|
|
I’ll start with the first of those points: the economy’s recent – and welcome – resilience. In the August 2016 IR, just after the Brexit vote, the MPC forecast that the economy would slow markedly, with quarterly growth close to zero in H2 2016 and early 2017. The BoE was not alone in expecting weaker growth: the IMF and OECD made similar forecasts, and in general the consensus was slightly gloomier than the BoE (see figure 1)1 . In practice, growth has remained solid. Indeed, on the current vintage of the ONS data, quarterly GDP growth in H2 2016 was slightly stronger than H1. Compared to the BoE’s forecasts from last August, the economy’s outperformance in H2 2016 was spread widely across consumer spending, housing and business investment. Exports and imports both did better than expected, but net trade overall was less supportive than forecast last August (see figure 2). In terms of contributions to QoQ GDP growth in H2 2016, the upside in business and housing investment combined was worth slightly more than the upside in consumer spending. Real wages turned out slightly softer than expected, with slightly higher inflation and lower pay growth than forecast. Household savings have fallen more than expected. Even with weakness in retail sales volumes, most recent evidence suggests that the economy continued to grow in Q1 of this year. For example, most business surveys are around longrun averages, job growth has picked up, and NIESR estimate Q1 growth at 0.5% QoQ. The latest survey from the Federation of Small Businesses is broadly consistent with this picture (see figure 3). The BoE’s post-referendum forecast reflected three major factors. First, the judgment that the possible longrun effects of Brexit would produce a modest adverse effect on actual and potential growth nearterm. Analysis by the IMF and OECD suggests that Brexit will reduce UK potential growth slightly over time, the next 15 years or so2 . Possible factors at work include reduced trade openness, reduced inward investment, reduced competition in some sectors, lower net inward migration, and the need to reallocate resources between different sectors. Of course, the long-run effects of EU exit are highly uncertain, and perceptions may well change markedly as the details become clearer.
The BoE’s forecast assumed these longrun Brexit effects would come in linearly, hence having some effect even in the first year or two, before Brexit actually occurs. Second, and more importantly, the BoE interpreted the sharp dive in business surveys just after the referendum as indicating that heightened uncertainty over the UK’s future trading arrangements would weigh heavily on nearterm growth, notably consumer spending, business investment, housing and commercial property. This appeared consistent with evidence that uncertainty often causes businesses and households to defer major spending decisions3 . No single measure can always reflect all aspects of uncertainty. The BoE’s approach is to combine various indicators, including financial market data, surveys of households and businesses, and media citations of uncertainty4 . There was a sharp spike in this and other uncertainty gauges just after the Brexit vote5 . This was accompanied by a marked deterioration in business surveys, commercial and residential real estate markets, and other signs (eg a preference for hiring temporary, rather than permanent, staff) that seemed to reflect heightened business caution (see figure 4)6 . The BoE’s mid-August forecast put a lot of weight on that sudden deterioration, although the forecast was actually less pessimistic than the surveys implied. The third factor was the sharp drop in sterling, which is down by 16-17% since late-2015, with about two thirds of that depreciation occurring after the referendum. The BoE judged that this depreciation would lift inflation above target in 2017-19, eroding consumers’ real incomes and spending. But, reflecting Brexit-related factors (including long run issues discussed above), the forecast assumed the lift to export volumes from sterling’s depreciation would be less than usual. [...]
In that spirit, I would like to highlight several factors that probably help explain the economy’s recent outperformance. First, Brexit-related uncertainty probably mattered less than expected. The post-referendum spike in the BoE’s uncertainty index proved short-lived and this measure now roughly matches its long-run norm. This has been reflected in a commensurate rebound in business confidence and hiring attitudes, with a relatively stable housing market. Moreover, the measurement of uncertainty is itself uncertain10 . The mid-2016 jump in the BoE uncertainty index was driven heavily by components that are less well correlated to economic growth (eg media mentions of “uncertainty”), see figure 5. The uncertainty indicators that are best correlated to economic growth (eg net balance of households expecting higher unemployment) generally did not deteriorate anything like so much, and maybe should be accorded higher weight in gauging the overall level of uncertainty. Second, UK monetary conditions loosened significantly in response to the Brexit vote. The MPC cut Bank Rate and resumed asset purchases, the Term Funding Scheme helped ensure that the rate cut was fully passed through, while the FPC’s reduction in the counter-cyclical capital buffer cut risks that the referendum outcome led to tighter credit conditions. All this, plus the lift to export profitability from sterling’s sharp depreciation, probably gave some boost to demand and confidence in recent quarters. Of course, forecasts made at the time largely included these factors, but their combined effect may have been greater than expected. The boost from these looser financial conditions is not yet over. Third, global activity has been stronger, with better trends in world trade, business surveys and consumer surveys. This has given some lift to UK export prospects – but, perhaps more importantly, probably also boosted asset prices and UK growth expectations11. This is nothing to do with the Brexit vote, it just happened to occur at around the same time. Fourth, the uncertainty spike was not – unlike prior episodes -- accompanied by a major deterioration in the cost and availability of credit, partly reflecting the factors noted above but perhaps also the greater resilience of the banking system12. The BoE Credit Conditions survey suggests there has been little overall change in credit availability for households and businesses since the referendum13: the FSB and Deloitte CFO surveys confirm that picture for businesses (see figure 6). UK bank lending spreads have been roughly stable and bank lending rates remain around record lows. [...]
And fifth, it is also possible that, unconnected to the Brexit vote, the UK expansion has developed greater momentum and resilience, reflecting low unemployment, several years of solid real income growth, and a reduced emphasis on balance sheet repair following the 2008-09 recession. Aggregate corporate liquidity is relatively high, the ratio of household wealth to income is at a record high, and private sector money growth has been reasonably firm since late-2012 (see figures 8 and 9). Once the post-Brexit uncertainty spike faded, these positive underlying drivers have helped sustain demand in recent quarters. These factors, some of which are mutually reinforcing, probably help to explain why demand has held up better than expected. Many of these factors are likely to carry forward, supported also by the reduced pace of UK fiscal tightening announced in the Autumn Statement and recent Budget. As a result, the BoE has several times raised its near-term growth forecasts, and our central forecast for 2017 growth (published in the February IR) is above the external consensus. As it became clear that the economy was doing better than expected, the MPC moved from an easing bias last August to a neutral position in late 2016. Our latest forecast, based on market rates, assumes a gently rising path of interest rates. [...]