Thrifty firms and thriftless households are risks to growth

• ? Structural changes in savings behaviour by households and especially firms in advanced economies in recent years pose threats to global growth. Household savings may have been compressed by high wealth levels, pointing to the risk of a sharp rise in saving and fall in spending if asset prices correct. One positive compared to a decade ago, however, is that US personal saving is less depressed than then.
• ? The bigger risk is arguably on the corporate side, where firms' net savings have risen on average by 2–3 percentage points of GDP since the early 1990s. This has been accompanied by weakening investment, especially in net terms. The reasons behind this are varied – post‐crisis caution, demographic factors and a shift to R&D intensive industry may all have played a role. But a key factor is likely to have been changes in incentives facing executives, leading them to prioritise stock buybacks over investment. This risks creating a long‐term low‐growth feedback loop.

     

Ageing's raging savings debate

• ? Demographic changes have played a crucial role in pushing savings rates up and real rates down in the advanced economies. Despite some voices to the contrary, we think such forces will remain in place for many years to come.
• ? For such a predictable process, it's amazing that the economic implications of ageing are so hotly debated. Ageing affects everyone's lifetime savings decisions and has an impact on macroeconomic variables through several direct and indirect channels, the strength of which varies over time.
• ? The impact of ageing on savings depends on interpreting two distinct long‐term drivers. On the one hand, aggregate savings may start to fall as the baby‐boomer “bulge” in advanced economies transitions from the peak period of saving to the phase of lower saving in retirement.
• ? On the other hand, rising life expectancy should lead individuals to save more during their working lives or wait longer to retire. Greater labour market participation by those close to the official retirement age suggests that rising life expectancy is already leading many to remain in the workforce for longer – a trend that is likely to continue.
• ? Accounting for ageing and rising life expectancy together, we find there will be no major decline in savings even as the elderly's share of the population rises further.
• ? The impact on future real interest rates will, if anything, be negative. Comprehensive studies have reached a similar conclusion, taking into account the impact of demographic changes on savings, investment and other factors affecting real interest rates.
• ? Ageing populations may be complicating the escape from low global inflation. Crosscountry evidence suggests older populations may prefer lower inflation. As societies in advanced economies continue age, there is a downside risk to the long‐term outlook for inflation and bond yields.

     

Why the surge in savings won't last

• ▀ The coronavirus lockdown caused UK savings to surge. We think the household saving ratio will average over 15% in 2020 - almost twice its long-run average - while the corporate sector is likely to run a hefty financial surplus.
• ▀ Evidence suggests that economic shocks usually push up the desire to save, to the detriment of growth. But the uniqueness of the Covid crisis and its aftereffects could limit the extent of the private sector's increased prudence.
• ▀ Scarred by recent events, consumers may remain thriftier as normality returns. But the short duration of the economic contraction and the windfall nature of lockdown savings mean any long-term rise in savings rates could be modest.
• ▀ Meanwhile, post-pandemic, a more cautious attitude to investment and efforts to repair balance sheets suggest higher saving by firms. But the prospect of weak corporate profit growth will, in our view, offset those forces.

     

The consequences of pandemic-driven inequality

• ▀ According to our analysis, the Covid-19 pandemic is likely to exaggerate global inequality, leading to more aggregate debt among lower earners and higher savings for those at the top. The surge in savings will raise demand for safe assets, which would put downward pressure on long-term government bond yields - already depressed from a chronic shortage of safe assets.
• ▀ Historically, pandemics can trigger a rise in inequality, even over medium-term periods. Pandemics damage confidence in using in-person services, which disproportionately exposes low-skilled work to displacement. A unique feature of this pandemic is that the ability to work from home is proving a key factor in determining job losses - those that can are typically in higher paid jobs.
• ▀ The poorest households spend more of their income on essentials such as housing and basic food. When their incomes fall, they still have to spend on these essentials and so are often forced to take on debt. Conversely, the richest often consume near maximum capacity, so any additional income goes into savings to support future consumption.
• ▀ Higher aggregate savings would, all else equal, drive up demand for safe assets and therefore lower interest rates. Other factors such as weak nominal GDP growth, demographics and a chronic shortage of safe assets will also contribute to keeping yields depressed over the next five years.

     

Wealth spurt unlikely to offer much aid to the consumer

• ? UK households are wealthier than ever, thanks to continued growth in house prices and a buoyant stock market. However, the nature and distribution of that wealth means that support for consumer spending from a ‘wealth effect’ is likely to be both small and less than in the past.
• ? In Q4 2016, households' holdings of owner‐occupied property and net holdings of financial assets amounted to £9.2tr, almost 8% up on the level a year earlier. This was equivalent to 719% of annual household gross disposable income, a near‐record high.
• ? A long‐established feature of economics is the concept of a ‘wealth effect’ – the premise that faced with rising wealth levels, households feel more comfortable and economically secure and hence spend more. But the economic literature differs on how large this effect is.
• ? Our own Global Model suggests that the wealth effect is modest, with a 10% rise in wealth boosting consumer spending by only around 0.2%. One reason is that about half of financial wealth consists of highly illiquid assets in pension funds. But this component has recently been the biggest source of growth in wealth.
• ? Given differences in the propensity to consume out of income and wealth, the concentration of financial and housing assets among better‐off households will also act to neuter the size of any wealth effect. The wealthiest one percent of households hold around 20% of household wealth. But the bottom quartile owns only 1.5%.
• ? Meanwhile, the housing market has created an ever‐greater concentration of wealth. The share of households owning their own property fell from 71% to 63% in the decade to 2015. But the share of private renters more than doubled in the same period, from 9% to just over 19%. And the pre‐crisis appetite to finance consumption by borrowing against the value of property shows no sign of returning.

     

Higher interest rates should not unduly faze consumers

• ? Though the MPC has signalled a more aggressive pace of interest hikes than previously anticipated, we still expect the impact on the consumer sector in aggregate to be modest. We estimate that household debt servicing costs will rise from the current level of 4.1% of household income to 5.0% by the end of 2019. This would still be only a little over half the pre‐crisis peak.
• ? We expect the MPC to hike interest rates twice in both 2018 and 2019, taking Bank Rate to 1.5% by the end of next year. Higher interest rates will impact on consumer spending by increasing debt servicing costs and reducing the attractiveness of credit (including mortgages), but savers will benefit from higher returns on their deposits.
• ? Mortgages account for 77% of loans to UK households and full pass through of a 100bp rise in Bank Rate to variable rate loans, implying an increase from 2.78% to 3.78%, would add £100 a month to the cost of servicing an average mortgage. But only two‐fifths of borrowers have a variable rate deal, so for many homeowners the adjustment to higher interest rates will not be immediate. And the proportion of houses which are owned via a mortgage has fallen over the past decade, suggesting that the household sector as a whole will be less sensitive to higher mortgage interest rates.
• ? Historically the relationship between Bank Rate and interest rates on unsecured lending has been weak and rates on credit cards and personal loans have not yet risen following November's rate hike. The link to deposit rates has been stronger and higher returns on savings will mitigate some of the damage to household income from higher debt servicing costs, although uneven distribution of debt and savings means that there will be winners and losers at a more disaggregated level.
• ? We have used the Oxford Economics Global Economic Model to run a counterfactual scenario where Bank Rate is kept at 0.5% throughout 2018 and 2019. The results suggest that the pace of rate hikes assumed in our baseline forecast would reduce the level of consumer spending by 0.2 percentage points by the end of 2019.

     

Fixed versus floating: vulnerability to rising interest rates

• ? Looking at different economies' exposure to fixed‐ and floating‐rate private‐sector debt reveals how vulnerable they could be to rising interest rates. Our analysis finds that Hong Kong, Sweden, China and Australia are potentially most exposed via floating rates to rising debt service costs. A 150bp rise in rates would also push several other countries' debt service ratios above the peaks of 2008. Less vulnerable economies include the US and Germany.
• ? High levels of floating‐rate debt imply a large and rapid pass‐through of rising interest rates to firms and households, with negative consequences. Exposure to floating‐rate debt as a share of GDP varies greatly: the highest levels are in Hong Kong, China, Sweden, Australia and Spain, with the lowest levels in the US, France and Germany.
• ? Growing shares of fixed‐rate housing debt in the US, Eurozone and UK mean the impact of higher interest rates may be less severe than a decade ago. Private deleveraging in countries such as the US, UK and Spain could also soften the impact.
• ? A rise of 100bp in short‐term interest rates would raise the debt service ratio after one year by around 2.5% of GDP in Hong Kong, with increases of 1.5–1.7% of GDP in Sweden, China and Australia. The smallest effects would be in the US and Germany.
• ? A 100–150bp rate rise would push debt service ratios in China, Hong Kong, Canada, France and the Netherlands well above their peaks of a decade ago. A similar rate rise would take debt service ratios in Sweden, South Korea and Australia close to, or above, previous peaks.
• ? The distribution of debt within economies, which our analysis does not cover, is also important. For example, there is some evidence that the US corporate sector has a high concentration of debt among borrowers with weak finances. Countries that are highly vulnerable to interest rate rises may see their central banks normalise policy rates more slowly than they otherwise would.

     

Consumers alone shouldn't trigger the next recession

• ? If consumers in advanced economies were to hike their savings and rein in their discretionary spending in response to fears about the economy, we project global GDP growth in 2020 would slow to a subdued 2.1% (below our 2.5% baseline forecast), but the world economy would avoid a recession.
• ? The risks of rising precautionary savings aren't uniform across the advanced economies. For instance, the UK, Australia, Canada and Japan look particularly vulnerable to slowing retail sales, which could trigger a stronger downturn in those countries.
• ? Economies with the highest rates of consumer debt are also likely to be more at risk to weaker discretionary spending. In this respect, Australia, the Netherlands and Canada stand out the most.

     

How robots change the world

• ? The global robotics revolution is rapidly accelerating, as fast‐paced technological advances converge. This will transform robots’ capabilities and their ability to replace human workers, including in services where robot use is also set to rise steeply.
• ? The global stock of industrial robots multiplied three‐fold over the past two decades, to 2.25 million. Trends suggest it will multiply even faster in the next 20 years, reaching as many as 20 million by 2030, with 14 million in China alone.
• ? The rise of the robots will boost productivity and growth, and create new jobs, some in yet‐to‐exist industries. We estimate a 30% rise in robot installations by 2030 would create around $5 trillion in additional global GDP in today's prices.
• ? But existing business models will be disrupted and jobs lost – we estimate up to 20 million manufacturing jobs by 2030. Each new industrial robot eliminates 1.6 manufacturing jobs on average and almost twice that in low‐skilled regions.
• ? Job losses will vary greatly across countries and regions, with the toll falling disproportionately on lower‐skilled workers and on poorer local economies, aggravating social and economic stresses. The challenges for governments and policy‐makers are daunting at a time of already growing political polarisation

     

The MPC is on the verge,but a hike would be a mistake

• ? In light of the MPC's recent signalling, we now think that November's meeting will deliver a 25 basis points rise in Bank Rate. But the case for tighter monetary policy is weak even by the Committee's own criteria. This article sets out six reasons why we think that the MPC is on the verge of making an error.
• ? A chorus of hawkish comments from MPC members suggests that the Committee's next meeting on 2 November will announce a rise in Bank Rate, a 10‐year first. Granted, the MPC has ‘cried wolf’ before. But this time around the rhetoric has been much stronger. And the Committee has set the bar for a hike very low. Moreover, with inflation likely to peak soon and then decline through 2018, November would offer a good opportunity, presentationally at least, to go for a hike.
• ? But caution should temper the Committee. The MPC cites “a continued erosion of slack and a gradual rise in underlying inflationary pressure ” in support of its view. However, while the unemployment rate has fallen below the Bank's ‘equilibrium’ estimate, the Bank has history in being compelled to progressively cut that estimate. With the jobless rate still well above post‐war lows and worker power cowed, joblessness could fall further without threatening inflation.
• ? Meanwhile, the rise in underlying inflationary pressure that should follow from diminishing slack is absent. The MPC's claim that pay growth is picking up is tenuous and conflicts with recent survey evidence from the Bank's own regional Agents. In any case, the idea that faster pay growth threatens higher inflation has surprisingly weak foundations.
• ? Admittedly, a small rate rise, in itself, would slow growth only modestly. But the message sent by such an action risks pushing the economy further into a low growth expectations trap. And the Bank has alternative tools for dealing with the adverse side‐effects of ultra‐low rates. A rate rise in November would be a mistake.

     

Looser financial conditions carry a sting in the tail

• ? Against a long list of headwinds facing the economy, one apparent saving grace has been a relaxation in financial conditions since the start of 2019. Our UK Financial Conditions Index presently sits at its loosest in 18 months.
• ? Some of the lowest real interest rates in the world and a currency trading close to the weakest on record represent, all else equal, are powerful reflationary forces for the UK. But the fundamental cause of loose conditions ‐ the rising no‐deal Brexit odds ‐ mean they carry a sting in the tail.
• ? Conversely, the likely rise in market interest rates and sterling that would follow our expectation of a Brexit deal implies an overall positive for the economy could be dampened by tighter financial conditions.

     

Slow but stable(ish) house price growth lies ahead

• ? UK house price growth is running out of steam. And with household incomes squeezed and the affordability of housing stretched, we think a prolonged period of very modest growth lies ahead. But the prospect of a crash is remote.
• ? At 2.6% in Q2 2017, annual house price growth is presently running at a four‐year low. This is a step change down from the recent peak of nearly 10% in mid‐2014 and average growth of 4% over the current economic expansion.
• ? Three developments are likely to lie behind this slowdown. The first is weak growth in households' real income, cutting the ability to save for a deposit or finance a move up the housing ladder. That said, past periods of sluggish income growth have not always been associated with low house price inflation.
• ? The second is the consequence of recent tax hikes imposed on buy‐to‐let investors and second‐home owners, which theory suggests should be capitalised in lower property prices.
• ? The third and perhaps most important reason is the increasing unaffordability of housing to an ever‐widening sub‐set of the population. The ratio of house prices to earnings is almost back at its pre‐crisis record. And the income of the average mortgage borrower is close to £60,000, more than double the average annual wage.
• ? This third factor has implications beyond price growth, suggesting both a permanently lower level of transactions and a further decline in the number of households with mortgages, continuing a trend which began at the beginning of the century.
• ? But set against these headwinds is the cushion provided by record lows for both mortgage rates and mortgage affordability. Overall, house prices are caught between a lack of traditional drivers of accelerating growth, but equally an absence of forces which have typically caused prices to fall. Hence, our expectation of a period of sluggish, but relatively stable, growth.

     

Coronavirus fallout threatens financial stability

• ▪ The coronavirus is set to sharply cut global growth and also risks sparking substantial levels of financial distress among both businesses and households, potentially cascading into the banking sector. The challenges this poses for policymakers is formidable.
• ▪ Financial vulnerabilities were evident in the corporate sector in many economies even before the virus struck, with rising debt, declining credit quality and worsening liquidity positions. Consumer-facing sectors especially at risk from the impact of the virus tend to have weaker financial positions.
• ▪ There are also household fragilities. A large fraction of households - often 40%–50% - have limited liquid assets to tide them over if they cannot work. Access to sickness and unemployment benefits varies widely across economies.
• ▪ High levels of bad loans in some banking systems, most notably in Italy, could be exacerbated by the virus impact, threatening financial stability. High dollar debts in many economies outside the US are another risk factor.
• ▪ As well as containing the virus, policymakers need to consider imaginative approaches to prevent financial distress worsening the economic downturn, potentially including a need to rapidly backstop banks, firms, and households.

     

Government bond yields will remain stuck to the floor

• ▀ The fallout from the pandemic will cement the trend of safe haven bond yields remaining low over the next five years. Any rises will be gradual and limited. We project bond yields will struggle to get much above end-2019 levels by 2024.
• ▀ Despite record issuance by US, UK, German, and Japanese governments, the rise in private sector savings and demand from central banks will keep the flows of demand and supply for safe assets in broad balance. Overall, the world will still suffer a chronic shortage of safe assets, keeping yields depressed.
• ▀ Weak prospects for nominal GDP growth, demographics, and rising inequality in the aftermath of the pandemic all point to low yields. This would also be consistent with historical experience following pandemics.

     

Five shocks that could lead to recession

• ? Although world growth has slowed sharply, we are still some way from global recession territory. Additional shocks, probably in combination, would be needed to tip the global economy into recession.
• ? Of the range of possible shocks, we consider five plausible candidates: rising oil prices, a sharp slump in equities, tightening credit standards, a financial shock to emerging markets, and further escalation in trade tensions.
• ? By themselves, each would need to be very large to trigger a recession. However, a combined set of plausibly‐sized shocks could well be enough. Historically it is common for different shocks to coincide or overlap, and there is a risk of a slowdown feeding on its own momentum.
• ? In our view, the risk of these shocks occurring is not insignificant, as we've seen in recent oil price volatility and given factors such as high equity valuations.

     

Don't rely on public investment to save the day

• ▀ We think public investment in the advanced economies should be increased, but we're also sceptical that, on its own, it can save economies from recession in the event of further negative shocks.
• ▀ Deteriorating infrastructure and low bond yields make a compelling case for sustained increases in public investment. And given the shortage of safe assets and the global savings glut, higher government borrowing seems unlikely to have big negative repercussions for private sector borrowers.
• ▀ But the scope for a sustained and aggressive rise in public investment to counter shocks is limited. Even when funds are plentiful, governments often struggle to meet capital spending targets due to other constraints.
• ▀ Spikes in public investment during downturns typically dry up the capital pipeline, leading to falling investment further ahead. As a result, it's harder to sustain increases in investment, compared to other government spending.

     

Why we aren't predicting a global recession

• ? Fears that the global economy is heading into a recession are rising. But while we cannot ignore the risks that a recession could be brewing, our baseline assumption is still for a modest growth slowdown from here.
• ? The global economy is in a similar position to 2012 and 2015, as mounting uncertainties dampen growth. This time, trade tensions are a high‐profile culprit rather than the possible collapse of the eurozone or a China hard landing.
• ? In the previous two cases global growth fell to around 2.5% ‐ around the rates seen in Q2 this year ‐ only to then rebound. Our baseline forecasts assume a similar mini cycle, albeit with only a modest growth rebound.
• ? We also assume that further major adverse shocks won't materialise, and that insurance policy moves by central banks will stop a plunge in investment and households from panicking.
• ? Still, recession fears should be taken seriously ‐ slowdowns can become self‐perpetuating. Once annual GDP growth has fallen by over 1ppt from its peak, the eventual decline typically ends up being much larger ‐ of the seven growth slowdowns since the late 1970s where annual growth slowed by over 1ppt ‐ four resulted in either a global recession or only a narrow escape from one.
• ? With US‐China tensions unlikely to recede and factors like the US yield curve inversion adding to the air of gloom, the latest downturn could gain momentum.
• ? Although reduced macro volatility and anchored inflation have made it easier for policymakers to deliver soft landings, the effectiveness of monetary policy has waned. And with China no longer acting as spender of last resort, it's vital that governments in advanced economies stand ready to pick up the slack

     

After the pandemic: Medium-term growth uncertainties

• ▪ The coronavirus is having a very large short-term negative impact on world growth. But the medium-term growth outlook is more uncertain. Much depends on the policy response - a strong and well-designed response could contain the medium-term output losses, but large and enduring damage is a risk.
• ▪ Growth patterns after significant GDP declines vary. Historical evidence points to large upfront effects from pandemics and natural disasters, but the medium-term outcomes are mixed, with policy responses a crucial determinant. Longer recessions and financial crises tend to lead to weaker medium-term growth.
• ▪ The coronavirus may trigger annual GDP declines among the worst seen in the last 100 years. Economies can bounce back sharply after such declines, but our analysis suggests output losses also endure in a significant number of cases.
• ▪ The US interwar experience shows the danger of allowing financial distress to snowball and exacerbate GDP declines. The weakness of medium-term growth after the global financial crisis confirms the long-term impact of such distress.
• ▪ A key risk for the medium term is that firms and/or households react to the coronavirus recession and disruption by raising savings, accelerating a shift toward ‘Japanification’ of major economies.
• ▪ Our baseline forecasts envisage moderate medium-term output losses due to coronavirus, in part reflecting rapid and large-scale policy interventions. But uncertainty around this forecast is significant and considerable variation is possible across economies.

     

World Economic Prospects Monthly

Overview: Reflation enthusiasm is tempered

• ? We have kept our world GDP growth forecasts unchanged this month, at 2.6% for 2017 and 2.9% in 2018. But our outlook for inflation has been lowered to 3.0% this year (from 3.3% last month) as inflation is close to a peak in several economies and oil prices have fallen recently.
• ? Global indicators continue to point to buoyant activity, driven by manufacturing. The global manufacturing PMI rose to its highest level in almost six years in February, which in turn is boosting world trade. Despite the exuberance shown by the surveys, we remain cautious. We continue to expect a slowdown in consumer spending as households are squeezed by higher prices.
• ? Although we still see GDP growth in the US accelerating this year, we have lowered our forecast to 2.1% as economic data have been weaker than expected at the start of the year. Large uncertainties around our central forecast persist given the unpredictability of President Trump's policies, and markets have tempered their initial enthusiasm regarding the success of ‘Trumponomics’.
• ? With the Federal Reserve now close to meeting its dual mandate, the pace of policy normalisation will accelerate. We now expect the Fed to raise interest rates this month and three times overall this year. This means that US bond yields are likely to continue to rise and the euro will remain under pressure due to the widening interest rate differential between the US and the Eurozone.
• ? The Eurozone economy remains resilient ahead of key elections in France, the Netherlands and Germany. Our view remains that populist fears are overstated and that Emmanuel Macron is still favourite to become the next French president.
• ? Many emerging markets have started 2017 with positive momentum, but caution remains the name of the game as the Fed prepares to raise rates faster than previously expected and the future of US trade policy remains uncertain.

     

Can Germany save itself from negative rates?

• German inflation looks set to rise in response to diminishing slack in the economy. But this will be a mixed blessing for those in Germany hit by negative policy rates and ECB asset purchases. Higher German inflation may eliminate the need for further ECB policy action, but it is unlikely to trigger imminent rate hikes. As a result, the rise in inflation will merely lower real interest rates for German savers.
• Structural cross‐country differences mean that the ECB is better able to hit its inflation target when the peripheral economies rather than Germany are the region's growth engine. A key reason for this is that the German Phillips curve is flat by Eurozone standards, meaning that policymakers need to work hard to generate sufficient inflation in Germany to offset sustained weakness elsewhere.
• Despite this, there is evidence to suggest that the tightening labour market is beginning to push German wage growth higher. And if productivity growth remains subdued, this will lead to faster unit labour cost growth.
• While firms could respond by lowering their margins, the strength of household spending suggests that firms may be more inclined than in the past to pass on higher costs to consumers.
• In all, we expect German inflation to rise more sharply than elsewhere to around 2% in 2017, meaning that the ECB will not unveil further unconventional policy support. But it would take much sharper rises in German wage growth and inflation than in our baseline forecast to prompt the ECB to bring forward interest rate rises.