In focus - Uncertainty and Brexit

Section 4 of the Monetary Policy Report - November 2019

Uncertainty about future outcomes is an important driver of economic behaviour, over and above central expectations. The Brexit process has already affected the UK economy. It has made some firms and households more pessimistic about the central outlook. It has also increased the uncertainty around that central outlook. Bank research suggests that these Brexit effects have depressed investment spending and weighed on productivity. The MPC’s latest projections assume that the progress of the Withdrawal Agreement removes some uncertainty. But some is likely to persist while the deal and the transition to it are negotiated.

People’s expectations about the economic outlook are important for spending and investment decisions. For example, people might spend less now if they think that their income is likely to be lower in future. In addition, there is a deep and long-standing literature showing that the degree of uncertainty around those expectations also has an important influence on behaviour. Higher uncertainty tends to weigh on investment and consumption (Chart 4.1), especially the former, and can reduce productive capacity.

Brexit will fundamentally change the nature of the UK’s relationship with its largest trading partner. The wide range of potential outcomes appears to have both increased uncertainty (Chart 4.2) and made people more pessimistic about the economic outlook. Those effects, which are difficult to separate, are already influencing the UK economy. They have lowered business investment in particular, and may have weighed on productivity and consumption.

This section summarises how uncertainty can affect the economy in principle (Section 4.1). It then assesses the impact of Brexit on indicators of uncertainty (Section 4.2) and how this has affected the UK economy (Section 4.3). Finally, it sets out how uncertainty is assumed to evolve in the MPC’s forecast (Section 4.4).

Chart 4.1 Uncertainty has a close relationship with spending

Annual growth of household consumption and business investment, and a measure of uncertainty

Sources: Bloomberg Finance L.P., CBI, Consensus Economics, Eikon from Refinitiv, GfK (research on behalf of the European Commission), Institutional Brokers’ Estimate System, ONS, and Bank calculations.

(a) Chained-volume measures. Business investment data are adjusted for the transfer of nuclear reactors from the public corporation sector to the central government in 2005 Q2.
(b) The first principal component extracted from the set of indicators: the average monthly standard deviation of external forecasts for GDP growth one and two years ahead; the standard deviation of analysts’ forecasts for corporate earnings growth over the next year; CBI survey measure of demand uncertainty as a factor likely to limit capital expenditure for manufacturing and services; an index of UK policy uncertainty based on newspaper articles; household survey responses on their personal financial situation and unemployment expectations; the three-month option-implied volatility for the FTSE 100 (realised volatility used prior to April 1992); a weighted average of the three-month option-implied volatility of the sterling-euro and sterling-dollar exchange rates. Data are shown to Q2.

Chart 4.2 The proportion of firms that cite Brexit as an important source of uncertainty is elevated

Brexit in top three current sources of uncertainty (a)

Sources: Decision Maker Panel (DMP) Survey and Bank calculations.

(a) Question: ‘How much has the result of the EU referendum affected the level of uncertainty affecting your business?’. Respondents can select: ‘Not important’; ‘One of many sources’; ‘Two or three top sources’; or ‘Top source of uncertainty’. Before August 2018, data are interpolated between waves and shown as three-month rolling averages. The DMP currently consists of around 8,000 businesses with around 3,000 responses a month being received.

4.1 The impact of uncertainty on the economy

Economic behaviour is forward looking.

When firms and households make decisions about investment and spending, they take the future into account. Their central expectations about the economic outlook — what they think is most likely to happen — are important. Uncertainty around those expectations also affects decisions being made now.

Uncertainty weighs on investment…

Uncertainty creates a value in waiting for news that might make the future outlook clearer. This means firms are more likely to delay investment decisions where the return will vary depending on how the economy evolves (see Bernanke (1983) and Dixit and Pindyck (1994)). Investment typically involves sunk costs and changes to processes, for example from installing a new production line. That makes it costly to reverse, and this irreversibility creates an incentive to delay decisions until more is known about the future, protecting firms from bad outcomes. The value of delaying is greater the worse the potential bad outcomes are.

The incentive to wait might be particularly strong if firms expect uncertainty to be resolved soon. Delaying investment until uncertainty falls comes with a cost. Firms forego the potential profits from the investment while they wait. The longer they wait, the higher the cost. Therefore delaying becomes more reasonable if uncertainty is expected to last only a short time.[1]

…including through its influence on financial conditions…

As uncertainty increases, investors are likely to demand a higher compensation for risk, causing credit conditions to tighten (see Whaley (2000) and Gilchrist, Sim and Zakrajšek (2014)). This can lower spending as credit becomes less readily available. But causality can also go the other way. A shock which tightens credit conditions can also lead to heightened uncertainty about the economic outlook. That occurred during the financial crisis, for example. As a result, it is important to separate the effect of changes in uncertainty from the effect of other shocks. In some of the empirical research, the estimated impact of uncertainty on demand falls once the effect of changes in credit conditions has been taken into account.[2]

…and can also reduce productivity and supply growth.

If investment is delayed, it affects the amount firms can produce in future through lower capital deepening and research and development (see Bonciani and Oh (2019)). Uncertainty can also cause firms to postpone hiring and firing decisions and can make workers reluctant to seek new jobs. The resulting fall in labour market churn could lower productivity if it results in labour being misallocated across firms (see Lazear and Spletzer (2012)).

Uncertainty might also affect consumption.

Higher uncertainty can cause risk-averse households to cut back on consumption temporarily (see Leland (1968) and Carroll (1997)). The literature suggests that uncertainty about job prospects is particularly important for household behaviour, as it encourages precautionary saving (see Benito (2004)). Households might delay spending that involves large costs, such as buying a house or car, similar to firms delaying investment. But because the majority of day-to-day household spending does not involve large one-off costs, the effect of uncertainty on total consumption may be smaller than for business investment.

4.2 The impact of Brexit on uncertainty

Brexit has increased uncertainty.

Brexit will fundamentally change the nature of the UK’s relationship with its largest trading partner. The impact of Brexit on indicators of uncertainty has been evident since the referendum. At least 30% of firms have cited Brexit in their top three sources of uncertainty in the Bank’s Decision Maker Panel (DMP) Survey since it began in 2016 (Chart 4.2). This has risen to around 55% of firms in more recent surveys. Brexit uncertainty has been widespread, including for firms which are not reliant on sales to the EU (Chart 4.3).

Chart 4.3 Brexit uncertainty has picked up for all firms, not just exporters to the EU

Brexit in top three current sources of uncertainty, by proportion of sales accounted for by exports to the EU (a)

Sources: DMP Survey and Bank calculations.

(a) See Chart 4.2 footnote.

Chart 4.4 Some measures suggest that uncertainty is close to post-crisis highs

Selected measures of uncertainty

Sources: Bloomberg Finance L.P., CBI, Consensus Economics, Deloitte, Eikon from Refinitiv, GfK (research on behalf of the European Commission), Institutional Brokers’ Estimate System, ONS, and Bank calculations.

(a) See Chart 4.1 footnote (b). Differences from average for principal component are since 1988.
(b) Monthly weighted average of the three-month option-implied volatility of the sterling-euro and sterling-dollar exchange rates. Series starts in September 2001.
(c) Proportion of firms reporting that the general level of external financial or economic uncertainty facing their business is ‘high’ or ‘very high’. Series starts in 2010 Q3. Not seasonally adjusted.

This is reflected in some more general uncertainty indicators…

As well as the DMP measure of Brexit-specific uncertainty (Chart 4.2), some indicators capturing general uncertainty have risen too. For example, an above-average proportion of respondents to the Deloitte CFO Survey — more than half — have reported high uncertainty in the past four quarters (Chart 4.4).

The implied volatility from sterling options — which captures perceived uncertainty around the exchange rate — has been elevated recently (Chart 4.4). This measure is likely to capture both political and business cycle uncertainty. It is suited to identifying UK-specific shocks, like Brexit, because the exchange rate reflects beliefs about relative economic prospects. It also has a historically reliable relationship with UK business investment and GDP growth.

In contrast, some other uncertainty indicators are less elevated. The range in Chart 4.4 shows a broad set of measures, summarised in a principal component.[3] The principal component is currently around its historical average. This might be because some indicators do not capture Brexit uncertainty fully. For example, one measure is based on a range of external forecasters’ central expectations for GDP growth. This might fail to pick up the degree of uncertainty that each forecaster has around their central projection.

…which suggest uncertainty is close to post-crisis highs.

Sterling implied volatility and uncertainty among CFOs rose substantially in the run-up to the Article 50 deadline in March 2019 (Chart 4.4). The proportion of firms which place Brexit in their top three sources of uncertainty also increased ahead of the March deadline and has remained elevated since (Chart 4.2). These measures suggest uncertainty in 2019 Q3 was close to post-crisis highs.

4.3 How has Brexit uncertainty affected the economy?

As well as increasing uncertainty, Brexit has made some households and firms more pessimistic…

Brexit appears to have made households and businesses more pessimistic about the economy, on average. The DMP Survey suggests that, on balance, firms expect that Brexit ultimately will have a negative impact on their sales (Chart 4.5). Household expectations for the general economic situation a year ahead have deteriorated since 2016, although confidence in their own financial situation has been less affected (Chart 2.18).

…and it is difficult to separate the impact of these two effects on the economy.

It is difficult to disentangle the influence of higher uncertainty from the effects of increased pessimism. Both can have similar effects on behaviour, weighing on consumption and investment. They have also moved together: the impact that firms expect Brexit will have on their sales is highly correlated with their uncertainty about Brexit (Chart 4.6). As a result, the estimated impact of Brexit uncertainty on the economy set out below might also be capturing some effects from increased pessimism.

Chart 4.5 On balance, firms expect Brexit to have a negative effect on their sales

Firms’ expected eventual impact of Brexit on sales (a)

Sources: DMP Survey and Bank calculations.

(a) Question: ‘How do you expect the eventual Brexit agreement to affect your sales once the UK has left the EU, compared to what would have been the case had the UK remained a member of the EU?’. Respondents are asked to provide the probability they place on each option, where a ‘large’ effect is more than 10% of sales. Responses collected between August and October 2019.

Chart 4.6 Firms with high uncertainty about Brexit expect it to weigh more on their sales

Firms’ expected eventual impact of Brexit on sales, by Brexit uncertainty (a)

Sources: DMP Survey and Bank calculations.

(a) See Chart 4.2 footnote for question about Brexit as a source of uncertainty and Chart 4.5 footnote for question on the expected impact on sales. Point estimates are constructed by attaching midpoints of 5% and 20% to the response categories for a ‘less than 10%’ and ‘10% or more’ impact respectively. Responses collected between August and October 2019.

Brexit has been a key factor in the stalling of business investment.

Heightened uncertainty has weighed on business investment in the UK. Over the past three years, cumulative growth in investment spending has been just 0.4%. This could be partly explained by slowing global growth, but investment has been weak even relative to other major advanced economies (Chart 2.15).

The biggest falls in investment growth have been by firms reporting high uncertainty about Brexit (Chart 4.7). Recent research (set out in Bloom et al (2019)) suggests that the level of business investment was around 11% lower in 2019 Q2 as a result of Brexit (Chart 4.8).

Business investment was still growing in the period soon after the referendum, albeit at a much slower rate than before, but it has fallen in almost every quarter since 2018. This further weakening could be partly explained by firms waiting to see if some uncertainty would be resolved by the March and October 2019 Brexit deadlines.

Chart 4.7 Investment growth by firms that are more uncertain about Brexit has fallen since the EU referendum

Average annual investment growth for firms, by Brexit uncertainty (a)

Sources: Bureau van Dijk, DMP Survey and Bank calculations.

(a) Sample uses DMP data where available (all post-referendum) and company accounts from Bureau van Dijk otherwise. See Chart 4.2 footnote for question about Brexit as a source of uncertainty. ‘High’ uncertainty is defined as placing Brexit in the top three sources of uncertainty. Data are unweighted averages across firms.

Chart 4.8 Uncertainty has weighed on business investment

Business investment and indicative estimate without the effect of the Brexit process

Sources: Bloom et al (2019), DMP Survey, ONS and Bank calculations.

(a) Counterfactual is based on estimates of the annual Brexit impacts set out in Table 3 in Bloom et al (2019). Annual impacts are linearly interpolated for quarterly figures. The swathe illustrates the 90% confidence interval.
(b) Chained-volume measure.

Consumer spending has been less affected so far…

Consumption has been more resilient than investment. While household expectations for the general economic outlook have fallen, uncertainty about job prospects has been more stable (Chart 2.18), and this is one of the most important channels through which uncertainty can affect spending. Spending has also been underpinned by real income growth and accommodative credit conditions.

…but there have been some signs of uncertainty dampening consumption.

There is some evidence that uncertainty has affected large purchases and discretionary spending by households. The housing market has been subdued since 2016 (Chart 2.19). Surveys suggest this is at least partly due to Brexit, as households delay purchases given the large costs involved. Uncertainty could also partly explain why car purchases have been weak, although changes in emissions regulations have made recent data difficult to interpret. More broadly, consumer spending that might be seen as non-essential[4] has risen by just 0.5% over the past year, the weakest growth since 2011. In addition, the saving ratio and household financial balance have drifted up a little over the past couple of years, which might suggest some precautionary saving.

Uncertainty may also have affected supply capacity.

Evidence from the DMP Survey suggests that the Brexit process has reduced the level of UK productivity by 2%.[5] Most of this effect comes from reduced productivity within firms. This could be a result of firms preparing for Brexit, which may have diverted resources away from productive output or making improvements. Firms with higher uncertainty about Brexit have spent more resources on planning (Chart 4.9). Lower business investment due to heightened uncertainty may have also weighed on labour productivity through reduced capital deepening.

Chart 4.9 Firms that are more uncertain about Brexit have spent more on Brexit preparation

Brexit-related spending over the past three years, by Brexit uncertainty (a)

Sources: Bureau van Dijk, DMP Survey and Bank calculations.

(a) Question on spending: ‘Approximately how much do you estimate that your business has spent on preparing for Brexit so far?’. See Chart 4.2 footnote for question about Brexit as a source of uncertainty. Responses collected between August and October 2019.

Chart 4.10 More productive firms tend to be more uncertain about Brexit

Brexit uncertainty, by average level of productivity (a)

Sources: Bureau van Dijk, DMP Survey and Bank calculations.

(a) Productivity is defined as real value added per employee. The Brexit Uncertainty Index shows the percentage of firms which reported that Brexit was in the top three current sources of uncertainty for their business, September 2016 to June 2019 average.

There may have also been a small effect on average productivity through changes in the relative growth rates between firms. Firms which were more productive before the EU referendum tend to report higher uncertainty about Brexit (Chart 4.10). The research suggests the Brexit process has weighed on the growth of more productive firms to a greater extent. Exporters tend to be more productive and demand for exports might have fallen by more than for domestic goods in response to Brexit. Intelligence from the Bank’s Agents suggests uncertainty has lowered export demand, as some overseas customers are reluctant to enter new deals until there is greater clarity about future trading arrangements (Box 3).

4.4 Uncertainty in the MPC’s forecast

The MPC projects that some uncertainty will be removed in the near term…

In October, the UK and EU agreed a Withdrawal Agreement and Political Declaration, and the UK House of Commons approved the second reading of the Bill which implements the agreement into law. The UK’s EU membership was also extended by up to a further three months. Those developments are likely to reduce the uncertainty that has been facing households and businesses in the near term. In part that will reflect a lower perceived likelihood of a no-deal Brexit. For example, betting odds suggest that the probability of a no-deal Brexit in 2019 has fallen markedly. Responses to the DMP Survey also suggest that the average likelihood that firms attach to a no-deal Brexit in 2019 fell after the second reading of the Withdrawal Agreement Bill was passed.

…though it remains somewhat elevated over the forecast period.

Some uncertainty is likely to persist, however. Brexit is a process rather than a single event. While the agreement sets out the broad parameters of the UK and EU’s future trading relationship, the range of potential outcomes is still relatively wide. Most companies reported that uncertainty was high in Q3 (Chart 4.11). And the proportion of firms expecting Brexit-related uncertainty to last until after next year has picked up recently (Chart 4.12). Uncertainty is assumed to decline gradually over the forecast period, as the details of the UK and EU’s eventual relationship emerge over time.

The reduction in uncertainty supports the recovery in UK demand growth.

As the dampening effect from Brexit-related uncertainties begins to dissipate, GDP growth is projected to pick up moderately. The decline in uncertainty is assumed to support business investment in particular. Four-quarter business investment growth is projected to rise from -1½% in 2019 Q2 to around 4% in 2021 (Section 1).

Chart 4.11 More firms have been persistently uncertain

Proportion of CFOs reporting ‘high’ or ‘very high’ levels of uncertainty (a)

Sources: Deloitte and Bank calculations.

(a) See Chart 4.4 footnote (c) for survey question.
(b) Firms reporting uncertainty is ‘high’ or ‘very high’ for four or more consecutive quarters.

Chart 4.12 Firms have been pushing back the date they expect Brexit-related uncertainty to be resolved

Date by which Brexit-related uncertainty is expected to be resolved (a)

Sources: DMP Survey and Bank calculations.

(a) Question: ‘When do you think it is most likely that Brexit-related uncertainty facing your business will be resolved?’. Data are for businesses that state that they are affected by Brexit-related uncertainty.

  1. This channel is covered in more detail in Broadbent, B (2019), ‘Investment and uncertainty: the value of waiting for news’.

  2. See Forbes, K (2016), ‘Uncertainty about uncertainty’.

  3. The principal component summarises the signals from the range of measures into the single variable that accounts for the greatest amount of covariation between them. For more details see Haddow, A, Hare, C, Hooley, J and Shakir, T (2013), ‘Macroeconomic uncertainty: what is it, how can we measure it and why does it matter?’, Bank of England Quarterly Bulletin, 2013 Q2.

  4. Spending excluding: most food and non-alcoholic beverages; housing, water and energy costs; repair of household appliances; non-durable household goods for routine maintenance; dwelling and transport insurance; and financial services not elsewhere classified.

  5. The research gives a range of 2%–5%. The estimate of 2% is calculated by weighting the results for each firm by its size, whereas 5% gives all firms equal weight. For this reason, the aggregate effect on UK productivity is likely to be closer to 2%. For more details, see Bloom et al (2019).

This page was last updated 31 January 2023