Rishi Kiroya and Lydia Henning
When you ask people what skills they have the most skills, you may receive answers such as “to fly”, “invisible”, or even “predict the future”. When you ask people who have worked in financial markets in particular, “predicting the future accurately” is probably the top of the list. From economic trends to political changes, market participants benefit from predicting what comes next. Using data collected from the Bank’s Market Participant Survey (MAPS), we will see how market forecasts develop later over periods of high uncertainty and volatility observed in the wake of the pandemic, and see how much prediction accuracy has changed depending on the period in question.
MAPS is a survey of monetary policy expectations carried out two weeks before all Monetary Policy Committees (MPC) meetings to gather information on topics related to MPCs. The map began as a pilot in mid-2020, then officially launched in February 2022, with the results published on the bank’s website 24 hours after each MPC’s decision (Check out Andrea Rosen’s speech).
I set up the scene, First, we look at the very close period of policy outlook, particularly the level of bank charges that arise from the latest MPC meetings. In Chart 1, the purple lines plot the median forecasts for “most likely” bank fees in each map survey (i.e. median MPCs for September 2024 recorded on the September 2024 map, and median MPCs for November 2024 map, etc.).
chart 1: Realization of bank fees against median expectations for future meetings
It can be seen that the median market participants correctly predict what will happen to bank charges at the next meeting of 19 of the 22 meetings explained in the previous sample.
How does this hold up when you expand the prediction window?
Chart 2 shows the average share of respondents with a record of bank fee forecasts that recorded 1, 2, and 3 policy announcements prior to the policy announcement in question. The average share of those who predicted the results of a general survey meeting is shown in deep purple for comparison. As you can imagine, the closer the market is to a decision, the more accurate the forecast is, as information is revealed and embedded in expectations.
chart 2: Prediction accuracy based on cycles
As you can imagine, looking at the sample, there is a higher tendency for forecasts to come true during periods when bank charges remained more constant than they were on the move.
What happens if I expand it further?
chart 3: Realization of bank charges against median recorded profiles
Chart 3 compares the median median profile of bank fees for the next 12 months, recorded at various points throughout the recent cycle, with the realisation path. This led to market participants tended to think about how high the price was until bank rates peaked. Interestingly, the median forecast for September 2023 was due to a slightly higher peak than what was realized.
Chart 4 compares subsequent realizations with median map predictions with the one-year horizon. If the criteria were set as accurate matches, such “hits” were observed at 19% of the time in the map samples. However, when allowing a 25 basis point threshold on both sides of the realized bank charge, the average accuracy was 40%.
chart 4: Average accuracy of median bank fees
Another alternative, a more generous benchmark only considers the direction of the bank rate path. That is, it goes up, goes down, or stays the same. On this scale (the darkest orange bar), we see that the median expected path for bank fees tends to evolve in the same direction as about 60% of the time.
Finally, we also see evidence that prediction accuracy differs between samples. It is clear that the proportion of median forecasts that were subsequently realized increased to early 2023 amid the tightening cycle of MPCs. It declines as bank charges peak and increase again during subsequent retention periods. This may be consistent with the respondent’s learning as respondents become accustomed to the cycle and adapted their expectations accordingly.
And we can’t talk about the other half of that, bank fees without talking about inflation.
chart 5: Average absolute deviation from realized inflation prints
Chart 5 uses the same approach as bank fees, comparing the expectations of the median map with subsequent realizations of inflation. The results show similar (and expected patterns) with the average deviation being lowest on the nearest horizon.
chart 6: Mean absolute deviation in cycle prediction
As can be seen in Chart 6, variations in material across the sample are also observed. This pattern is more monotonous than for bank rates, where there is a gap between forecasting (between the horizon for the year) and realizations that narrowly shorten the period. Splitting inflation expectations by the calendar year in which the map occurred, the average absolute deviation from the realised print of MAPS respondents fell approximately 3.5 times between 2022 and 2024, with respondents adapting to the upward surge in inflation followed by subsequent declines and relative leveling.
Chart 7 plots the median profile expected for inflation at various points during this cycle, achieving inflation on white. Similar to bank charges, median profiles tended to monitor what was realized below until realized inflation peaked. Furthermore, as the market was recalibrated through the sample, their expectations approached realised inflation.
Chart 7: Realized inflation for recorded median profiles
First I raised the question, “Fothic King or 220 foresight in the dark?” Evidence from the map (of course) shows that neither of them clearly apply. However, it must be said that the time frames included in the sample constitute an unprecedented period of volatility that must be highlighted in the report card, and the observation that market participants appear to be adapted to and “learning” in an evolving environment.
Rishi Kiroya and Lydia Henning are working in the bank’s market information and analysis department.
If you would like to contact us, please email us bankunderground@bankofengland.co.uk Or leave a comment below.
comment It will only be displayed after it has been approved by the moderator and only published if the full name is provided. Bank Underground is a blog where Bank of England staff can share opinions that challenge or support policy orthodoxy in general. The views expressed here are those of the authors and not necessarily those of the Bank of England or its policy committees.
Share the post “Fosic King or 220 Foresight in the Dark?”
