Navigating market signals: MaPS for policy makers – remarks by Andrea Rosen

Speech

Introduction

Good morning and many thanks to AFME for having me here today.

Before joining the Bank of England as Head of Market Intelligence and Analysis, I spent my entire career at an investment bank. As many of you can relate, my job involved a tremendous amount of time focused on what central banks said and did, and what that meant for markets. I often wondered how much policymakers understood market dynamics and how that fed into their decisions and communications. Now that I am on the other side, I get those questions all of the time.

By design, much of what happens inside a central bank is not public. But, given the importance of markets to central banks and vice versa, I think there is value in financial market participants better understanding how market intelligence is gathered; how it is analysed; and – crucially – how it is used by policymakers at the Bank of England.

The sheer breadth of what our Market Intelligence and Analysis function covers – servicing the Monetary Policy Committee (MPC), the Financial Policy Committee (FPC) and the Bank’s Governors, spanning the market conjuncture but also policy design, horizon scanning and structural issues – means that I could be here all day (Figure A).

Figure A: Our Market Intelligence and Analysis function

Instead, I’m going to focus my remarks on the work that we do to inform the MPC on financial markets, through a mix of discussions with market participants, data analytics and the new Market Participants Survey (MaPS).

Why financial markets matter to monetary policy makers

Monetary policy decisions are complex and depend on a large number of factors. Financial markets are also complex, and constantly evolving. It is certainly not the job of policymakers to just mechanically do what markets expect or want them to. But the MPC does invest heavily in trying to understand global markets, the signals they are sending us about the outlook, and the impact of asset prices and financial conditions for the real economy^{footnote [1]}.

This is not always straightforward: as you know, understanding markets does not stop at tracking price movements. It involves keeping up with the global market narrative, challenging what we hear with data, attempting to strip signals from the noise, taking liquidity and functioning into account, identifying potential biases and putting UK moves in a global, cross-asset context. It is also a highly iterative process.

It is important for the MPC to recognise market signals and expectations, without being governed by them. There are many reasons for this, but let me list a few:

• First, market pricing and the accompanying narrative provide an enormous amount of information that can help corroborate, as well as challenge policymakers’ views on the economic outlook, the international context and the perceived balance of risks. Ultimately everyone is trying to understand the economy and the outlook, so it is useful for the MPC to decode the views of thousands of professionals who do this across global markets, day in and day out.
• Second, market curves, rather than the policy makers’ own views about the path of Bank Rate, are used as a conditioning assumption for the MPC’s forecasts. Market pricing affects where the forecast ends up, and hence the signal it sends back.
• Third, financial markets are one of the main channels through which monetary policy actions can influence conditions to meet the inflation target. Market behaviours can and do change the trajectory of the real economy, as we saw during the Global Financial Crisis. And ultimately, households and firms borrow and lend at market rates. It is therefore essential for the MPC to be able to monitor the impact of its actions through markets, as it calibrates its reaction function.

Market channels are arguably more important in the UK, where the financial sector is unusually large and diverse relative to the size of the overall economy. Our status as a small open economy means that we don’t operate in a vacuum^{footnote [2]}, making it crucial to understand UK developments in an international framework. There are many examples of why, not least the global co-movement of risk free rates; the fact that equity risk premia are also correlated across countries; and the impact on sterling when global risk aversion leads to dollar strength.

Ultimately the MPC sets policy to meet its objective having synthesised a vast array of inputs. Sometimes that view will coincide with central expectations in the market, other times it may involve a policy “surprise”. But either way, market intelligence and analysis are key inputs to the information set that the Committee consider.

How the Bank’s Market Intelligence and Analysis function works in practice

The work we do to inform the MPC combines market data and our models with a rich set of qualitative inputs - based on extensive discussions with market participants and our own presence in markets, through sterling operations and foreign exchange reserve management^{footnote [3]}. Taken together, we call this Market Intelligence or “MI”.

Our robust approach to marrying feedback with data has been further supplemented by a new tool called the Market Participants Survey (MaPS)^{footnote [4]} , which aims to provide even more systematic, repeatable and representative aggregate feedback from a broad range of financial market participants on a subset of questions relevant to the MPC.

Let’s take the MI, data analytics and MaPS in turn.

Qualitative Market Intelligence

Direct interactions with - and feedback from - market participants allow us to put important context and nuance around market moves and data.

Needless to say, we can’t and don’t take what we hear at face value. We speak with a wide range of contacts for a given issue, and systematically aggregate the feedback, distil the conclusions and challenge them against our own judgement, as well as a comprehensive set of data and models.

We have dedicated MI teams for specific markets and sectors, as well as a cross-asset overlay (Figure B). We supplement the views from our external contact base with that of specialist experts within our own staff – both our traders^{footnote [5]}, as well as a number of subject matter specialists across the Bank. By frequently sharing and testing all of that information, a broader risk picture emerges.

Figure B: Market Intelligence at the Bank of England

We have an agile approach: primarily focusing on rates and FX markets, including a dedicated team for UK rate expectations, but with a global focus and built-in flexibility. For example, work on commodity markets has featured more prominently in recent MPC deliberations, as has our attempt to track how global growth expectations are feeding into equity and credit market conditions.

Our structure also enables us to look into emerging trends, such as the ways in which digital currencies may interact with the broader financial system, and the role of algorithmic trading in market functioning. Whilst we form our own judgments on what’s important, we are always interested to hear what topics are at the forefront of the mind of our contacts.

But who are these contacts? We interact with a wide array of market participants, spanning the buy- and sell-side, as well as industry bodies and providers of market infrastructure (Figure B). Our Meeting Varied People initiative, launched by
Governor Andrew Bailey in 2021, is aimed at seeking even greater diversity in both the institutions and individuals we engage with^{footnote [6]}. This diversity of thought ensures that we get the best information possible to fulfil our mission.

Our interactions are governed by the Market Intelligence Charter, which ensures rigour and transparency in terms of who we speak with and what we do with the information we receive.

Data analytics and modelling

An important part of our ability to challenge or support what we hear from market participants comes from examining market data, and testing views against our internal models.

Data helps us attempt to complete the picture when answering familiar questions like: how much is central expectation vs. balance of risks; are international metrics co-moving unusually; how can you decompose drivers in a given market; does the joint movement across asset classes corroborate one narrative vs another, etc.

There is much to learn by identifying the underlying drivers of movements and placing them in a cross-market, cross-asset context. For example, looking at bond/equity correlations tells us a richer story than just examining each market in isolation. And analysing the relationship between sterling, gilts and inflation metrics adds to our understanding of market views on the UK economy.

Another important use of market data analytics is to provide a quantitative framework to complement MI for monitoring market functioning. Liquidity, volatility and trading conditions often feature in our discussions with market participants, as they should, given the importance of market functioning to the transmission of monetary policy and as an indicator of financial stability. We take these discussions very seriously, but it is our ability to test what we hear with data that makes this insight even more powerful.

Amongst our tools, we can track UK market positioning and depth through trade-level and aggregate data. The Bank also has a number of dashboards that draw together a variety of relevant indicators across UK gilts, derivatives, FX and money markets. These dashboards give us a useful snapshot of, for example, whether UK metrics have moved sharply in a historical context; whether these moves are large relative to other markets; how each market is functioning; and what level of “noise” are we seeing in the data
(Chart 1).

We have a process in place to overlay what we see in the dashboards with what we hear and judge from markets, and to flag any issues with implications for market functioning directly to Governors. Given the volatility of the past couple of years, we engage with Governors frequently on this matter.

Chart 1: Evolution of Illustrative gilt market functioning indicators around the ‘Dash for Cash’ in 2020

The Market Participants Survey (MaPS)

Despite the huge value of the MI plus data approach, we perceived the need to add a layer of systematic and repeatable feedback from market active participants in the form of a regular survey. This would enable the MPC to quantify and probe market expectations for monetary policy beyond what pricing is already tells us, and would allow for cross-meeting comparisons to identify trends.

The Bank’s Market Participants Survey (MaPS) started as a pilot during in mid-2020 and was formally launched in February 2022^{footnote [7]}, with the results published on our website 24 hours after every MPC decision^{footnote [8]}. (Figure C)

Figure C: What is the Market Participants Survey (MaPS)?

Survey respondents comprise a broad set of buy- and sell-side market participants and have been selected by the Bank based on a number of criteria, including: (i) relevant market activity in UK rates, FX or money markets, (ii) external market committee membership^{footnote [9]}, (iii) expertise in UK markets and/or UK monetary policy, and (iv) willingness to participate regularly in the survey and in the Bank’s market intelligence activity^{footnote [10]}. Currently, we have around 40 respondents per round. The participants list will continue to evolve to reflect the composition of the markets we track, and we are open to considering additional respondents.

MaPS questions are formulated by Bank staff, involve topics that are widely discussed in the public domain and never presume any specific policy action. Some of the questions remain constant over time (for example Charts 2, 3, 4); others are drafted by staff to target the prevailing context. MaPS participants also have the opportunity to provide free-form comments as feedback, and follow up on answers during our regular market intelligence engagements.

Chart 2: Expectations for GBPUSD (MaPS – June 2022) ^(a)

Footnotes

• (a) MaPS question: What is your central expectation for GBPUSD at the following points in the future?

Chart 3: Expectations for the 10-year gilt-yield (MaPS – June 2022 ^(b)

Footnotes

• (b) MaPS Question: What is your central expectation for the 10-year gilt yield at the following points in the future (%)?

Of course, MaPS results are one of many pieces of information that the MPC takes into account when formulating policy. But comparing MaPS against our MI and data brings an additional dimension for the MPC to consider.

This has been particularly useful in analysing market pricing. If you think about what a market curve tells you, it is basically a combination of central (or modal) expectations for UK rates, plus assessments about the balance of risks, global spill overs, and technical factors such as liquidity and trading conditions.

Obviously, what is priced is the most relevant measure for MPC and the metric of the transmission mechanism. But it is also useful to have a sense of how much of that is central expectations for Bank Rate, and how much can be attributed to other factors. Options markets give us some information, although layered with caveats given persistent liquidity challenges.The last few MaPS surveys have shown a premium in the market-implied profile of SONIA swaps and futures at the time the survey is taken, versus the curve derived by plotting what MaPS respondents say they expect from UK rates. In short, the MaPS curve has consistently appeared below the markets curve (Chart 5).

Chart 5: Evolution of the difference between the median MaPS path for Bank Rate and the OIS curve (MaPS – Feb/Mar/May/Jun 2022)(a) (b)

Footnotes

• (a) Market curves correspond to the timing when the respective MaPS surveys were taken, rather than published. Source: Bloomberg and Bank calculations.(b) MaPS Question: What is your central expectation for Bank Rate after the following MPC meetings (%)?

To a significant degree, our market contacts have attributed this gap – through MaPS and MI - to the fact that, in their view, the risks remained skewed towards a higher path of Bank Rate (Chart 6), whether driven by domestic fundamentals, global spill overs and/or trading conditions (Chart 7). We have seen some of this play out in recent days, as curves continue to adjust based on global inflation and central bank expectations.

MaPS has added to our collective understanding of what may be driving the UK curve, quantifying the perceived impact of the balance of risks and allowing us to track its evolution over time.

Chart 6: Evolution of balance of risk perceptions surrounding Bank Rate expectations at one year, two year, and three year time horizons (MaPS – Mar/May/Jun) (a)

Footnotes

• (a) MaPS question: how would you describe the balance of risks surrounding your expectations for Bank Rate at the following horizons?

Chart 7: Reported drivers of the divergence between MaPS Bank Rate expectations and the market curve (MaPS - Jun 2022) (b)

Footnotes

• (b) MaPS question: It was noted in the May MPC Minutes that the respondents to the Market Participants Survey ‘expected a much lower central path for Bank Rate than the market-implied path further ahead’. Please weight the following factors in order of importance in affecting this gap between the Market Participants Survey central path for Bank Rate and the market-implied path.

Another area where MaPS and MI can be helpful tools is on market expectations about the future path of the Bank of England balance sheet. One can directly infer market expectations for the UK rates outlook by looking at the traded price of a range of instruments. But inferring the outlook for the Bank’s balance sheet is more difficult to measure from market prices alone, particularly at the point where a balance sheet policy is still being considered^{footnote [11]} and there are a number of global dynamics at play.

Long rates are affected by a number of factors that are all endogenous to one another, and you could only extract an estimate of one if you control for the others. There is also a wide range of views of the impact of balance sheet policies on yields curves, so disentangling expectations is very difficult outside of narrow event windows.

MaPS and MI can help. For example, MaPS asks about respondents’ expectations for the stock of asset purchases (Chart 8), as well as their perceptions of the possible impact of balance sheet unwind on yield curve pricing. Tracking the evolution of these responses has given the MPC valuable information.

In parallel, our market intelligence function has been seeking views from a wide range of stakeholders on the operational considerations related to any future gilt sales, ahead of the Committee’s update in August and decisions at any subsequent meetings.^{footnote [12]} We expect views to continue evolving as more information is known, and will feed that accordingly to the MPC.

Chart 8: Views from MaPS respondents on the evolution of the Bank of England’s gilt holdings (MaPS – Mar/May/Jun 2022)(a)

Connecting Markets to the MPC

In short, the systematic set of measures on core concepts that MaPS provides acts as a powerful cross-check to our qualitative feedback and data analysis functions. The

combination of all three ensures that the MPC is well informed not just on what markets are pricing, but also on what they are telling us and why it matters.

So how do we pull all these strands together?

Our Market Intelligence and Analysis function is completely integrated into the 6-weekly MPC rounds (Figure D^{footnote [13]}) and we are in regular contact with the Committee in between.

Staff from our Markets teams provide daily and sometimes intra-day updates on market conditions to Governors and MPC members. We engage with policy makers frequently through both formal and informal discussions about market conditions, and provide a series of written and verbal briefings to the Committee in the lead up to each MPC round.

Through MI, MaPS and data analytics, our briefings build on top of the basic data and give a sense of the evolving market narratives, the global, cross-asset picture, the decomposition of moves through our economic models, and how all of that feeds into expectations for a specific policy decision.

Our work continues throughout the round, and the collective views of market participants are often referenced in the Minutes, the Monetary Policy Report and related communications. Our insight into market conditions also directly informs policy design, operational strategy and implementation.

Figure D: Timeline of MPC meeting schedule with supporting MI gathering and MaPS

Conclusion

I have laid out how - although markets are important to the MPC through a number of different channels - market expectations are an input, rather than a driver in the policymaking process.

Markets can be noisy, and Committee members are invested in attempting to understand and navigate them as best as possible. I hope that I have added to your understanding of the work that goes into making that happen.

Having spent the majority of my career in the City, I have been struck by the level of sophistication, experience and depth of market knowledge that exists within Threadneedle Street. I have seen first-hand how much focus and attention the Bank directs towards

markets; how extensive its market analysis, intelligence and operations are; and how important the market transmission mechanism is to the Bank’s mission.

As markets evolve, so do our market analysis and intelligence activities. We continue to invest in our data analytics and technology, to enhance and improve MaPS, and to remain agile in our thematic focus. Importantly, we wish to continue expanding our network of market contacts, so that they are even more diverse^{footnote [14]} and reflective of the market setup.

We value thoughtful, open, and candid input from market participants such as you. By necessity, some of our discussions tend to be one-way. But we must also be open where we can. The more our contacts understand what we do, the better they can help us to promote the good of the people of the United Kingdom by maintaining monetary and financial stability.

I am very grateful to Ben Bowry and Helena Patterson for their assistance in preparing these remarks.

Thanks to Andrew Bailey, Dave Ramsden, Ben Broadbent, Andrew Hauser,
Catherine Mann, Silvana Tenreyro, Harry Austin, Kiman Bassi, Jonathan Bridges,
Alan Castle, Michael Foster, Sumita Ghosh, David Glanville, Robert Hills, Tom Jennings, Rafael Kinston, Sean Maloney, Amber McAlone, Nick McLaren, Arif Merali, Rhys Phillips, Will Rawstorne, Matt Roberts-Sklar and Tom Smith for their helpful advice and comments.