ordination. How should monetary policy and macroprudential policies be co-ordinated in response to such a shock? Relatedly, the rise in the debt burden and the use of new monetary policy instruments call for a renewed work on how fiscal and monetary policies should be co-ordinated. There is also a broader question over the appropriate scope and boundary for each policy tool.
Effective lower bound
The expansion of the monetary policy toolkit partly came about because rates have been in the vicinity of an effective lower bound. A key issue is the optimal exit strategy from that bound, and the sequencing of instruments used to unwind stimulus.
Could other measures be used to aid the exit? Helicopter money, temporary price level targeting, yield curve control and negative interest rates have been suggested in the literature but more analysis on their relative merits is needed.
Understanding when, how often and why the effective lower bound binds is crucial. Empirical estimates of the level of equilibrium real interest rates are central here. So too is a deeper understanding of how the structure of the economy, the international financial system or the policy regime determines equilibrium rates.
This also raises questions about the role, level and nature of the monetary policy target. The costs of inflation, especially as financial technology develops, need to be estimated to assess the appropriateness of the target.
Departures from a representative agent/firm world can yield important new insights. New modelling approaches are needed to understand how heterogeneity among banks, firms and households affects the transmission mechanism of monetary policy.
This calls for more work on the distributional effects of low interest rates and unconventional tools. Does heterogeneity affect the transmission mechanisms of policy measures, and does heterogeneity create additional unintended consequences of policies?
Central banks have extensively reformed their communication strategies over the past decade. But the literature examining the impact of those reforms is still relatively underdeveloped. It would also be valuable to understand how communication can be best formulated to target different audiences.
A better understanding is needed of the trade-offs between clarity of communication and diversity of views expressed. Similarly, the optimal amount of information to publish remains an underexplored area.
Communication is also intrinsically linked to expectation formation. Further examination of what determines the inflation expectations of firms, households and investors, and whether these expectations matter, would be helpful. There are many open questions over how expectation formation may affect the properties of optimal monetary policy.
Priority topics for 2021
- What is the appropriate perimeter for central bank asset purchases?
- What is the right set of tools for central banks to provide liquidity support beyond the banking sector, and how do these shape the incentives of financial market participants?
- What are the transmission mechanisms of unconventional monetary policy tools, and how effective are they?
- What are the implications of a negative central bank policy rate?
+ Other work on unconventional monetary tools and their interaction with other policy instruments.
The Open Economy
Keywords: trade, global value chains, dollar dominance, international currencies, swap lines, capital controls, deglobalisation, fragmentation, exchange rates, capital flows.
The nature of international trade is changing. A greater share of trade is in components rather than finished products and some goods cross borders many times during their production. More work is needed to understand how this growth in intermediate goods trade changes the workings of the international economy.
New forces of de-globalisation may be emerging. In the UK, Brexit could change the shape of trading relationships substantially. Globally, new technologies such as 3D printing, the emergence of risks such as pandemics, and trade wars may change the calculus of offshoring vs onshoring, and affect the overall resilience of the economy to external shocks. How large are these effects, and how might they affect the structure and properties of the economy? How might these forces affect the behaviour of the economy and asset prices?
Trade is also a potential determinant of productivity, through both a more efficient allocation of resources and through the transfer of ideas. Assessing the strength of these channels is valuable. Many of these issues apply to trade in services as well as goods. A deeper examination of the drivers of services trade would also be useful.
International monetary system
The dollar remains dominant in the international monetary system, and widely used beyond the US for trade invoicing, cross-border credit and sovereign reserves. How dollar dominance alters the transmission of shocks is a key question for research. Does this alter how policymakers should appraise and react to shocks and does this create scope for more international co-ordination?
There are also deeper questions regarding the determinants of dominant currencies. A key one is the role of financial markets, trade and other factors in determining which currency is dominant. Perhaps alternatives to the dollar are on the horizon from private issuers, co-ordinated sovereign action, or individual states. Is this desirable, how should policy respond, and what would the transition look like?
The international financial system’s safety net is currently a combination of IMF credit, swap lines and sovereign reserves. Future research should explore whether the current setup is appropriate and suggest potential reforms. Is there a need to create a deep market for state-contingent sovereign bonds? Should an international body be responsible for managing the swap line network? Further work is also needed to examine how risks transmit between countries, and how each country should deal with spillovers from crisis abroad.
The forces that drive exchange rates need re-evaluating in the light of recent experiences. Work is required on assessing whether and how exchange rates have ‘reconnected’ to macroeconomic variables. How might frictions in financial markets drive exchange rates away from canonical textbook relationships based on differences in interest rates? Given how exchange rates pass through to the macroeconomy and inflation, this opens the question of the optimal monetary policy response to exchange rate movements and whether that response depends on the underlying shock.
Evaluating the role of capital controls and other macroprudential tools to deal with capital flows will be an important research stream in the years ahead. New and richer models are needed on the determinants of capital flows, including the role of non-bank actors and whether the system is becoming more fragmented. Flows of capital also pin down the price of capital: ie the neutral real interest rate. More work is required on the global drivers of the equilibrium real interest rates.
Priority topics for 2021
- How will new technologies, Brexit or Covid-19 affect trade in goods and services? How will they affect international capital flows?
- What are the potential gains and losses from changing trading patterns, and how are they distributed?
- What are the macroeconomic and financial stability implications of de-globalisation?
+ Other work on international spillovers of central bank policies to support domestic economic recovery and to stabilise financial markets.
The Prudential Framework
Keywords: capital requirements, liquidity requirements, optimal policy, stress testing, resolution, borrower resilience, complexity, policy interactions, insurers, regulatory perimeter, funding markets.
Macroprudential policy has evolved radically and it is now a well-established part of many central banks’ policy apparatus. But more work is required to augment the analytical apparatus underpinning the day-to-day setting of macroprudential tools and to understand the inefficiencies that motivate them. Improving the modelling framework more broadly is desirable, in order to estimate costs and benefits of policy and to conduct welfare and optimal policy analysis.
Further development of stress test modelling is also needed. This comprises several elements. One is evaluating the robustness of the tests to scenario design. Another is understanding how stress tests should be used, if at all, in a situation where the financial system is already stressed. Further modelling of the interactions between different parts of the financial system in a stress and their interlinkages with the real economy is also needed. For example, borrower resilience in the corporate sector has come into focus in the recent past due to concerns over the health of the syndicated loan market and rising corporate leverage. Are there externalities from corporate borrowing, and if so how should prudential policies respond?
Capital and liquidity requirements interact with each other. How do they jointly affect the provision of credit, the cost of crises and the likelihood of runs?
The prolonged period of low rates could create new risks, raising the question of whether and how macro and microprudential policies should respond. Links also run in the other direction: analysis is needed on whether the risks in the financial system affect long-run equilibrium real interest rates.
Further research is needed to examine whether new regulations aimed at curbing excessive risk-taking behaviour at the individual level, including the remuneration rules and Senior Managers Regime, have achieved the intended aims. The role of behavioural and psychological factors underpinning risk-taking decisions needs further exploration.
With multiple policy instruments comes the question of whether the prudential policy regime is too complex. For instance, capital requirements combine multiple buffers and add-ons stacked upon a flat minimum. Could this ‘capital stack’ be simplified? The new resolution regime may have altered the trade-offs between going concern versus gone concern capital requirements. More broadly, further research is needed on any potential unintended consequences of regulations.
Research could also inform the way supervision is conducted. For example, should the extent of deposit insurance coverage influence the intensity of supervision? How should supervisors prioritise data collection? And how can the effectiveness of supervision be measured? An important research agenda is also how the operational resilience of an institution and the financial system could be monitored.
Outside the banking sector but within the Bank’s regulatory perimeter lie insurers. Further research is needed to examine the interlinkages between the insurance sector and the banking sector, and to quantify the sources of systemic risk arising from the insurance sector. The question of how to regulate insurers of different sizes and types is also a fertile area for study. What are the socially optimal solvency requirements for insurers, and how do these compare with observed ratios? Such an analysis will require an assessment of how the solvency requirements influence systemic and prudential risks, as well as competition.
There are also risks stemming from actors outside the Bank’s regulatory perimeter. These include leveraged institutions such as hedge funds, as well as mutual funds and pension funds. Further work is needed to examine how different financial institutions behave in a crisis, and whether their collective behaviour amplifies or dampens the cyclicality of financial system and the real economy. Additional work is required on whether there is a trade-off between market liquidity in normal times and resilience in times of stress.
Priority topics for 2021
- What features are needed in models that can be used to assess the costs and benefits of prudential policy responses to specific shocks?
- How can the interactions between the different parts of the financial system and the real economy be modelled to assess the economy’s response to large shocks?
- Can financial policy support long-term growth and prosperity objectives more broadly, beyond ensuring the soundness and resilience of the financial system?
+ Other work developing modelling tools for stress testing and supervision, and empirical work using the Bank’s regulatory data and other microdata.
The Future of Finance
Keywords: fintech, digital currencies, suptech, artificial intelligence, blockchain, market fragility, financial market infrastructure, productive finance, competition, efficiency.
Technological changes create a need for fresh research streams to understand whether and how they create new risks for financial markets and systems.
Financial actors are now using machine learning, artificial intelligence and other techniques to process big data to inform decision making. These technologies could improve efficiency and lower regulatory burdens but they may also throw up new risks. For example, could algorithmic trading generate self-reinforcing loops in a stress or collusive pricing? How robust are they to small perturbations in the data? Similarly, new research should explore whether cheap deposit searching could make banks more prone to runs, or whether artificial intelligence can be used to game stress tests. There is also work to be done in assessing the possible benefits of algorithms. For example, there could be scope for encouraging co-ordination on good equilibria, thereby averting downward spirals.
These technologies can also be used to inform central bank policy. Supervisors might deploy artificial intelligence and machine learning to aid their work. Large amounts of supervisory data collected by central banks have the potential to identify risks, but more analysis is required on ways of operationalising this. It is also important to understand how the deployment of such techniques by supervisors might interact with firms’ use of similar techniques. Applying text analytics to large bodies of written communication is another fruitful avenue here.
Technology has the potential to change the nature of money. Central banks could respond to the digital world by opening their balance sheets to deposits from all sectors of the economy (not just banks) in the form of a digital currency. The trade-offs associated with issuing central bank digital currency (CBDC), including the extent to which central banks’ balance sheet is open to different actors in the economy, still need to be worked out. More broadly, there are important unresolved questions over the potential impact of CBDC on the monetary transmission mechanism, financial stability, and the industrial organisation of the financial system – including the implications of separating financial intermediation from offering means of payment. CBDC’s potential for overcoming the effective lower bound on the policy rate also needs to be explored.
Private cryptocurrencies have become increasingly popular. Do these currencies also require regulation? If so, this raises the issues of where the regulatory perimeter should lie and how policy should be structured. More work is needed on whether innovations behind cryptocurrencies such as distributed ledgers, permissionless systems and blockchain can be deployed elsewhere to improve the financial system.
New business models are emerging in the market for retail financial services. This competition has implications for the incumbent financial institutions, as well as the types of products available to consumers. A more fundamental issue is about the appropriate regulatory environment for new forms of retail financial services, including whether regulation should be set at the firm or product level.
Potential disruption from financial innovations raises the question about the competitiveness of the market for financial services, and the broader impact on the economy as a whole. How do technological changes and new business models affect competition in the financial sector? And what are the broader consequences of traditional financial intermediation being disrupted or replaced?
The end goal of financial policy is to make sure that the financial system allocates capital efficiently to support the real economy, while taking all the relevant risks into account. This calls for investigations into any potential causes of funding misallocation, including unintended consequences of regulation and other policies.
Priority topics for 2021
- What are the implications of deployment of machine learning and artificial intelligence by firms for the financial system?
- How can machine learning and artificial intelligence be deployed by supervisors?
- What are the implications of digital currencies issued by central banks and private entities on the monetary transmission mechanism and financial stability?
- How do disruptive entrants into retail financial services affect competition and the stability of the financial system?
+ Other work on using big data to inform decision making by central banks.
The Transformed World
Keywords: gig economy, economic measurement, intangible capital, income distribution, firm growth, competition, automation, climate change, pandemics, uncertainty.
The Covid-19 pandemic has created many challenges for policymakers and exposed new types of fragilities in the economy. Lessons can be learned from the pandemic about how financial markets and the real economy react to stressful situations, including how different sectors and agents interact. Moreover, the myriads of different macroeconomic and regulatory policies put in place are in need of careful evaluation. This also opens up a rich seam of future research to understand the macroeconomic consequences of the pandemic, beyond the immediate short-term fall in output and employment. Will Covid-19 also lead to long-term structural changes in the economy, eg acceleration of automation and online shopping?
The pandemic has also raised the broader question about who should bear risks from extreme events. In response, governments have provided extensive support to severely affected firms and households during the pandemic. This may lead to a re-evaluation of the macroeconomics of sovereign indebtedness. Further research on how financial markets can contribute to better distribution of such fundamental risks is also needed. How should insurance policies be structured to ensure better distribution of risks? Financial securities, such as catastrophe bonds, may also play a role.
Climate change and the transition to a low-carbon economy could create new risks and opportunities for the economy and the financial system. How do climate-related risks affect assets and liabilities of insurers and banks, and how should these be measured? How should central banks take into account these risks in their operations? A further understanding is needed of how the adoption of low-carbon technologies and energy sources alters economic dynamics. More work is also needed to examine how financial markets are adapting to the need to channel investment to support a transition to a low-carbon economy.
A deeper question is whether economies are entering a situation where uncertainty is structurally high. Uncertainty could stem from medical, natural, technological, and geopolitical sources, or from the financial system. The consequences of different types of uncertainty for saving and spending decisions, and for appropriate policy responses, need to be assessed.
The new economy
Traditional economic models struggle to capture many emerging features of modern economies. The shift away from manufacturing towards the so-called ‘new economy’ based on services, knowledge and technology poses challenges around quantifying output and investment. Such a shift could also change economic dynamics. For example, increased importance of intangible capital could affect collateralised borrowing and the transmission of policy shocks. Measurement-related uncertainty over the underlying productive capacity of the economy raises challenges for the policy process in estimating slack and inflationary pressures, which carry important implications for the setting of policy.
The new economy may change how the economic pie is divided. The so-called ‘gig economy’ has emerged, with many people operating outside of traditional firm-worker employment relations. Changing demographics may also have implications for both labour and capital markets. What will be the share of income accruing to capital and labour in the future? What factors will determine the distribution of capital and labour income across firms and households? What are the knock-on effects on growth and real interest rates?
Alongside a changing industrial mix, within industries the competitive environment may be shifting. Several potential trends require more investigation. Are mark-ups and market concentration rising, and are the most productive firms increasing their market share? If so, this raises the question of how evidence of declining competition can be reconciled with record rates of new firm creation in the UK in recent years.
These potential trends also raise issues of which firms will drive future productivity growth. Changes in the competitive landscape could also alter inflation dynamics, with knock-on policy implications. Most models of monetary policy are based on monopolistic competition, and so this naturally calls into question whether such an approach is still valid.
Is the nature of today’s technological change going to play out in an economically similar way to the innovations of the past few decades? Alternatively, some have argued that a ‘fourth industrial revolution’ is underway which is fundamentally transforming the economy. The first task for economists is to understand the relative merits of each view.
Further work is also needed to understand the effects of automation on individual product and labour markets. An important avenue for research is the effects of automation on employment in different sectors, skills and income distribution, and economic dynamics. How will automation affect the propagation of shocks, and how should policymakers respond?
Priority topics for 2021
- What has been the impact of Covid-19 and the macroeconomic policy responses on aggregate economic variables? How did this impact vary across different types of consumers, workers and firms?
- What are the policy trade-offs between short-term stability and long-term dynamism when an economy is hit by a large shock?
- What role can the different actors in the financial sector play in distributing, managing and mitigating risks arising from pandemics, climate change, and other fundamental shocks?
- How can the financial risks from climate change be evaluated?
+ Other work on the economic impact of Covid-19.