CPI inflation and activity have evolved broadly in line with the MPC’s expectations at the time of the May Report. Inflation was 2.4% in June, above the target due to sterling’s past depreciation and higher energy prices. Those external cost pressures are projected to dissipate over the forecast period, though at a slightly slower rate than projected in May following the further fall in the sterling exchange rate over the past three months. GDP growth appears to have recovered in Q2. UK demand is expected to continue to grow at a modest pace which, given subdued potential supply growth, is likely to be more than sufficient to use up the very limited degree of slack remaining in the economy. Conditioned on a path for Bank Rate that rises to 1.1% over the next three years, a small margin of excess demand is likely to emerge by late 2019 and to build thereafter, raising domestic inflationary pressures. Taken together with diminishing external pressures, CPI inflation is projected to decline towards the target, reaching 2% in the third year of the forecast period.
UK GDP growth is reported to have dipped to 0.2% in Q1, partly reflecting the temporary impact of adverse weather. But it is estimated to have recovered to 0.4% in Q2, as anticipated in the May Report.
Global GDP growth was slightly weaker than expected in Q2 and the outlook has moderated slightly over the past three months. Nevertheless, most indicators of global activity suggest that growth should remain relatively robust at above-potential rates (Key Judgement 1).
Along with accommodative financial conditions, relatively robust global growth is projected to support UK demand, particularly investment and net trade. Uncertainty around the United Kingdom’s future trading arrangements continues to weigh on business investment, however (Section 2). And while the fall in sterling associated with the EU referendum is boosting net trade, it is still having some dampening effects on the growth rates of household real income and consumption (Key Judgement 2). The sterling ERI was around 2½% lower than in the run-up to the May Report and around 17% below its late-2015 peak. As in previous Reports, the MPC’s projections are conditioned on the average of a range of possible outcomes for the United Kingdom’s eventual trading relationship with the European Union. They also assume that households and companies base their decisions on the expectation of a smooth adjustment to those new trading arrangements.
The MPC’s projections, summarised in Table 5.A, are conditioned on those assumptions and a path for Bank Rate that reaches 1.1% by mid-2021 (Table 5.B), just over 10 basis points lower than in May.1 Four-quarter GDP growth is projected to recover from its temporary weakness at the start of 2018, and average around 1¾% over the forecast period (Chart 5.1), similar to May. Potential supply growth is subdued relative to pre-crisis norms, reflecting continued weakness in productivity growth and slower growth of the working-age population. The pace of demand growth is more than sufficient to absorb the very limited degree of spare capacity that currently remains in the economy (Key Judgement 3). A range of indicators suggest that the labour market is tight.
Unemployment remains low, and is projected to fall a little further (Chart 5.2). The economy moves into excess demand by late 2019. That leads to a continuing firming of wage growth and domestic inflationary pressures.
While domestic inflationary pressures firm over the forecast period, external cost pressures ease (Key Judgement 4). The rise in import prices following sterling’s referendum-related depreciation and higher energy prices have accounted for above-target inflation since the beginning of 2017, but their combined impact is likely to subside in coming years. Inflation is projected to decline towards the target (Chart 5.3), although the depreciation of sterling since the May Report slows the projected pace of decline slightly compared with the previous forecast (Chart 5.4). Inflation reaches 2% in the third year of the forecast period.
At its meeting ending on 1 August 2018, the MPC voted to increase Bank Rate by 0.25 percentage points, to 0.75%, to maintain the stock of sterling non-financial investment grade corporate bond purchases, financed by the issuance of central bank reserves, at £10 billion and to maintain the stock of UK government bond purchases, financed by the issuance of central bank reserves, at £435 billion. The factors behind that decision are set out in the Monetary Policy Summary on page i of this Report and in more detail in the Minutes of the meeting.2 The remainder of this section sets out the MPC’s projections and the risks around them in more detail.
- a Modal projections for GDP, CPI inflation, LFS unemployment and excess supply/excess demand. Figures in parentheses show the corresponding projections in the May 2018 Inflation Report. Projections were only available to 2021 Q2 in May.
b The August projections have been conditioned on the assumptions that the stock of purchased gilts remains at £435 billion and the stock of purchased corporate bonds remains at £10 billion throughout the forecast period, and on the Term Funding Scheme (TFS); all three of which are financed by the issuance of central bank reserves. The May projections were conditioned on the same asset purchase and TFS assumptions.
c Four-quarter growth in real GDP. The growth rates reported in the table exclude the backcast for GDP. Including the backcast 2018 Q3 growth is 1.6%, 2019 Q3 growth is 1.8%, 2020 Q3 growth is 1.7% and 2021 Q3 growth is 1.7%. This compares to 1.7% in 2018 Q3, 1.7% in 2019 Q3 and 1.7% in 2020 Q3 in the May 2018 Inflation Report.
d Four-quarter inflation rate.
e Per cent of potential GDP. A negative figure implies output is below potential and a positive figure that it is above.
f Per cent. The path for Bank Rate implied by forward market interest rates. The curves are based on overnight index swap rates.
Conditioning path for Bank Rate implied by forward market interest ratesa
- a The data are 15 working day averages of one‑day forward rates to 25 July 2018 and 2 May 2018 respectively. The curve is based on overnight index swap rates.
b August figure for 2018 Q3 is an average of realised overnight rates to 25 July 2018, and forward rates thereafter.
GDP projection based on market interest rate expectations, other policy measures as announced
- The fan chart depicts the probability of various outcomes for GDP growth. It has been conditioned on the assumptions in Table 5.A footnote b. To the left of the vertical dashed line, the distribution reflects uncertainty around revisions to the data over the past. To aid comparability with the official data, it does not include the backcast for expected revisions, which is available at Data from the August 2018 Inflation Report. To the right of the vertical line, the distribution reflects uncertainty over the evolution of GDP growth in the future. If economic circumstances identical to today’s were to prevail on 100 occasions, the MPC’s best collective judgement is that the mature estimate of GDP growth would lie within the darkest central band on only 30 of those occasions. The fan chart is constructed so that outturns are also expected to lie within each pair of the lighter green areas on 30 occasions. In any particular quarter of the forecast period, GDP growth is therefore expected to lie somewhere within the fan on 90 out of 100 occasions. And on the remaining 10 out of 100 occasions GDP growth can fall anywhere outside the green area of the fan chart. Over the forecast period, this has been depicted by the light grey background. See the box on page 39 of the November 2007 Inflation Report for a fuller description of the fan chart and what it represents.
Unemployment projection based on market interest rate expectations, other policy measures as announced
- The fan chart depicts the probability of various outcomes for LFS unemployment. It has been conditioned on the assumptions in Table 5.A footnote b. The coloured bands have the same interpretation as in Chart 5.1, and portray 90% of the probability distribution. The calibration of this fan chart takes account of the likely path dependency of the economy, where, for example, it is judged that shocks to unemployment in one quarter will continue to have some effect on unemployment in successive quarters. The fan begins in 2018 Q2, a quarter earlier than the fan for CPI inflation. That is because Q2 is a staff projection for the unemployment rate, based in part on data for April and May. The unemployment rate was 4.2% in the three months to May, and is projected to be 4.1% in Q2 as a whole. A significant proportion of this distribution lies below Bank staff’s current estimate of the long-term equilibrium unemployment rate. There is therefore uncertainty about the precise calibration of this fan chart.
CPI inflation projection based on market interest rate expectations, other policy measures as announced
CPI inflation projection in May based on market interest rate expectations, other policy measures as announced
- Charts 5.3 and 5.4 depict the probability of various outcomes for CPI inflation in the future. They have been conditioned on the assumptions in Table 5.A footnote b. If economic circumstances identical to today’s were to prevail on 100 occasions, the MPC’s best collective judgement is that inflation in any particular quarter would lie within the darkest central band on only 30 of those occasions. The fan charts are constructed so that outturns of inflation are also expected to lie within each pair of the lighter red areas on 30 occasions. In any particular quarter of the forecast period, inflation is therefore expected to lie somewhere within the fans on 90 out of 100 occasions. And on the remaining 10 out of 100 occasions inflation can fall anywhere outside the red area of the fan chart. Over the forecast period, this has been depicted by the light grey background. See the box on pages 48–49 of the May 2002 Inflation Report for a fuller description of the fan chart and what it represents.
5.1 The MPC's key judgements and risks
Key Judgement 1: global demand grows at above-potential rates
Global economic growth has remained above trend in the first half of 2018, with four-quarter growth, using PPP weights, a little under 4%. World goods trade growth has decelerated in recent months, however, and the outlook for global activity and trade appears to have moderated slightly since the May Report. This reflects a recent pickup in geopolitical uncertainty, in part related to an intensification of trade tensions, alongside tightening financial conditions particularly in some emerging economies. Nonetheless, financial conditions remain accommodative by historical standards and, alongside healthy business and consumer confidence, should support global growth at rates above potential over the forecast period.
Quarterly euro-area growth dipped in the first half of 2018, averaging 0.4%. That was slower than the 0.7% average rate experienced in 2017 and below the May forecast. Some of the slowing appears to reflect temporary factors, however, and indicators suggest that underlying demand growth remains healthy (Section 1). As a result, growth is expected to tick up in the near term, albeit to slightly lower rates than projected in May (Table 5.C). Thereafter, it slows towards potential. Above-trend growth will lead to the gradual absorption of spare capacity. As that slack is absorbed, core inflation is projected to rise to just over 1½% by the end of the forecast, slightly lower than expected at the time of the May Report.
Quarterly US growth was 1% in Q2, up from 0.5% in Q1. Q2 GDP was higher than expected in May and well above trend rates. Looking ahead, the tariffs that have been implemented and proposed on bilateral trade between the US and its trading partners, including China, are likely to weigh on growth (Section 1). Taken together, the MPC’s central projection for US activity is little changed from May. As in May, growth is expected to ease over the forecast, partly reflecting a declining boost from recent fiscal measures. Nevertheless, excess demand builds, and inflation is expected to remain above 2% throughout the forecast period.
While GDP growth in China and other emerging economies has in aggregate remained relatively robust, the outlook has softened compared with the May Report. In China, the tariffs on trade with the US, and an associated fall in equity prices, are expected to dampen growth a little. In other emerging economies, tightening monetary policy in the US and increasing trade tensions — as well as idiosyncratic institutional and political developments in some countries — have resulted in tighter financial conditions. On average, emerging economy government and corporate bond spreads have risen by around 60 basis points and equity prices have fallen by 8% since the May Report. Moreover, emerging economy currencies have depreciated by around 6% against the dollar, on average, and some central banks have increased policy rates to try to stem capital outflows. That tightening in financial conditions weighs on the growth projection, which is a little lower than in the May Report.
Global growth — based on PPP weights — is projected to be 3¾% in 2018, before slowing to 3½% (Chart 5.5). Weighted by UK export shares, growth is projected to slow from 2¾% in 2018 to 2¼% in 2020 (Table 5.C). Those projections are a little lower than three months ago, reflecting the impact of the implemented and proposed tariffs on trade and tighter financial conditions in emerging economies. The potential for a continued increase in the barriers to trade and for a further tightening in financial conditions create downside risks to the outlook. On the upside, there remains a possibility of a larger recovery in productivity growth, which would allow economies to grow more quickly without leading to inflationary pressure.
Key Judgement 2: net trade and business investment continue to support UK activity, while consumption growth remains modest
The outlook for UK demand is very similar to that in May. After some temporary weakness in the near term reflecting the slowing in quarterly growth in Q1, four-quarter GDP growth is expected to average around 1¾%. Net trade and business investment support GDP growth, while consumption grows modestly, in line with real incomes.
Net trade contributed positively to growth in 2017 and 2018 Q1. Although it is expected to have subtracted from growth in 2018 Q2 — which will depress its contribution to GDP growth in 2018 as a whole — that appears to reflect erratically weak export growth. Survey indicators of export growth remain robust (Section 2) and net trade is expected to improve in the second half of the year (Table 5.D). Export demand will continue to benefit from relatively robust growth in the global economy. Net trade will also be supported by the lower sterling exchange rate.
Business investment has also been supported by external demand, as well as by the low cost of finance, the relatively high rate of return on capital, and the incentive to invest to expand capacity. The pace of growth has been dampened, however, by the anticipation of, and uncertainty over, Brexit. Business investment is currently estimated to have fallen in 2018 Q1. This may in part reflect the impact of adverse weather on construction-related investment (Section 2). Growth is expected to have recovered in Q2, and to remain a little above its past average rate over the rest of 2018. That rate is subdued relative to past recoveries, reflecting the drag from uncertainty. As that wanes, business investment growth is expected to rise a little further (Table 5.E), although there is a risk that uncertainty increases and weighs more heavily on growth.
Consumption growth has been modest since 2017. In large part, that reflects weak growth in households’ real incomes, which have been squeezed by the rising cost of imports following sterling’s referendum-related depreciation. Consumption growth has declined by less than real income growth over the recent past, however, as households have reduced their rate of saving, supported by rising net wealth (see Box 4), high employment and accommodative financial conditions (Section 2). Real income growth is expected to be higher over the forecast period than it has been in recent years, as nominal wage growth continues to pick up (Key Judgement 3) and the drag from import price inflation continues to fade (Key Judgement 4). Consumption is expected to grow broadly in line with real incomes, such that the saving ratio remains broadly unchanged.
The housing market has also been subdued. Price inflation has fallen over the past year and activity has been flat. That is likely in part to reflect weak real income growth, in common with consumption. Developments in the housing market can also affect consumption growth independently through wealth and collateral channels. Thus far, the weakness in the housing market appears to be concentrated in London (Section 2). The MPC judges that has partly reflected idiosyncratic factors, and so is unlikely to have significant spillovers to wider UK housing markets. Annual UK house price inflation is projected to recover to just over 3% over the forecast period.
There is uncertainty about the extent to which households will adjust their spending and saving. The saving ratio is estimated to have fallen to a low level over the past few years and households might choose to build savings at a somewhat faster rate as real income growth rises, depressing spending. Unemployment remains low, however, and households currently report a historically high degree of job security. So it is possible that households could lower their saving rate further, boosting consumption growth, with the additional support of generally healthy balance sheets.
Key Judgement 3: demand growth outstrips subdued potential supply growth, and a margin of excess demand emerges, pushing up domestic cost growth
The speed at which demand can grow before it puts upward pressure on inflation depends on the amount of slack in the economy and on the growth rate of potential supply. The MPC judges that there is currently only a very limited degree of slack remaining in the economy, given the tightness reported in the labour market (Section 3) and little evidence of much spare capacity within companies. Given that, demand can only sustainably grow at rates in line with the expansion of potential supply.
As set out in its assessment of supply-side conditions in February, the MPC judges that growth in potential supply will remain subdued relative to pre-crisis rates, at around 1½% per year on average. Within that, labour supply growth is likely to be modest, and a little slower than current rates. That slowing in part reflects an expected decline in net inward migration in line with the ONS projections on which the MPC’s forecasts are conditioned. Offsetting that, structural productivity growth is projected to improve somewhat to close to 1%, although that remains around 1 percentage point lower than pre-crisis norms. The pickup in productivity growth over the forecast partly reflects higher investment feeding through to an increase in the capital stock.
There are significant risks to the outlook for productivity. On the downside, productivity growth has been lower than expected since the financial crisis. It could fail to pick up again if, for example, lower-than-expected investment weighs on the growth of the capital stock. On the upside, productivity growth could pick up to closer to historical norms. That could be driven by a boost to productivity growth from higher investment, if companies substitute towards capital and away from labour in a tight labour market, for example.
Conditional on market interest rate expectations of Bank Rate rising to 1.1% over the forecast period, demand is projected to grow a little faster than potential supply over the forecast period, such that a small margin of excess demand emerges by late 2019 and continues to build thereafter.
The absorption of slack and emergence of excess demand will put upward pressure on domestic cost growth. Domestic inflationary pressures have been subdued over the past few years while there has been some degree of slack in the economy. Labour costs, which are the largest domestic component of companies’ costs, have generally grown at modest rates for much of the post-crisis period.
Pay growth has picked up over the past year in both the private and public sectors. The labour market has tightened and companies have found it harder to recruit and retain staff (Section 3). That is apparent in a range of survey indicators of firms’ hiring difficulties and pay — including the Agents’ recent survey (see Box 5). Higher wage growth will lead to increased inflationary pressure to the extent that it occurs without a commensurate pickup in productivity growth. Over the past year or so, the increase in annual pay growth has not been matched by higher growth in productivity per head, such that unit wage and labour cost growth have risen. Over the forecast period, private sector and whole-economy unit labour cost growth are projected to remain higher than they have been over the past few years, leading to a gradual firming in domestic inflationary pressures.
Key Judgement 4: domestic inflationary pressures continue to build over the forecast period, while external cost pressures ease
CPI inflation has fallen back since the start of 2018. In 2018 Q2 it was 2.4%, in line with the MPC’s expectation at the time of the May Report. The decline in inflation over 2018 has partly reflected a waning impact from import prices, which rose substantially following the referendum-related depreciation of sterling. The contribution from import prices to CPI inflation is estimated to remain elevated, however, and can account for much of the current overshoot of inflation relative to the target. The MPC judges that it is likely that the effect of sterling’s depreciation around the EU referendum on import prices has already come through, although the further pass-through to consumer prices has some way to run. Import price inflation is expected to rise again over the next few quarters, by slightly more than expected in May (Chart5.6), reflecting the recent depreciation of sterling and an acceleration in global export prices (Section 1). It will take time for those rises to feed through fully into CPI inflation and import prices will carry on boosting inflation over the forecast period, although the overall boost will wane.
CPI inflation has also been affected by movements in energy prices. The oil price, which affects fuel prices rapidly and so can have marked temporary effects on inflation, has been volatile over the past few months. The sterling oil price rose sharply around the time of the May Report, pushing up petrol prices. It has fallen back a little more recently, however, which reduces its contribution to CPI inflation in the near term. Further ahead, the oil futures curve on which the MPC’s forecast is conditioned is downward sloping, such that the contribution from fuel prices to inflation turns negative.
Retail gas and electricity prices have risen over the past few months, reflecting previous increases in wholesale energy and other costs. Recent further increases in wholesale costs are likely to put upwards pressure on retail prices over the forecast period. Acting in the opposite direction is the likely impact of the forthcoming cap on households’ energy tariffs (Section 4). The MPC judges that the effects of those factors on CPI are broadly offsetting, although the size and timing of both are uncertain.
Taken together, the impact of these external factors on inflation is expected to dissipate over the forecast period. The projected strengthening in domestic inflationary pressures as excess demand builds (Key Judgement 3) only partially offsets that dissipation in external cost pressures, such that in the central projection CPI inflation declines towards the 2% target. Inflation is judged likely to be a little above the target over much of the forecast period, before returning to 2% in the third year. That is slightly higher than expected in May, mainly reflecting the impact of the recent depreciation of sterling.
MPC key judgementsab
- Sources: Bank of England, BDRC Continental SME Finance Monitor, Bloomberg Finance L.P., British Household Panel Survey, Department for Business, Energy and Industrial Strategy, Eurostat, ICE/BoAML Global Research (used with permission), IMF World Economic Outlook (WEO), ONS, US Bureau of Economic Analysis and Bank calculations.
a The MPC’s projections for GDP growth, CPI inflation and unemployment (as presented in the fan charts) are underpinned by four key judgements. The mapping from the key judgements to individual variables is not precise, but the profiles in the table should be viewed as broadly consistent with the MPC’s key judgements.
b Figures show annual average growth rates unless otherwise stated. Figures in parentheses show the corresponding projections in the May 2018 Inflation Report.
c Chained-volume measure. Constructed using real GDP growth rates of 180 countries weighted according to their shares in UK exports.
d Chained-volume measure. Constructed using real GDP growth rates of 181 countries weighted according to their shares in world GDP using the IMF’s purchasing power parity (PPP) weights.
e Chained-volume measure. Forecast was finalised before the release of the preliminary flash estimate of GDP for Q2, so that is not incorporated.
f Chained-volume measure.
g Chained-volume measure.
h Annual average. Chained-volume business investment as a percentage of GDP.
i Chained-volume measure. Exports less imports.
(j) Chained-volume measure. Includes non-profit institutions serving households.
(k) Level in Q4. Percentage point spread over reference rates. Based on a weighted average of household and corporate loan and deposit spreads over appropriate risk-free rates. Indexed to equal zero in 2007 Q3.
(l) Based on the weighted average of spreads for households and large companies over 2003 and 2004 relative to the level in 2007 Q3. Data used to construct the SME spread are not available for that period. The period is chosen as broadly representative of one where spreads were neither unusually tight nor unusually loose.
(m) Annual average. Percentage of total available household resources.
(n) GDP per hour worked.
(o) Level in Q4. Percentage of the 16+ population.
(p) Level in Q4. Average weekly hours worked, in main job and second job.
(q) Four-quarter growth in unit labour costs in Q4. Whole-economy total labour costs divided by GDP at market prices, based on the mode of the MPC’s GDP backcast. Total labour costs comprise compensation of employees and the labour share multiplied by mixed income.
(r) Four-quarter growth in unit wage costs in Q4. Whole-economy total wage costs divided by GDP at market prices, based on the mode of the MPC’s GDP backcast. Total wage costs are wages and salaries and the labour share multiplied by mixed income.
(s) Four-quarter inflation rate in Q4.
(t) Average level in Q4. Dollars per barrel. Projection based on monthly Brent futures prices.
World GDP (PPP‑weighted)a
- Sources: IMF WEO and Bank calculations.
a Annual average growth rates. Chained‐volume measure. Constructed using real GDP growth rates of 181 countries weighted according to their shares in world GDP using the IMF’s purchasing power parity (PPP) weights.
Indicative projections consistent with the MPC’s modal projectionsa
- a These projections are produced by Bank staff for the MPC to be consistent with the MPC’s modal projections for GDP growth, CPI inflation and unemployment. Figures show annual average growth rates unless otherwise stated. Figures in parentheses show the corresponding projections in the May 2018 Inflation Report.
b Chained-volume measure. Includes non-profit institutions serving households.
c Chained-volume measure.
d Chained-volume measure. Whole-economy measure. Includes new dwellings, improvements and spending on services associated with the sale and purchase of property.
e Chained-volume measure. The historical data exclude the impact of missing trader intra-community (MTIC) fraud.
f Total available household resources deflated by the consumer expenditure deflator.
g Whole-economy total pay.
Import price inflationa
- Sources: ONS and Bank calculations.
a Projections are four-quarter inflation rate in Q4. Excludes the impact of MTIC fraud.
5.2 The projections for demand, unemployment and inflation
Based on these judgements, under the market path for Bank Rate and the assumption of an unchanged stock of purchased assets, the MPC projects four-quarter GDP growth to average around 1¾% over the forecast period (Table 5.F ). That projection is similar to the May forecast (Chart 5.7). Within demand, consumption growth is projected to remain modest relative to historical rates, with net trade and investment supporting growth, in turn supported by relatively robust global growth. The risks around the projection are balanced, as in May.
The economy’s supply capacity is judged likely to grow at a subdued pace — of around 1½% per year on average — over the forecast period. That is slightly lower than demand growth, with unemployment projected to fall a little further as a result and the economy moving into excess demand by late 2019.
CPI inflation has fallen back since the beginning of 2018, but remains above the MPC’s 2% target. The inflation overshoot reflects the impact of external cost pressures from import and energy prices. Inflation is projected to fall further towards the target as those effects wane, more than offsetting building domestic inflationary pressures. Under the market path for Bank Rate, inflation is judged likely to decline towards the target, reaching 2% in the third year of the forecast period (Chart 5.8). The projection is higher than in May (Table 5.G), mainly reflecting the effect of the recent depreciation of sterling. The risks around the inflation projection remain balanced (Chart 5.9).
Chart 5.10 and Chart 5.11 show the MPC’s projections under the alternative constant rate assumption and an unchanged stock of purchased assets. That assumption is that Bank Rate remains at 0.75% throughout the three years of the forecast period, before rising towards the market path over the subsequent three years. Under that path, GDP growth is stronger, and a greater degree of excess demand emerges, with unemployment falling towards 3½%. Inflation is higher and ends the forecast period above the target at 2.2%.
1 Unless otherwise stated, the projections shown in this section are conditioned on: Bank Rate following a path implied by market yields; the stock of purchased gilts remaining at £435 billion and the stock of purchased corporate bonds remaining at £10 billion throughout the forecast period and the Term Funding Scheme (TFS), all three of which are financed by the issuance of central bank reserves; the Recommendations of the Financial Policy Committee and the current regulatory plans of the Prudential Regulation Authority; the Government’s tax and spending plans as set out in the Spring Statement 2018; commodity prices following market paths; and the sterling exchange rate remaining broadly flat. For more details, see the ‘Data from the August 2018 Inflation Report’ section.
2 The Minutes are available here.
Annual average GDP growth rates of modal, median and mean pathsa
- a The table shows the projections for annual average GDP growth rates of modal, median and mean projections for four‑quarter growth of real GDP implied by the fan chart. The figures in parentheses show the corresponding projections in the May 2018 Inflation Report excluding the backcast. The projections have been conditioned on the assumptions in Table 5.A footnote b.
Projected probabilities of GDP growth in 2020 Q3 (central 90% of the distribution)a
- a Chart 5.7 represents the cross‑section of the GDP growth fan chart in 2020 Q3 for the market interest rate projection. The grey outline represents the corresponding cross‑section of the May 2018 Inflation Report fan chart for the market interest rate projection excluding the backcast. The projections have been conditioned on the assumptions in Table 5.A footnote b. The coloured bands in Chart 5.7 have a similar interpretation to those on the fan charts. Like the fan charts, they portray the central 90% of the probability distribution.
b Average probability within each band; the figures on the y‑axis indicate the probability of growth being within ±0.05 percentage points of any given growth rate, specified to one decimal place.
Inflation probabilities relative to the target
- The August and May swathes in this chart are derived from the same distributions as Charts 5.3 and 5.4 respectively. They indicate the assessed probability of inflation relative to the target in each quarter of the forecast period. The 5 percentage points width of the swathes reflects the fact that there is uncertainty about the precise probability in any given quarter, but they should not be interpreted as confidence intervals.
Q4 CPI inflation
- The table shows projections for Q4 four‑quarter CPI inflation. The figures in parentheses show the corresponding projections in the May 2018 Inflation Report. The projections have been conditioned on the assumptions in Table 5.A footnote b.
Projected probabilities of CPI inflation in 2020 Q3 (central 90% of the distribution)a
- a Chart 5.9 represents the cross‑section of the CPI inflation fan chart in 2020 Q3 for the market interest rate projection. The grey outline represents the corresponding cross‑section of the May 2018 Inflation Report fan chart for the market interest rate projection. The projections have been conditioned on the assumptions in Table 5.A footnote b. The coloured bands in Chart 5.9 have a similar interpretation to those on the fan charts. Like the fan charts, they portray the central 90% of the probability distribution.
b Average probability within each band; the figures on the y‑axis indicate the probability of inflation being within ±0.05 percentage points of any given inflation rate, specified to one decimal place.
GDP projection based on constant nominal interest rates at 0.75%, other policy measures as announced
- See footnote to Chart 5.1.
CPI inflation projection based on constant nominal interest rates at 0.75%, other policy measures as announced
- See footnote to Chart 5.3
Table 5.D Monitoring risks to the Committee’s key judgements
The Committee’s projections are underpinned by four key judgements. Risks surround all of these, and the MPC will monitor a broad range of variables to assess the degree to which the risks are crystallising. The table below shows Bank staff’s indicative near-term projections that are consistent with the judgements in the MPC’s central view evolving as expected.
Likely developments in 2018 Q3 to 2019 Q1 if judgements evolve as expected
|1: global demand grows at
|2: net trade and business
investment continue to
support UK activity,
growth remains modest
|3: demand growth
potential supply growth,
and a margin of excess
pushing up domestic cost
|4: domestic inflationary
pressures continue to
build over the forecast
period, while external
cost pressures ease
Box 6: The equilibrium interest rate
The MPC sets monetary policy to meet the 2% inflation target, and in a way that helps to sustain growth and employment. The policy rate required to achieve these objectives will vary over time. Prior to the financial crisis, the level of Bank Rate set by the Committee averaged 5%.1 During the crisis, however, Bank Rate needed to be cut sharply in order for the MPC to meet its objectives in the face of major headwinds to demand. While Bank Rate is expected to need to rise gradually as those headwinds diminish and slack is absorbed, longer-term structural influences mean that those rises are expected to be limited and it is unlikely that Bank Rate will return to pre-crisis levels anytime soon. This box presents the MPC’s latest assessment of those influences on Bank Rate and the outlook for coming years.
Understanding the equilibrium interest rate
The ‘equilibrium interest rate’ is the interest rate that, if the economy starts from a position with no output gap and inflation at the target, would sustain output at potential and inflation at the target. Setting Bank Rate equal to the equilibrium interest rate may not be sufficient to meet the 2% inflation target at all times, however. For example, if output were below (above) its potential, Bank Rate would need to be set below (above) the equilibrium rate for a period of time in order to return output to potential. Furthermore, some shocks may create a trade-off between the speed at which inflation is returned to the target and the support provided to jobs and activity.2
The equilibrium interest rate cannot be directly observed, and so is not used by monetary policy makers as a direct guide to setting policy. But to the extent that it — and the factors driving it — can be estimated, it may help to explain the evolution of interest rates over the past and provide an indication of the outlook for interest rates over coming years.
One framework for understanding the equilibrium real interest rate (r*) — the equilibrium interest rate adjusted for inflation — decomposes it into a longer-run, or ‘trend’, component (R*) and a shorter-term component (s*):
The trend real rate, R*, is determined by slow-moving structural factors that affect the balance between the demand for capital and the stock of wealth available to finance it. For an open economy like the UK, those factors will reflect global influences as well as domestic ones. The upward-sloping red lines in Figure A show that the quantity of wealth individuals want to hold tends to increase as real interest rates rise, since a higher real interest rate implies a greater return on saving. In contrast, the demand for capital (shown in the blue lines) tends to be downward sloping: as the quantity of capital rises, its marginal product falls and so each extra unit of wealth will earn a lower return.
The trend real rate has fallen over the past few decades across advanced economies. Much of that fall can be explained by structural factors such as an ageing global population, which — for a given real interest rate — have raised the quantity of wealth individuals wish to hold (a rightward shift in the red line in Figure A). Other structural factors, including a decline in trend productivity growth, have reduced businesses’ demand for capital (a leftward shift in the blue line). Combined, these developments have reduced the trend real interest rate required to bring actual stocks of wealth and capital into line.
Over the nearer term, the equilibrium real interest rate, r*, can fluctuate around its trend level as a result of shorter-term influences on the economy, s*. During the financial crisis, a number of headwinds to demand — including a rise in uncertainty and a tightening in the financial conditions facing households and firms — meant that the equilibrium real rate fell sharply. These headwinds to demand are taking many years to dissipate, meaning that the equilibrium real rate remains well below the trend real rate R*, which is itself well below its longer-term average.
The outlook for both the trend real rate and these shorter-term headwinds will have implications for the likely future path of Bank Rate. To the extent that the structural shifts in desired wealth and demand for capital persist, Bank Rate is likely to remain materially below the 5% level set on average by the Committee prior to the crisis. Over the shorter term, however, Bank Rate is likely to need to rise gradually as the headwinds to demand recede and the current margin of spare capacity is absorbed (see above).
Estimating the trend real interest rate, R*
The trend real interest rate, R*, cannot be directly observed and is difficult to estimate with precision. There are a number of ways in which it can be estimated. One approach is to use market-based measures implied by long-term government bond yields. As explained in the box on pages 6–7 of the May 2017 Report, ‘term structure’ models can be used to decompose these yields into expected future short-term interest rates and term premia, which are the additional compensation that investors require for holding longer-maturity assets. The average of a range of measures of the short-term interest rate component currently implies a forecast for R* of around ½% in 10 years’ time (Table 1).
Macroeconomic models can also be used to estimate how R* has evolved over the past. Table 1 presents estimates from a range of studies for different regions based on a variety of such models. These estimates span a wide range of values, in part reflecting the fact that they are likely to capture some aspects of s* as well as R*, as well as substantial variation in modelling techniques. There are few existing estimates of R* for the UK, although the openness of the UK economy means that the level of R* in the UK is likely to be highly correlated with the level in other advanced economies.
Bank staff have developed a new approach to estimating R* for the UK.3 An advantage of this approach, compared to others, is that it takes account of the fact that some determinants may overlap and so reduces the possibility that these will be double counted. One drawback, however, is that a number of simplifying assumptions are needed in order to capture these determinants within a tractable framework, and some potential determinants are not easy to model in this way. Uncertainty around the central estimates due to these assumptions, as well as the potential impacts of some omitted determinants, are discussed below.
The estimation approach follows two stages:
- First, a statistical filter is used to estimate R* over the past. Over long enough periods of time, shorter-term movements in real interest rates should average zero. The filter removes these fluctuations, leaving an estimate of R* which will reflect slow-moving structural factors. The filter cannot be used over more recent periods, however, as it relies on data over both earlier and subsequent years to provide an estimate for any given year.
- Second, an overlapping generations model is used to estimate the change in R* since 1990 and to produce a projection.4 In this model, R* is the rate at which the wealth accumulated by households is sufficient to finance the capital demanded by firms. Given the openness of the UK economy to global capital flows, global changes in desired wealth and capital are likely to be important in determining the trend real rate in the UK. The model used by Bank staff allows UK R* to depend on both global and domestic structural developments.
Results based on this approach suggest that R* in the UK has fallen by more than 2 percentage points since 1990. Allowing for uncertainty around the precise starting point and filter length, R* in real terms is estimated to have fallen from around 2¼%–3¼% (with a modal estimate of around 2½%) to around 0%–1% currently (with a modal estimate of around ¼%).5 Adding the 2% inflation target in order to convert those numbers into nominal terms results in a current estimate of nominal R* in the range of 2%–3%. 6 As explained in more detail below, shorter-term forces currently acting on the UK economy have pushed nominal r* below this level.
There is substantial additional uncertainty around these estimates, however, depending on the modelling assumptions and the paths for the main determinants of R* in the model. One calibration of those uncertainties is shown in Chart A. The red line shows the estimated probability attached to the current level of R* having a particular value, while the blue line shows the historical distribution of real interest rates over the 20th century. The leftward shift from the blue line to the red line implies that there has been a material fall in the level of R* over time. But the range of potential values of current R* covered by the red line shows that plausible changes in assumptions can translate into substantial differences in its estimated level.
Many of the factors estimated to have lowered R* over the past are likely to persist for many years, such that R* is projected to remain around its current level for some time to come. There is a wide degree of uncertainty around the persistence of the fall, however. Factors that are omitted from the model estimates may also play a role, and are discussed below.
Influences on R*
Part of the estimated decline in R* can be explained by an increase in the amount of wealth individuals wish to hold for a given real rate of interest. In turn, much of that increase has been due to the effects of changing global demographics (Table 2). Net wealth typically rises over an individual’s working life, since people tend to accumulate savings gradually in order to finance spending in their retirement. Over the past few decades, global population growth has slowed as birth rates have declined and life expectancy has increased. Together, these developments have contributed towards a rising average age of the population of advanced economies (Chart B),7 which has meant that a greater proportion of the population is at stages of life associated with higher levels of wealth. The corresponding rise in wealth holdings is estimated to have reduced R* since 1990.
The slowing in global population growth and rise in life expectancy are projected to continue, which will continue to weigh on R* relative to the past. The median UN projection implies that the average age of the population in advanced economies will rise from 41 years in 2015 to 45 years over the subsequent two and a half decades (Chart B).
The extent to which the rising average age of the global population reduces R* will depend on the degree to which older people adjust their working patterns. As explained in the box on pages 30–31 of the November 2014 Inflation Report, the proportion of older people in work has risen steadily in recent decades across advanced economies. That will have limited the number of years in retirement that individuals are likely to face, and consequently the amount of wealth required in order to finance their retirement spending. Rising old-age employment is estimated to have pushed up slightly on R*, relative to the past, partly offsetting the downward pressure from slower population growth and rising life expectancy (Table 2). In the event that future policy changes result in further rises in the average retirement age, that could also provide support to R* in the future.
Another factor estimated to have pushed up slightly on R* is a rise in the stock of government debt relative to GDP across advanced economies in recent decades (Table 2), which, for a given real interest rate, has reduced the quantity of wealth available to finance investment in the capital stock. To the extent that these higher levels of debt persist, that will continue to support R* in the future.8
Part of the estimated decline in R* can also be explained by a fall in the demand for capital for a given real rate of interest. One reason for that fall has been a slowing in the growth rate of productivity (Chart C), and hence the potential returns on new investment. To the extent that productivity growth picks up, R* will increase in coming years, all else equal.
Another factor estimated to have reduced the demand for capital — and hence R* — is a rise in the cost of financial intermediation (Table 2). Part of that rise has reflected heightened perceptions of risk following the financial crisis (see below). The rise has also reflected reforms implemented since the crisis to increase the resilience of the financial system and to end the problem of ‘too big to fail’. For example, as part of a broader set of measures, UK banks are now required to hold far greater loss-absorbing resources and their capital requirements have increased tenfold. By raising the cost of financing, these reforms are likely to require R* to be lower than in the past for a given level of demand for capital from businesses.
Desired wealth and the demand for capital may be affected by other factors that are not included within the central estimates:
i Changes in risk perceptions and risk attitudes may have reduced R* over the past. That could have stemmed from an increase in the perceived risk of large falls in output (tail risk).9 Greater risk aversion may have contributed to the build-up of savings by Asian economies following the crises of the 1990s and early 2000s, as they looked to self-insure against the possibility of similar events in future. More recently, the global financial crisis may have encouraged banks, companies and households to hold a larger proportion of their portfolio in safe assets.10
(ii) Acting in the opposite direction, the increasing financial integration of lower-income countries may push up R* in coming years. One reason for that is that the population of lower-income countries is younger on average than that of advanced economies. In addition, increased financial integration could boost R* by supporting a pickup in overall global productivity growth. While rising productivity within lower-income economies in recent decades has closed some of the productivity gap with advanced economies, there remains significant scope for further catch-up.
(iii) An increase in automation could also affect R*. On the one hand, further automation could raise the productivity of the capital stock, which could in turn raise businesses’ demand for capital. On the other hand, to the extent that automation results in a lower share of income accruing to lower-skilled employees, that could raise income and wealth inequality. Since wealthier households tend to save a larger proportion of their incomes on average, that could weigh on R* relative to the past.
The equilibrium real interest rate over the shorter term, r*
Over the shorter term, the equilibrium real interest rate, r*, can fluctuate around its trend level, R*, as a result of shorter-term influences on the economy. A temporary fall in desired spending relative to potential output, for example, would be likely to reduce the level of r* required to maintain output at its potential and to keep inflation around the target.
r* is very difficult to estimate with precision. Figure B presents a stylised illustration of how it has probably been evolving relative to its trend, R*.
During the financial crisis, spending fell sharply in the UK as uncertainty rose, growth in other advanced economies slowed, and increased risk aversion and heightened perceptions of risk on the part of investors resulted in a tightening in financial conditions. Both the public and private sectors sought to reduce their pace of borrowing. These shorter-term factors pushed r* down sharply and it became negative (Figure B).11 In such circumstances, setting the effective real policy rate below r* helps support the economy and close the output gap. Following the crisis, the MPC supported demand by cutting Bank Rate to 0.5% and purchasing assets to provide further monetary stimulus.
Many of the headwinds to spending — including the need for balance sheet repair following the financial crisis — have persisted for some time. Nevertheless, it is likely that r* has risen gradually in recent years, with easier financial conditions, a reduced fiscal drag, the end of private sector deleveraging in the UK and, in the last couple of years, stronger global growth.
The outlook for interest rates
The outlook for the equilibrium real interest rate in coming years depends on the persistence of the fall in the trend real rate, R*, as well as the extent to which the shorter-term factors (s*) currently pushing down the equilibrium real rate, r*, continue to unwind.
While any rise in trend productivity growth over coming years could raise R* somewhat, many of the structural factors currently weighing on R* — in particular, changes in demographics — are likely to persist for many years to come. That means that over the longer term, R*, and hence r*, are likely to remain lower than in the past.
In the near term, however, the MPC judges that, while recent headwinds to demand have meant that r* has remained below R*, r* is likely to rise gradually towards R* if uncertainty dissipates and as the fiscal consolidation imparts less of a drag on growth than on average in recent years.
The expected rise in r* over coming years, combined with the absorption of spare capacity over the forecast period (above), means that — even as inflation is projected to fall back towards 2% — Bank Rate is likely to need to rise gradually in order to keep inflation at the target. But the persistence of the fall in the trend real rate means that any rises in Bank Rate are expected to be limited, and interest rates are likely to need to remain low by historical standards for some time to come.
1 Indeed, over the entire period since the Bank of England was founded in 1694, Bank Rate has averaged close to 5%.
2 For more details, see Carney, M (2017), ‘Lambda’ and Carney, M (2018), ‘Guidance, contingencies and Brexit’.
3 One rationale for producing these estimates was that existing work has tended to focus either on global or US trend real interest rates. In addition, existing estimates for the UK — for example Holston, K, Laubach, T and Williams, J (2017), ‘Measuring the natural rate of interest: international trends and determinants’— have not investigated the drivers, or potential future path, of the trend real interest rate.
4 For more details on such models, see for example Lisack, N, Sajedi, R and Thwaites, G (2017), ‘Demographic trends and the real interest rate’. Demographic shifts are important determinants of R*, and one benefit of this modelling approach is that it allows different generations of individuals to be modelled separately and to interact with each other.
5 The 1990 range represents the 5th and 95th percentiles of estimates based on slightly different starting years and a distribution around the time period to which the statistical filter is applied. The range around the 2018 estimate represents the impact of using those different estimates for the starting value in 1990. Given the additional uncertainty around the estimate for 2018, the equivalent interval shown in Chart A is wider.
6 A more accurate calculation of the trend rate in nominal terms would involve adding a measure of CPI inflation expectations rather than the 2% inflation target. Note that index-linked gilts are indexed to RPI inflation.
7 For more details see Lisack, N, Sajedi, R and Thwaites, G (2017), ibid.
8 Under strong assumptions, the level of government debt has no effect on real interest rates. This ‘Ricardian equivalence’ result relies on households and firms recognising that higher government debt will ultimately be paid for by increased future taxes. In more general settings, such as the model used by Bank staff, some households and firms may not be liable for higher future taxes (for example, because they may die), and so government debt holdings represent net wealth to those households and firms.
9 For more details, see Kozlowski, J, Veldkamp, L and Venkateswaran, V (2018), ‘The tail that keeps the riskless rate low’, and Vlieghe, G (2017), ‘Real interest rates and risk’.
10 For more details, see Bernanke, B (2005), ‘The global saving glut and the US current account deficit’ and Caballero, R, Farhi, E and Gourinchas, P-O (2017), ‘The safe assets shortage conundrum’.
11 Various studies support the idea that the equilibrium real interest rate became negative during the financial crisis, both in the UK and other advanced economies. See for example Del Negro, M, Giannoni, M, Cocci, M, Shahanaghi, S and Smith, M (2015), ‘Why are interest rates so low?’ .
In the long run, real interest rates are determined by the balance between desired wealth and capital holdings
Stylised diagram of shifts in desired wealth and capital schedules
There is a wide range of existing estimates of r* and R*
Existing estimates and projections of the equilibrium real interest ratea
- a All estimates and ranges are reported to the nearest 25 basis points.
b Range based on four different term structure models. The central estimate is the average of the results from these models. See the box on pages 6–7 of the May 2017 Report for more details.
c Range based on +/–1 standard error. Figures based on the authors’ latest available estimates.
d As reported in the Federal Reserve July 2018 Monetary Policy Report.
e 95% confidence interval.
f 90% confidence interval.
There is a wide degree of uncertainty around the estimated level of R*
Estimated distribution of the trend real ratea
- a The blue line shows a smoothed distribution of the low-frequency component of actual UK real interest rates. The red line shows a smoothed distribution of the UK trend real rate generated by Bank staff’s model under the range of possible values of R* in 1990 (discussed in footnote 5), alternative assumptions for the model parameters, and the paths of the main determinants.
Changing demographics and a fall in trend productivity growth have reduced R*
Indicative contributions to changes in the trend real interest ratea
- a The direction of the arrows indicates the direction of the impact of each factor on R* since 1990, estimated using the approach outlined above. Two arrows indicate that a factor has had a relatively large impact on R*.
The average age of the population is projected to rise further
Age distribution of the population in advanced economiesa
- Source: United Nations (UN) World Population Prospects.
a Countries included are classified as ‘more developed regions’. Dashed line and faded bars are UN median variant projections.
Productivity growth in advanced economies has slowed
Productivity growth in the G7 economiesa
- Sources: OECD, ONS and Bank calculations.
a Rolling three-year averages for annual growth in output per worker in Canada, France, Germany, Italy, Japan, the UK and US.
r* fell sharply during the financial crisis
Stylised diagram of the equilibrium real interest rate
Box 7: Other forecasters' expectations
This box reports the results of the Bank’s most recent survey of external forecasters, carried out in July.1 On average, respondents had expected four-quarter GDP growth to pick up gradually to 1.7% in three years’ time (Table 1). That was unchanged relative to expectations three months ago. External forecasters’ central projections for the unemployment rate had declined (Chart A), although they remained higher, on average, than the equivalent Inflation Report forecast (above), and implied a rise in unemployment over the next three years.
External forecasters, on average, had expected CPI inflation to be at the 2% target at all horizons. And the average probability placed on inflation being more than 1 percentage point away from the target in two years’ time — either below 1% or at least 3% — had fallen to around historically low levels (Chart B).
External forecasters’ central expectations for Bank Rate at the two and three-year horizons had increased, on average, relative to three months ago. And those average central projections were higher than the market-implied path for interest rates (Chart C). As in May, almost all forecasters expected the current stock of gilt and corporate bond purchases to remain unchanged over the next two years, although the former was expected to have fallen a little by the three-year horizon.
1 For detailed distributions, see ‘Other forecasters’ expectations’.
Averages of other forecasters’ central projectionsa
- Source: Projections of outside forecasters as of 18 July 2018.
a For 2019 Q3, there were 22 forecasts for CPI inflation, 22 for GDP growth, 19 for the unemployment rate, 21 for Bank Rate, 16 for the stock of gilt purchases, 13 for the stock of corporate bond purchases and 10 for the sterling ERI. For 2020 Q3, there were 18 forecasts for CPI inflation, 18 for GDP growth, 16 for the unemployment rate, 18 for Bank Rate, 14 for the stock of gilt purchases, 12 for the stock of corporate bond purchases and 9 for the sterling ERI. For 2021 Q3, there were 15 forecasts for CPI inflation, 16 for GDP growth, 14 for the unemployment rate, 16 for Bank Rate, 11 for the stock of gilt purchases, 10 for the stock of corporate bond purchases and 8 for the sterling ERI.
b Twelve-month rate.
c Four-quarter percentage change.
d Original purchase value. Purchased via the creation of central bank reserves.
External forecasters’ unemployment projections had fallen
Averages of forecasters’ central projections of the unemployment rate
- Sources: Projections of outside forecasters provided for Inflation Reports between February 2014 and August 2018.
The average probability attached to CPI inflation being more than 1 percentage point away from the target in two years’ time had fallen
Averages of forecasters’ probabilities attached to CPI inflation outturns in two years’ time
- Sources: Projections of outside forecasters provided for Inflation Reports between February 2007 and August 2018.
Expectations of Bank Rate were higher than implied by financial market prices
Market interest rates and averages of forecasters’ central projections of Bank Rate
- Sources: Bloomberg Finance L.P., projections of outside forecasters provided for Inflation Reports in May 2018 and August 2018 and Bank calculations.
a Estimated using instantaneous forward overnight index swap rates in the 15 working days to 2 May 2018 and 25 July 2018 respectively.