Recent developments

Public finances continued to be better than expected in the final quarter of 2018, with public sector borrowing undershooting borrowing over the same period in 2017 by £13 billion and market expectations by £7 billion (figure 4). Between April and December 2018, the government borrowed £36 billion, the lowest over a comparable period in 16 years. For the financial year as whole, the Office for Budget Responsibility (Reference ChadhaOBR, October 2018) expects borrowing to fall to £26 billion, or 1.2 per cent of GDP, while independent forecasters surveyed by HM Treasury (2019) in January expect £29 billion. Public sector net debt in December remained at 84 per cent of GDP, a decrease of 0.5 percentage points compared to a year earlier. The more favourable state of public finances can in part be explained by better-than-expected labour market outcomes, which increased the tax base and reduced social security spending. It may signal that GDP growth has been underestimated. For a discussion of changes in the fiscal outlook in response to revisions to economic data and forecasts see Reference Chadha, Hantzsche, Pabst, Lazarowicz and YoungChadha et al. (2018a).

Figure 4. Surprises to public sector net borrowing in 2018–19 to date

In October, the OBR provided a revised fiscal forecast that, prior to taking new policy measures into account, implied that public sector net borrowing as a percentage of GDP would be 0.7 percentage points smaller on average than expected in March 2018 for the period 2018–19 to 2022–23. The downward revision reflected a more optimistic judgement that currently strong tax receipts as a share of economic output persist over the forecast horizon as well as more optimistic assumptions about unemployment, revised down from an average of 4.6 per cent of the labour force over 2019 to 2022 to an average of 3.8 per cent, boosting tax receipts and reducing welfare payments.

As we anticipated in our November Review, the government used the additional fiscal space to announce new spending measures, adding an average of £13 billion per year to borrowing between 2018–19 and 2023–24. The largest part of this increase reflected an increase in spending on the National Health Service in England (and knock-on effects for devolved administrations), summing up to £27.6 billion a year by 2023–24. Downward revisions to borrowing due to more optimistic economic forecasts and upward revisions due to new spending measures on the OBR's calculations nearly cancel each other out, such that forecast public sector net borrowing as a share of GDP remained at an average of 1 per cent a year, only 0.2 percentage points smaller than the OBR had forecast in March 2018 (figure 5). Public sector net debt was projected by the OBR to fall to 74.1 per cent of GDP by 2023–24, at a somewhat faster pace than previously expected.

Figure 5. Public sector net borrowing

Soft Brexit central forecast for public finances

Our public finances projection deviates from that of the OBR in two dimensions: our economic forecast and our judgement of public expenditure needs. The OBR continued to apply broad-brush judgements on productivity, trade and migration that average over different possible Brexit outcomes. By contrast, our central forecast is based on a soft Brexit accompanied by elevated levels of uncertainty during the first half of 2019 (Box A). While differences between our real GDP and inflation forecasts are small relative to the OBR's, we hold less optimistic assumptions about the unemployment rate (while forecasting somewhat stronger wage growth).

At the same time, we believe that additional spending measures may suffice to fill immediate gaps in health care spending but will not be enough to accommodate sustainably the needs of an ageing population and maintain the quality of public services in the long run (Reference Hantzsche and YoungHantzsche and Young, 2018). We therefore assume that total managed expenditure at a minimum will have to remain close to its long-run average of 38–39 per cent of GDP, rather than fall below 38 per cent as planned by the government.

These assumptions imply that, on average over 2019–20 to 2023–24, the government will have an additional £30 billion per year at hand to meet rising spending needs. It also means that the public deficit will remain elevated for longer. We forecast public sector net borrowing to stay around 2 per cent over the forecast horizon (figure 5). Public sector net debt is projected to reach around 73 per cent of GDP by 2023–24, which is similar to the OBR's outlook given that we forecast nominal GDP to be around 1 per cent higher over that period.

Risks to the forecast

Were the UK to exit the EU without a deal in March, we would expect a sharper slowdown in economic activity than in the central case. How much the economy responds would depend on the extent to which policy steps in and eases the transition to a new trading equilibrium (see discussion in Box B). Figure 5 illustrates that, all other spending commitments held equal, an active fiscal policy intervention that would lift borrowing as a share of GDP by 2 percentage points by 2023–4, or £50–60 billion, in addition to the £500 million already earmarked for Brexit preparations in the 2018 Budget for 2019–20.

Revision of the fiscal policy framework?

The government has set itself a number of fiscal targets. The targets constrain spending of government departments in the short run but are by no means necessarily optimal from a general welfare point of view and have been changed frequently in the past, reflecting different fiscal requirements in the aftermath of the financial crisis and in the face of subdued productivity growth (for the latter see research articles published in this Review). The fiscal mandate requires the government to reduce the cyclically adjusted deficit to below 2 per cent of GDP by 2020–21. On our central forecast, using current definitions of the public deficit and assuming growth close to potential, this is achievable, as is the supplementary target of having the share of net debt over GDP fall by 2020–21. However, meeting the so-called fiscal objective of achieving overall fiscal balance by the middle of the next decade does not look likely without tax increases. A no-deal Brexit would make it much harder to meet the fiscal mandate in the short run.

We have argued before that a more comprehensive review of the current fiscal framework will be needed that also includes a review of the tax system. In our view, the rise in future spending needs and a preference for balancing the budget will require additional tax revenue that cannot be raised with increases in single tax rates alone. Instead, a comprehensive overhaul of taxation and how the government charges for public services is needed that aims at raising revenue more efficiently and equitably than under the current system. The need for such a view is likely to become more urgent as a consequence of a significant change to the accounting treatment of student loans.

In December, the Office for National Statistics announced a change to the treatment of student loans in the national accounts and public finance statistics from September 2019 onwards (Reference ChadhaONS, 2018). This will have important implications for headline fiscal figures. Currently student loans enter official statistics as any other loans provided by the government. However, around 70 per cent of loans are expected to be cancelled rather than repaid, mainly in cases where graduate earnings remain below the earnings threshold for the entire 30-year period relevant for repayment. Although the expected loss is known at the outset, the current practice is to recognise the loss with a long delay and, as a result, the near-term deficit data is flattered. Furthermore, the government is able to sell off tranches of loans below nominal value without impacting government expenditure at any time. Instead, a partitioned approach will be applied in the future, treating student loans partly as genuine loans as some portion will be repaid and partly as capital transfers. While public sector net debt remains unaffected, given the cash flow is the same, the OBR (2018) estimates that this accounting change will increase public sector net borrowing figures by around £15 billion a year between 2018–19 and 2023–24. Figure 5 illustrates that our forecast for the fiscal deficit as a share of GDP would be around 0.6 percentage points larger, purely as a result of accounting changes. The government is very unlikely to tighten fiscal policy to accommodate the introduction of student loans in the deficit measure and will either raise the 2 per cent ceiling for the cyclically adjusted deficit that it has set for itself under the fiscal mandate or ignore the breach if it is temporary. The Chancellor may use the Spring Statement on 13 March to clarify his position on the fiscal rules.

News since the November forecast

The main news from the economy since our last Review concerns the near-term outlook for UK economic growth. The intensification of Brexit uncertainty appears to have played an important role, particularly for investment spending. The latest vintage of ONS data shows business investment falling in the first three quarters of 2018, implying erosion in the future supply capacity of the economy. Surveys suggest that investment spending will remain weak for as long as the nature of the future relationship between the UK and the EU is unclear. Brexit-related worries have also permeated into consumer confidence and the housing market. Consumer confidence surveys, such as from GfK, continue to trend lower and the latest RICS housing market survey also points to lower prices and transaction volumes because of Brexit.

Some of that overall weakness in output is driven by the moderation in global growth and, in particular, Euro Area growth where specific one-off factors such as the disruption caused by the new emissions test appear to have played a role. The impact on the UK is most evident in the manufacturing sector where output loss was broad-based. Output fell in all the four sub-sectors of the manufacturing sector in the three months to November according to official data – the first time this has happened since October 2012.

Annual consumer price inflation has fallen since the November forecast to 2.3 per cent and is set to drop below the target rate of 2 per cent in the first quarter of this year on our central forecast. The most proximate reason for lower inflation is the fall in oil prices. The price of Brent crude is some 30 per cent lower in sterling terms compared with September and this will have a direct impact on petrol prices.

Against this, signs of a tightening labour market are building. The employment rate is at a post-1971 high and the unemployment rate has not been lower at any time since early-1975. Annual labour productivity growth slowed once again in the third quarter to just 0.2 per cent on the output per hour worked measure – the lowest four-quarter growth rate in two years. At the same time, annual growth in regular pay has strengthened from around 2.8 per cent in the middle of last year to 3.3 per cent in the latest data; the pick-up is driven by both private and public sector pay. Unit labour costs rose by 2.8 per cent in the third quarter of 2018 relative to a year earlier, as labour costs rose faster than the rate of growth of labour productivity (as measured by hours).

Looking ahead, we expect economic growth to recover from the current soft patch if the UK continues to uphold a favourable trading relationship with the EU. This, together with the low levels of unemployment and rising wages, will require a gradual increase in the Bank Rate and as such our forecast is conditioned on a 25 basis point rise in August 2018.

Financial stability and Brexit

At its November meeting, the Financial Policy Committee (FPC) held the countercyclical capital buffer (CCyb) rate unchanged at 1 per cent. As before, Brexit remains a key risk for the FPC and, according to the FPC, the 2018 stress test showed that the UK banks are in a position to serve UK households and businesses in the face of an intense shock that could be triggered by a disorderly Brexit where the UK exits EU without a deal and without a transition period.

Box A. Assumptions about Brexit

On 15 January 2019, Parliament rejected the withdrawal agreement and political declaration negotiated between the UK government and their EU counterpart with an overwhelming majority of 230 votes. As this Review goes to press, there appear to be two broad possibilities in the short term. The first is that the UK maintains a similar relationship with the EU over the next two or three years as it does now. The second is that there is a hard Brexit. An average forecast would not sufficiently reflect either of these possibilities (cf. Reference ChadhaChadha, 2018). We therefore continue to provide forecasts for two alternative sequences of possible events. Our central forecast is based on the assumption of a ‘soft’ Brexit preceded by elevated levels of uncertainty. From the range of alternative scenarios, we also provide a forecast for the case of an ‘orderly no deal’. This box explains the economic assumptions underlying both variants and sketches how they may come about.

Soft Brexit central forecast

In this scenario, the UK continues to have barrier-free access to the EU single market and customs union over most of our forecast horizon. This scenario would be broadly consistent with a number of political outcomes, including a two-year long transition, as specified in the withdrawal agreement, or an extension of the Article 50 period, independent of whether in the long run these options lead to a comprehensive free trade agreement between the UK and the EU, continued EU membership of the UK in a customs union with the EU and/or the single market, or a reversal of the decision to exit altogether. Given that the final outcome may not be known for some time, political uncertainty remains high in this scenario, in particular during the first half of 2019.

Table A1 summarises the key Brexit-related assumptions. As a result of elevated political uncertainty, investment decisions will be held back in the earlier parts of the year. The implementation of contingency measures will continue until soft Brexit emerges as the final outcome and thus, unproductive resources will be spent and stocks built up. We further apply negative residuals to export and import equations to reflect the fact that a share of exporters will refrain from renewing contracts with European counterparts and supply chains may be thinned out. Hiring is assumed to slow down somewhat in the first half of 2019 but freedom of movement continues and in the long run labour force growth is determined by population growth which develops according to principal population projections provided by the Office for National Statistics (ONS).

Orderly no-deal variant

In this alternative scenario, the clock runs out without the government, Parliament or the EU agreeing on a deal. On 29 March, the UK reverts to trade under WTO terms for the foreseeable future. This would mean that customs duties are collected on UK borders and the regulatory regime of the EU single market no longer applies such that certain goods and services can no longer be exported. However, we assume that until the end of the Article 50 period, contingency planning will be stepped up. Both the UK government, the EU and member state governments have provided information to citizens and businesses, to assist them with withdrawal preparations (European Commission, 2018a, annex 3; Department for Exiting the European Union, 2018). In November, while it made clear that the UK would be treated as a third country, the European Commission proposed “a limited number of contingency measures to mitigate significant disruptions in some narrowly defined areas” (European Commission, 2018a), some of which the Commission started to implement in December (European Commission, 2018b). On 18 December, the UK government made no-deal preparations an operational priority accelerating legislation, and staffing and infrastructure decisions. Table A2 provides an overview of selected measures adopted or planned that will determine the economic impact of no deal in the short term. It is important to note that most of these measures would only be effective if reciprocated by both sides. A number of EU measures, for instance regarding financial services, are explicitly temporary and serve the purpose of allowing EU importers to switch from UK to EU suppliers. Putting procedures, infrastructure and staff in place to facilitate visa applications and customs checks will help mitigate temporary disruptions and queues but does not change the fact that freedom of movement will end, a number of services may no longer be tradable and the cross-border costs of goods trade will increase.

Long-run assumptions underlying our no-deal scenario are laid out in detail in Reference Hantzsche, Kara and YoungHantzsche et al. (2018).^{Footnote 1} We assume that goods trade with the EU will be 50 per cent smaller compared to continued EU membership, services trade will be 65 per cent smaller and foreign direct investment will be 24 per cent lower leading to overall business investment being 3.5 per cent lower (table A1, right column). We further assume that net migration halves and the combination of lower investment, reduced levels of international competition and the potential lack of skilled labour from abroad reduce productivity by 1.6 per cent in the long run. To reflect contingency measures announced by both sides and assuming reciprocal treatment, we phase in trade shocks gradually over 2019 and 2020 such that half of the total shock materialises within two years. We then allow for further gradual adjustment over the course of a decade to account for regulatory divergence over time. Economic uncertainty around the exit date is reflected in higher investment, equity and term premiums. The short-run impact of a no-deal Brexit depends very much on assumptions about monetary and fiscal policy which are discussed in Box B.

Notes

1 As discussed in Reference Hantzsche, Kara and YoungHantzsche et al. (2018), the baseline for our no-deal scenario is formed by a scenario in which the UK continues its membership of the European Union. Under the assumption that such a decision is made decisively in the first half of 2019, uncertainty is fully lifted, business investment rebounds from 2019 onwards, labour productivity growth picks up to just above 1.5 per cent a year and sterling appreciates to $1.40 by 2020 while growth in trade steadies. This allows us to base our no-deal assumptions on estimates about trade, foreign direct investment and productivity effects from the empirical literature which contrasts third-country status with EU membership rather than a soft Brexit case with uncertainty about trade priced on financial markets and reflected in trade and investment data.

This box was prepared by Arno Hantzsche.

Box B. Policy options for a no-deal Brexit

There is still a material risk that the UK leaves the EU at the end of March without a deal or a comprehensive transition arrangement in place. How will policymakers react to that scenario? Their response will depend on the scale and specific nature of the disruption and the reaction of financial markets to it (see also Reference ChadhaChadha, 2018). We focus here on the macroeconomics, i.e. the response of inflation and output to the Brexit shock and the mitigating action that the Chancellor and the Monetary Policy Committee might take to stabilise the economy. Our main conclusion is that policymakers have room to inject monetary and fiscal stimulus in order to stabilise output if inflation expectations and wage growth are anchored (and also thought to be anchored by policymakers) at a level that is consistent with the medium-term 2 per cent inflation target, and if fiscal rules are adjusted to allow for higher government spending. Our findings suggest that policymakers are in a position to help stabilise GDP growth in the short term but not in the medium and long term. This option would not be available in a scenario where wage growth picks up and policymakers believed that inflation expectations would be dislodged if monetary policy did not actively and immediately offset a Brexit-related spike in inflation. As such, the focus in this box lies on the short run and how policymakers can ease the transition of the economy to a new trading equilibrium by delaying some of the economic impact that is bound to materialise in the future. It complements our analysis of the long-run economic impact of a no-deal Brexit (Reference Hantzsche, Kara and YoungHantzsche et al., 2018). The long-run impact arises mainly from a slowdown in capital, employment and productivity growth and therefore leaves little room for monetary and conventional fiscal policy to respond.

We start with our central forecast which is conditioned on a soft Brexit outcome and apply a no-deal Brexit scenario that in the short term is characterised by an interruption to trade and productivity as well as a rise in risk premia (Brexit-related economic assumptions underpinning our central forecast and no-deal scenario are explained in Box A). In this scenario, the productive capacity becomes constrained immediately after exit, for instance because supply chains are interrupted and border barriers erected. Investment, interest rate and equity risk premia dampen economic sentiment and thus, aggregate demand. Policymakers have a wide range of instruments at their disposable and should deploy the tools that most effectively mitigate the dislocation. From a fiscal point of view, these tools range from tax cuts, spending measures and guarantees and from the point of the view of the central bank, there are various macro-prudential measures, Bank Rate, quantitative easing, liquidity injections, foreign currency swap lines etc. Here, we focus on taxes, transfers and Bank Rate. Using NIESR's global econometric model NiGEM, we assess the impact of these levers on inflation and GDP growth assuming that policymakers will deploy these tools depending on their perception of the size of the shock and the persistence of inflation. We present the results of four alternative scenarios: 1) a non-accommodative monetary policy response, 2) accommodative monetary policy with wage growth picking up, 3) accommodative monetary policy with restrained wage inflation, 4) additional fiscal expansion.

Scenario 1: Non-accommodative monetary policy response

As a result of interruptions to trade and lower confidence, GDP growth falls close to zero in the first two years after a no-deal Brexit (figure B1). This is partly due to supply-side constraints as productivity growth slows but also driven by a lack of demand in the face of heightened uncertainty. We would expect contingency measures to ease the initial adjustment somewhat but not prevent an economic slowdown. The effective exchange rate depreciates by around 5 per cent within a year after a no-deal Brexit, and on our analysis is 10 per cent lower than in the soft Brexit case within four years. As a result, import prices rise and consumer price inflation picks up by around 1 percentage point per annum one year after leaving the EU (figure B2). If the central bank fears that this rise in inflation would dislodge inflation expectations, it may respond mechanically to deviations of inflation from target and the fall in output relative to potential. The assumption of a mechanical response is made by the Bank of England (2018) in their assessment of the effects of different Brexit outcomes. Similar to their results, we find that Bank Rate would have to rise by 2 percentage points if it followed a standard policy rule. This, however, would exacerbate the economic slowdown (dashed red lines). The impact of automatic fiscal stabilisers would be small in this scenario. Therefore, scenarios based on the assumption of a non-accommodative policy response provide good tests of how resilient the economy is when faced with the worst case but not necessarily good forecasts.

Figure B1. GDP growth under no deal and different policies

Figure B2. Inflation under no deal and different policies

Scenario 2: Accommodative monetary policy when wages respond to higher inflation

Instead, monetary policymakers have in the past often looked through episodes of temporary inflationary pressure. For instance, in response to the financial crisis, fiscal policy provided unprecedented financial support to the banking sector while monetary policy employed unconventional tools to ease economic disruption. A more accommodative stance may be appropriate also in the event of a no-deal Brexit. In this scenario, we assume that Bank Rate would not deviate from the path we project for a soft Brexit outcome during the first year of no deal and moves only little away thereafter. In other words, the central bank takes a more accommodative stance than it would take if its interest rate policy were to follow a mechanical reaction function. As a result, GDP growth could be stabilised in the short run (solid red line) and the probability of a recession substantially reduced. The risk of such a strategy is that inflation expectations may no longer be anchored, pushing up wages and prices and keeping inflation above the Bank of England's 2 per cent target for an extended period of time.

Scenario 3: Accommodative monetary policy when wage growth is restrained

However, recent episodes have shown that nominal wages may be less responsive to economic shocks than in the past (e.g. Reference HantzscheHantzsche, 2018). Based on this evidence, we assume that nominal wage growth does not respond to the rise in inflation for an extended period of time. This would also be consistent with long-run inflation expectations being anchored. Wage setters accept reductions in real earnings for accommodative monetary policy to be effective. This is modelled by fixing nominal wages to the path they would take if monetary policy were to follow the mechanical rule described above. Under these assumptions we would expect headline inflation to subside. This would render the accommodative monetary stance effective in that initial stabilisation measures are not offset by a subsequent GDP growth slowdown (red line + crosses) and rise in unemployment.

Scenario 4: Additional fiscal loosening

Nevertheless, real disposable income will be lower than under a soft Brexit outcome, not just as a result of lower productivity growth but also because of higher import and consumer prices. This is where fiscal policy could be used to ease the burden on households. Measures could be taken that provide direct support to household income, once monetary stimulus wears off. In our analysis, we consider a combination of income tax reductions and higher transfers to households and apply it to the case where monetary policy remains accommodative while inflation expectations are anchored, assuming the government takes a more flexible approach to existing fiscal rules. We find that such a combination of expansionary fiscal measures could, while permanently increasing the public budget deficit, stabilise real disposable income, consumption and thus, GDP growth over a period of 2–3 years (red line + circles). We estimate that public sector borrowing would have to rise by 2 per cent of GDP a year to finance these policies. While such fiscal measures add up to half a percentage point to inflationary pressure, we do not find that these effects are particularly long-lasting.

This analysis suggests that a mix between accommodative monetary policy and expansionary fiscal policy has the potential to prevent the economy from a sharp slowdown in activity and should therefore be adopted in the case of a no-deal outcome, as long as wages do not respond to temporary increases in inflation and remain consistent with anchored long-run inflation expectations. It should be noted that such a policy mix will not directly resolve any disruptions to supply as a result of trade restrictions and interrupted value chains or change the fact that a no-deal Brexit would create winners and losers. But monetary and fiscal measures as ‘blunt’ instruments can be used temporarily to ease the transition of the economy as a whole to a new trading equilibrium. In the long run however, monetary and fiscal policy will not be capable of addressing structural changes arising from the new trading relationship. While leading to a somewhat smoother adjustment, expansionary monetary and fiscal policy measures would not come without a longer-term cost. As a result of looser borrowing conditions, the risk of asset price inflation rises and levels of private and public debt would increase further from currently elevated levels. Altogether this would make the economy more vulnerable to financial shocks and reduce the space available to monetary and fiscal policy to react to shocks unrelated to Brexit. The analysis could thus be extended to consider tools aimed at safeguarding financial stability, for instance using NiGEM's macroprudential modelling suite (Reference Davis, Liadze and PiggottDavis et al., 2018). The policy mix proposed here may be considerably less effective if a no-deal Brexit leads to structural disruptions to economic relationships that our modelling approach is not able to pick up.

This box was prepared by Arno Hantzsche and Amit Kara.

Equity market

Over the course of 2018, UK equity markets have tracked the European market and underperformed the US market. The FTSE All-share fell by 11 per cent over this period compared with the S&P500 which also dropped, but by just 6 per cent. Companies listed in the UK with a large sales exposure to the European Union have tended to underperform other companies. Over the past three months however, in spite of the intensification of Brexit uncertainty, the UK equity market performance has mirrored the US and Euro Area markets – each falling by around 4 per cent (figure 7).

Figure 7. European and US equity markets (local currency)

Bond market

There is, however, evidence of UK-specific Brexit stress in the sovereign credit default swap market. The cost of protecting against default has risen in the UK compared with other major economies (figure 8), but the premium remains below the levels reached at the time of the 2016 EU referendum and well below the levels reached during the global financial crisis. The sovereign bond yield curve has fallen in the UK, but not by as much as the US, implying that the spread between the US and UK bond yields has widened. For example, the benchmark 10-year yield has dropped by around 20 basis points in the UK compared with a 46 basis points drop in the US. The slope of the yield curve, as measured by the spread between the ten- and two-year gilts, is commonly considered as a lead indicator of economic growth. That spread has been falling in the UK and elsewhere in line with broader concerns about economic growth and inflation prospects. Consistent with that, market expectation of policy tightening in the UK is delayed (see Monetary Policy section).

Figure 8. Five-year sovereign CDS spread

Gilt yields are likely to have been additionally impacted by Brexit developments. Brexit uncertainty has intensified once it became clear that the government's deal was likely to be rejected by the House of Commons. That uncertainty will raise risk premiums and at the same time also raise the likelihood of an offsetting monetary policy expansion (Reference ChadhaChadha et al., 2018b).

The borrowing costs for sub-investment grade companies issuing bonds in sterling, dollar and the euro has risen in 2018 in response to economic growth concerns, the end of ECB corporate bond purchases and worries about the credit market more generally (figure 9). Concerns related to Brexit do not appear to have had a material impact on the borrowing costs of companies issuing sterling-denominated bonds however; issuance of bonds, particularly by UK companies, had dried up recently, most likely because of Brexit uncertainty.

Figure 9. BBB Corporate bond spread

Sterling exchange rate

Brexit-related news tends to have a large impact on sterling exchange rates. The currency has depreciated by more than 15 per cent since the peak in 2015 with the sharpest decline in June 2016 when the results of the EU referendum result became known (figure 10). The currency has remained relatively stable since our last forecast in November. The picture of calm that is evident in headline exchange rates, however, masks a high degree of uncertainty about the future value of the currency. This is evident in measures of volatility and risk reversals that point to expectations of large movements ahead and also a higher possibility of a depreciation rather than an appreciation.

Figure 10. Sterling exchange rate

ONS data revisions and the new narrative

The ONS published a revised set of national accounts data in December. The overall picture for headline GDP has not changed but that masks important changes to the underlying picture. The most important of these is the narrative around the rebalancing of the economy since the EU referendum. Whereas the earlier data vintage indicated a positive 0.8 percentage points contribution to GDP growth for the period starting in the second quarter of 2016 up until the third quarter of 2018, the revised data, which is still prone to revision, shows that net trade instead subtracted 0.2 percentage points over this period (figure 11). In other words, the latest data suggest that net trade did not make a positive contribution to real GDP growth in spite of a weaker exchange rate and robust growth in the Euro Area. Gross capital formation and, in particular, government investment and household consumption, were revised higher such that the overall growth picture remained largely unchanged.

Figure 11. Net trade cumulative contributions to quarterly GDP growth

The UK economy was volatile in 2018. GDP growth recovered in the second and third quarter after the weather-related disruption in the first quarter when the economy more or less stagnated. Economic growth slowed once again in the final quarter according to the NIESR Monthly GDP Tracker. Abstracting through the quarterly volatility, economic growth is set to average 1.4 per cent in 2018 if final quarter GDP growth is in line with our forecast. If correct, this will be the weakest calendar year growth rate since 2009.

Box C. Forecasting with a benchmark: the Warwick Business School forecasting system

We provide benchmark forecasts to help understand and contextualise the forecasts presented in this Review. The box presents density forecasts for UK GDP annual growth and inflation, and reports the probabilities of a range of output and inflation events occurring, as calculated using the Warwick Business School Forecasting System (WBSFS).^{Footnote 1}

To reflect the uncertainties inherent in economic forecasting, and following the practice of the NIESR and other forecasters such as the Bank of England and OBR, the WBSFS provides probabilistic forecasts. The WBSFS forecasts are produced by explicitly combining density forecasts from a set of 24, statistically motivated, univariate and multivariate econometric models commonly used in the academic literature. The use of combination forecasts or model averaging reflects the view, supported by research (e.g., see Reference Bates and GrangerBates and Granger, 1969; Reference WallisWallis, 2011; Reference Geweke and AmisanoGeweke and Amisano, 2012; Reference Rossi, Elliott and TimmermannRossi, 2013), that because any single model may be mis-specified there may be gains from the use of combination forecasts.

Comparison of the Institute's forecasts with the probabilistic forecasts from the WBSFS may be interpreted as providing an approximate indicator of the importance of expert judgement, which may include views on the underlying structure of the macroeconomy. This is because the WBSFS forecasts are computed by exploiting regularities in past data with the aid of automated time-series models; they do not take an explicit, structural or theoretical view about how the macroeconomy works; and they do not rely on (subjective) expert judgement to the same degree as those presented by the Institute. The forecasts from the WBSFS are not altered once produced; they are deemed ‘simply’ to represent the data's view of what will happen to the macroeconomy in the future.

Figure C1 presents WBSFS's latest (as of 17 January 2019) probabilistic forecasts for real GDP growth and inflation – defined as year-on-year growth rates for 2019Q4 and 2020Q4 – as histograms. The information set used to produce these forecasts includes information on GDP growth up to 2018Q3 and data on CPI inflation up to December 2018.

Figure C1. WBSFS forecast probabilities for real GDP growth and inflation, year-on-year

Table C1 extracts from these histogram forecasts the probabilities of specific output growth and inflation events. The events considered are the probability of output growth being less than 0 per cent, 1per cent and 2 per cent, and of inflation lying outside the 1–3 per cent target range (i.e., the probability of the Bank of England's Governor having to write a letter explaining how and why inflation has breached its target range). Also reported are the individual probabilities of inflation being less than 1 per cent and greater than 3 per cent, to indicate which side of the target range is most likely to be breached.

Inspection of the output growth forecasts for 2019Q4 suggests that, compared with our previous forecasts made one quarter ago, very little has changed; see table C1. The most likely outcome, with a forecasted probability of 31 per cent, is that growth continues at between 1 per cent and 2 per cent in 2019Q4. And the risk of ‘low’ growth (growth less than 1 per cent) in 2019Q4 is essentially forecast to be the same as one quarter ago: the predictive probability of this event has fallen from 28 to 27 per cent. Looking out further to 2020Q4, we observe a shift in the forecast density to one suggesting slightly higher growth. So output growth, of between 1 and 2 per cent, is now less likely at 26 per cent (versus 31 per cent for 2019Q4); and marginally more likely at around 27 per cent (versus 25 per cent for 2019Q4) to be between 2 and 3 per cent. We also observe a fairly symmetric forecast distribution for economic growth in 2020Q4, with risks evenly balanced around output growth of approximately 2 per cent.

For inflation, our probabilistic forecasts for 2019Q4 have changed between October and January only to the extent that the probability of inflation being outside the range targeted by the Bank of England has increased by a modest 3 percentage points, from 36 to 39 per cent. This change is attributable to the probability of inflation falling below 1 per cent having increased from 11 to 14 per cent. It constitutes a modest continuation of the downward movement observed in our inflation forecasts last year. An inflation rate of between 2 and 3 per cent remains the most likely outcome in the year ending 2019Q4, with a 34 per cent probability, down from around 45 per cent predicted in 2018. This slight downward trend in inflation is forecast to continue through to 2020Q4, where the probability of being outside the target range is still higher at 42 per cent, but where this risk largely comes on the downside: there is now a 16 per cent probability of inflation less than 1 per cent. The range of probable outcomes for inflation is forecast to be wider in 2020Q4 than in 2019Q4, as we also observe a small increase (of just 1 percentage point) in the forecasted probability of inflation being greater than 3 per cent.

This Box was prepared by Ana Galvão, Anthony Garratt and James Mitchell.

Note

1 WBSFS forecasts for UK output growth and inflation have been released every quarter since November 2014. Details of the releases are available at https://www2.warwick.ac.uk/fac/soc/wbs/subjects/emf/forecasting/ and a description of the models in the system and of the indicators employed is available at https://www2.warwick.ac.uk/fac/soc/wbs/subjects/emf/forecasting/summary_of_wbs_forecastng_system.pdf.

At first glance, this suggests that the economy underperformed in 2018, but on a different metric – average quarterly GDP growth over the four quarters of 2018 – the economy expanded by 0.4 per cent which is very similar to the period since 2015 and similar to our estimate of potential growth.

There have been a number of countervailing global factors at play over this period. The global economy and more specifically Euro Area GDP growth slowed sharply in the second half of 2018 (see World section). The weakness can be attributed to one-off factors such as the disruption caused to motor vehicle output by the new emissions standards, protests in France and the US government shut-down as well as more persistent drivers such as tighter monetary policy (US, Euro Area and the UK), the slowdown in China and trade barriers. There were tailwinds from oil prices, which fell by around 30 per cent since our previous forecast in November (Prices section), as well as US fiscal policy, which was expansionary for most of 2018.

The most important domestic influence behind the recent economic performance is the uncertainty around the UK's future relationship with the EU. There is clear evidence of the impact of that uncertainty on the economy in surveys as well as official data. Among surveys, the latest Bank of England Decision Maker's Panel shows Brexit as one of the top sources of uncertainty for UK companies. The latest CBI Industrial Trends survey highlights political and economic factors are amongst the main factors holding back activity over the next three months. That uncertainty, together with a potential rise in risk aversion, is also discouraging investment spending. Consistent with that message, the latest vintage of ONS data shows that business investment has fallen in each of the first three quarters of 2018 and research suggests that overseas investment into the UK is some 19 per cent lower because of Brexit (Reference Serwicka and TamberiSerwicka and Tamberi, 2018). There is also evidence of above average levels of stock building in surveys and official data as firms prepare for a disruptive no-deal Brexit.

The outlook under a soft Brexit scenario

Looking ahead, under our benign soft Brexit central scenario, we see economic growth of 1.5 per cent in 2019, revised lower from 1.9 per cent because of Brexit uncertainty and the downward revision to global GDP growth.

Household consumption is set to make the largest positive contribution to our real GDP growth forecast mainly because of rising real incomes. Official data points to a recovery in average earnings which is driven by a rising minimum/living wage and more generous public sector wage settlements (see Supply section). At the same time, inflation has surprised to the downside and is expected to remain subdued over the forecast period. Set against these tailwinds is the headwind from Brexit. Consumer confidence has been falling since the EU referendum according the GfK measure and has fallen further since our November forecast. Reinforcing this headwind is the housing market where price growth and activity, as measured by mortgage approvals, remain subdued and the household saving ratio which dropped sharply after the EU referendum in response to rising employment and lower real personal disposable income growth (figure 12). Consistent with that, total individual insolvencies rose to a 7-year high in 2018.

Figure 12. RICS Housing market survey and UK consumer confidence

We have revised lower our forecast for business investment over the period 2018–20 because of back data revisions as well as a more cautious view on negotiations that will ultimately settle the relationship with the EU. We now see business investment falling by 0.5 per cent in 2018 instead of expanding by 1.2 per cent as was the case in our November forecast; rising by 0.3 per cent in 2019 (2.9 per cent previously) and 1.3 per cent in 2020 (1.9 per cent previously). Business investment is around 4 per cent lower in levels by the end of 2020 compared with the earlier forecast.

Our judgement is that government spending will need to grow more than in the recent past to accommodate the needs of an ageing population and maintain the quality of public services (see Public finances section). As a result, we project government consumption to add 0.2 percentage points to real GDP growth in 2019, rising to 0.4 percentage points in 2020.