How is this recession different?


We are in the midst of an unprecedented global recession triggered by the Covid-19 virus. The speed, scale, and breadth of the contraction in economic activity are unlike anything we have seen in the post-war era. However, we can still look to history to provide some guidance on how this recession may propagate through the economy, how policy can and should respond, and what the long term implications of the shock may be.

In recently published research exploring whether economies experience “stall speed” and “escape velocity”, we built and analysed a dataset covering developed market recessions in the post-war period. We use these data in this note to say something about what a “typical” recession is like in terms of magnitude, duration, cause and consequence, and compare to the current contraction.

The vast majority of recessions haven been caused by one of three main types of shock. Supply-driven oil price shocks triggered by production cuts; monetary policy tightening in response to inflation; and financial shocks characterised by the unwinding of large-scale imbalances. Inventory swings have played a role in exaggerating downturns as well, especially before modern supply chain management techniques. Most recessions are of short duration and shallow depth, but there is a long tail of more severe downturns, especially when they involve the unwinding of financial imbalances. Pandemics are implicated in some global downturns, and typically involve a steep drop in the level of economic activity, but strong rebounds subsequently.

The current crisis is a deliberate policy-induced cessation of economic activity to suppress the spread of the Covid-19 virus. The resulting recession is characterised by a demand, supply, and financial market shock occurring concurrently, interacting with and magnifying each other. Social distancing measures mean vast swathes of economic activity simply cannot occur. But at the same time economic activity is stopped, debts still need serviced. This mismatch in income and liability servicing presents a liquidity problem that can easily become a solvency problem. This in turn forces up risk premia and causes a sudden fall in asset values, tightening financial conditions and pushing down on demand still further.

For all that this shock is uniquely severe, it is possible to see light at the end of the tunnel and have some confidence in a robust rebound in activity levels. The example of some East Asian economies, where the virus is (seemingly!) under control shows how suppression can work and eventually, albeit partially at first, be lifted. And the size of offsetting monetary and fiscal policy responses almost everywhere should help to cushion (but not offset entirely) the extent of permanent supply destruction during the current recession and eventually stimulate a recovery in demand.

However, it is not difficult to foresee a more adverse scenario with a deeper recession and a more limited recovery. Western economies may not be able to follow the same exit path as Asian economies where suppression measures began earlier in the outbreak and adherence has been higher; while the latter may yet experience a second wave of infections. And for all their speed and size, monetary and fiscal policy responses may just be insufficient to protect the economy from long-term scarring. Meanwhile, our work on stall speed shows that self-perpetuating downturns are a feature of economic activity, where weakness begets further economic weakness, perhaps through adverse feedback loops. That same research also found that escape velocity bounces out of recessions are few and far between; instead the legacy of recessions tends to be sustained weakness in economic growth even after recovery has set in.

Given that all significant and disruptive recessions have had long-lived consequences, we look to the potential long-run implications of this crisis. More activist fiscal policy, funded in times of crisis by monetary policy; deeper cleavages in the US/China relationship; a bigger, more interventionist state, especially when it comes to personal surveillance; further moves towards deglobalisation; more economic activity taking place online; and a potential shift in the politics around climate mitigation policies, are all explored as potential outcomes.

These many changes to the economic and political environment will have important implications for return drivers of financial assets. The equilibrium, or natural rate, of interest is likely to fall even lower, continuing the secular trend characterising the interest rate environment for decades. This means lower returns across asset markets, both risky and risk-free. However, from where we stand right now, the steep correction in equity market prices means that expected returns are actually boosted by attractive valuations. This is especially true if markets tend to overshoot when responding to incoming information in part due to behavioural biases and in part to provide the steep discounts required to incentivise investors to once again hold riskier assets. Meanwhile, while the most likely outcome of the crisis is a sustained period of low inflation, the opposing tail risks of a deflationary trap and of much higher inflation have probably both increased in probability. Given the increased risk of both these scenarios, the inflation risk component of bond prices and asset prices more generally is likely to increase as this reflects the volatility and unpredictability of inflation. Ultimately, though, we don’t expect what is a health and economic crisis to become a financial crisis.




Download this article as a PDF

Download  arrow

Paul Diggle
Luke Bartholomew


Lessons from history

Chapter 1

Any comparison of how the current recession is different from the historical experience must start with a brief consideration of that historical experience. In this chapter, we review the causes and consequences of past recessions, concluding with a particular focus on pandemic disease outbreaks.

Our starting point is the dataset of post-war recessions in the 36 OECD economies that we collated for our recent CEPR Discussion Paper, Stall Speed and Escape Velocity: Empty Metaphors or Empirical Realities?, December 20191. This dataset yields 152 recessions in total. These recessions are, in general, highly grouped in time. Two-thirds of all post-1950 OECD recessions occurred within 12 distinct historical clusters (see Figure 1). Many major economies have only ever been in recession during these wider periods of global economic stress. Contractions in these economies could thus be a sign of a wider malaise. This includes the US, which itself has been a major propagator of these downturns.

Major historical pandemics involving at least 100,000 deaths have long-lived labour market and interest rate implications.

Figure 1: Share of OECD economies in recession

Source: Thomson Reuters Datastream, Aberdeen Standard Investments

What is the typical recession like?

On average, these recessions last 3.4 quarters and subtract 4% from the level of GDP. But these averages mask considerable variation. Most recessions are actually just two quarters in length and see peak-to-trough declines in GDP of 2% or less. But the longest recessions in our sample - Greece and Spain during the sovereign debt crisis of 2010-12 - lasted fully three years; while the deepest recessions are the aforementioned downturn in Greece, and in Latvia and Estonia during the global financial crisis of 2007-09, all of which subtracted 20% or more from GDP.

Figure 2 plots the relationship between recession length and depth across our sample. The bulk of recessions are in the bottom left hand corner, lasting 2-4 quarters and subtracting 4% or less from GDP. But there is a long tail of more severe recessions.

Figure 2: Length and depth of 152 post-1950 OECD recessions

Source: Thomson Reuters Datastream, Aberdeen Standard Investments

What causes recessions?

Pre-industrial revolution, most economic downturns were simply the result of bad harvests, but the role of agriculture in driving the economy is clearly much lower now. Downturns that weren’t triggered by the vagaries of the weather often had their cause in mass-mobilisation military conflicts, including those of the early 20th century, which are hopefully a thing of the past as well. Large swings in inventories have also played a role in exacerbating historical downturns, but have become less important in recent history as inventory management techniques improved (it’s an interesting question, considered in Chapter 4, whether inventories will be systematically higher after this current shock and hence more prone to driving the business cycle). A small group of recessions were the result of pandemic disease outbreaks, and these are considered in the next section. Meanwhile, recessions in emerging market economies typically have a different set of triggers, including sudden stops in external funding and disorderly movements in exchange rates.

Three common causes are implicated in a disproportionate share of recessions from the post-war developed economies in our sample, however: supply-driven oil price shocks triggered by production cuts and causing a big hit to real incomes; monetary policy tightening in response to inflation pressure; and financial shocks characterised by large-scale imbalances and amplification through the banking sector. Oftentimes, these causes are present in different combinations, and interact in interesting ways.

Clusters of recession occurred across the developed markets in 1974-75 and 1980. These are the most obvious examples of oil price shocks driving downturns. As a result of the Yom Kippur war in the first instance, and the Iranian Revolution in the second, real oil prices rose sharply. This caused a large increase in energy prices in many economies (in excess of 30% y/y in the G7 - Figure 3). In turn, these higher energy prices led to a major squeeze on real incomes. Unsurprisingly, widespread recessions followed. But work by Hamilton2 has shown a role for a run-up in oil prices in numerous recessions. Indeed, casually inspecting Figure 3, we can see that this is the case.

Figure 3: G7 energy price growth

Source: Thomson Reuters Datastream, Aberdeen Standard Investments

Monetary policy tightening has also played a role in many historical recessions. It is often related to the potentially inflationary effect of a sharp rise in oil prices. Under such circumstances the business cycle might follow a set pattern. First, economic expansions reduce the amount of spare capacity in the economy, eventually resulting in inflation pressures. These cause the central bank to raise interest rates. causes This in turn weighs on economic activity and often causes a recession.

Figure 4 suggests that there is a lot of truth in this basic example of monetary policy’s role in triggering recessions. We have used the ex-post US real policy rate (the fed funds rates minus core CPI inflation) as a a proxy for the global real policy rate. It rose at least 200 basis points in the years before most of the major recession clusters in the post-war OECD economies.

Figure 4: US real policy rate

Source: Thomson Reuters Datastream, Aberdeen Standard Investments

Finally, asset price bubbles, which then unwind, often due to monetary policy tightening, have been central to a number of the recessions in our sample. So too have other large-scale financial imbalances. Identifying financial imbalances and bubbles is no easy task --if it were, they might not develop in the first place!, One simple measure, however, is the share of private non-financial debt relative to its long-term trend. Figure 5 shows the significant build-up in US private sector debt ahead of the early-1990s recession clusters, as well as the global financial crisis. We can see that there were similar sharp run-ups in private sector debt ahead of the sovereign debt crisis.

Figure 5: US private sector non-financial debt, deviation from trend, % of GDP

Source: Thomson Reuters Datastream, Aberdeen Standard Investments

A large body of empirical work, most obviously that by Reinhart and Rogoff3,shows that recessions involving large-scale financial imbalances unwinding have a disproportionately larger and longer-lasting impact on economic activity. In Figure 2, we show the distribution of recessions by length and depth. All of the recessions in the top-right quadrant involve the unwinding of financial imbalances.

The lessons from pandemic-induced recessions

What of recessions closer in origin to the current one, with their cause in pandemic diseases?

Early on, observers drew comparisons between the Covid-19 pandemic and SARS. At first, that comparison was reassuring. The bounce-back in the most affected Asian economies was rapid. Meanwhile, the World Bank estimates that it knocked just 0.1% off the level of global GDP. But Covid-19 and its associated economic consequences are now vastly more serious than SARS. This means we should look back at the bigger global pandemics in history. The Hong Kong flu of 1968-9, the Asian flu of 1957-58, and the Spanish flu of 1918-19 caused overall deaths that are (and will hopefully remain) much greater than from Covid-19. Nevertheless, the scale of global economic disruption they wrought is probably more comparable.

Figure 6: Deaths during previous global pandemics (millions)


Source: Various, Aberdeen Standard Investments

The limitations of historical data make any assessment of economic impact rough-and-ready at best. There are also the difficulties of disentangling the effect of the pandemic from other things happening at the time. The most obvious of these is the case of Spanish flu in the aftermath of the First World War. But we can compare GDP growth in G7 countries in the year of the pandemic to GDP growth in the five years beforehand. This suggests an impact of between 2% and 4% off the level of economic activity (Figure 7). The profile of GDP growth around these pandemics follows an extended “U” shape. Growth dips sharply in the year of the outbreak, but recovers subsequently.

Figure 7: G7 GDP growth relative to five year average, around global pandemics


Source: Maddison database, Aberdeen Standard Investments

Moreover, these rough-and-ready estimates broadly chime with the World Bank’s 2006 estimates[1] of the damage that would be inflicted on the global economy if there were a new pandemic with the characteristics of the 1918/19, 1957/8 or 1968/9 episodes (Figure 8).

Figure 8: Effect of global pandemics on world GDP, %


Source: Maddison database, Aberdeen Standard Investments

Work by the Federal Reserve Bank of San Francisco has shown that major historical pandemics involving at least 100,000 deaths have long-lived labour market and interest rate implications. Real wages are elevated in the aftermath of pandemics, consistent with pandemics inducing labour scarcity or raising precautionary savings rates. This is in notable contrast to military conflicts that involve similar death tolls, where the concurrent destruction of capital (a feature generally not present in pandemics) lowers the marginal worker’s productivity and hence real wage.

Meanwhile, the natural rate of interest declines for decades after a pandemic, reaching its nadir about 20 years later, when the natural rate is some 150 basis points lower than had the pandemic not taken place. It is not until 40 years after the pandemic that the natural rate returns to the level it would have been absent the pandemic. The upshot is that real rates of return on a wide range of financial assets are substantially depressed for up to four decades after a pandemic



Major historical pandemics involving at least 100,000 deaths have long-lived labour market and interest rate implications.

This recession

Chapter 2

For all that we can learn from historical comparisons, there is no doubt that the global economy is currently experiencing a recession which is unprecedented in the post-war era in its scale, speed and breadth. Literally overnight vast swathes of the global economy have been turned off as social distancing measures mean huge areas of economic activity simply cannot operate. Moreover, this a severe financial shock.

A supply and demand sudden stop

The coronavirus shock originated as a major services sector disruption in China’s Wuhan province. Here, workers could not get to their jobs and various forms of economic activity became impossible. For Western economies, this was originally felt as a classic supply shock – i.e. a shock that reduced the contemporaneous productive capacity of the economy. The shutdown in China impaired long and complex supply chains, making it harder for firms who run extremely efficient inventory management techniques to continue to source components and operate their processes. This, in turn, caused a dramatic increase in suppliers’ delivery times across the globe.

The economic clock has stopped ticking, but the financial clock is still running, creating a mismatch.

However, as the virus spread around the world and more and more countries imposed suppression measures (‘lockdowns’), the shock became a more profound hit to the supply and demand side of the global economy. Social distancing measures make it impossible for the economy to continue to produce a huge range of goods and services. This includes those in recreation, tourism, travel, and large swathes of the retail sector. This is a clear supply shock, but what is interesting about this historically is the impact on the services sector. The manufacturing sector tends to be more cyclically sensitive, especially those sectors involved in the production of durable consumer goods; while the service sector tends to decline less sharply in downturns as less expenditure is discretionary. Nonetheless, it is the service sector that is most affected by the collapse in person-to-person contact because this kind of interaction is crucial to the very service being provided. So this time we are seeing a much sharper decline in services activity than in manufacturing (Figure 9).

Figure 9: Global manufacturing & services PMIs


Source: Haver, Aberdeen Standard Investments

The lost spending and income in these sectors will be experienced as an enormous demand shock for other sectors of the economy as the circular flow of income around the economy is impaired. Indeed, the scale of the collapse in new orders is easily comparable to the Great Financial Crisis in its depth –and even more rapid.

In a sense, the ‘economic clock’ has stopped ticking, as the normal flow of spending and income that circles through the economy has experienced a sudden stop. Clearly, vastly fewer hours are needed or can be worked in the economy at the moment. However, this shock to hours can manifest in unemployment, furlough, or widespread hours reductions across the workforce with limited cost to employment depending on the relative flexibility of labour markets. For example, the astronomical increase in initial jobless claims in the US compared to the much smaller increase in joblessness in Germany shows how similar shocks can have quite different labour market outcomes depending on institutional arrangements. The more the shock is felt in hours worked rather than in unemployment, then the greater the chance the initial demand shock does not continue to propagate through the economy and do lasting damage by breaking valuable and hard-to-replace employment matches.

At the same time as the economic clock has stopped, the ‘financial clock’ is still ticking, in the sense that financial obligations such as mortgage and other debt payments still become due. It is this fundamental mismatch between the economic and financial clocks that creates deep downside risks for the economy. In the short term, it can cause liquidity mismatches as households and businesses do not receive the income and cashflow they need to service their liabilities. In the medium term, this liquidity problem can quickly become a solvency problem forcing otherwise viable businesses, contracts and other economic matches into default.

At least imbalances were less severe on the eve of this recession

As our look back at historical recessions revealed, the existence of large scale financial imbalances can be a catalyst for household and firm balance sheet deleveraging during a recession, which tends to have very long-lived consequences for economic activity even once the initial shock has passed. One crumb of comfort in the current recession is that the state of financial imbalances is less severe than on the eve of financial crises of the past.

Admittedly, the absolute level of global private non-financial debt to GDP is at an all-time high of 93%, which does increase the vulnerability of the economy to income shocks. But global non-financial debt to GDP relative to its long-term trend does not look particularly stretched in the major economies (Figure 10). In any case, there may be are good reasons why debt ratios are sustainably higher than in the past. With equilibrium interest rates low, higher debt levels can be supported for a given level of income. Moreover, the liberalisation of domestic financial sectors and financial globalisation has led to a greater competitive supply of credit for a given level of demand for debt. And there has been a shift in preferences regarding issuing debt rather than equity.

Figure 10: Private non-financial debt, deviation from trend, % of GDP


Source: Thomson Reuters Datastream, Aberdeen Standard Investments

Nevertheless, there are certainly pockets of excess that were concerning ahead of this recession, and which we have analysed in detail previously 4 5. The US sovereign sector and corporate leveraged loan market; the Chinese non-financial corporate and household sectors; the sovereign sector in the Eurozone periphery; and the household sectors in smaller advanced economies like Australia, Canada and the Nordics – all look over-leveraged and vulnerable to unwind. Digging into just one of these pockets of vulnerability, public-sector debt relative to GDP was already highest in those Eurozone economies with the lowest trend growth rates and often unstable political dynamics. Many of these economies – not the least of which is Italy – have also been hit hardest by Covid-19 infections.

Meanwhile, the global banking system has entered this shock well capitalised, and in a much better state than in 2008. The improvement in core tier 1 capital ratios at systemically important global banks means they are better able to withstand a sudden deterioration in asset quality. The upshot is that while the financial system and the tightening in financial conditions can still be a transmission mechanism for this shock, a full-on banking crisis seems more of a risk than a baseline expectation.

But the multitude of shocks can interact in damaging ways

Nevertheless, the various shocks working their way through the global economy can feed off each other to deliver even more damage. The heightened risk of liquidity and solvency problems increases the riskiness of various financial forces, driving risk premia across a range of assets suddenly higher. The rapid downward adjustment in prices can impair financial market functioning and credit creation, making it even harder for firms to access bridging support and other normal financing. And emerging markets have scrambled to find scarce dollar financing to service hard currency loans in the context of huge capital flight. All of this increases the risk of bankruptcy and permanent capital destruction weighing down on long-run growth. Tighter financial conditions also weigh on demand through a host of other channels including through balance sheet strength and the ‘wealth effect’. It can also adversely affect animal spirits, as lower asset prices can not only reflect reduced confidence about the future but also generate these weaker expectations.

The expectation of lower productivity growth in the future due to an impaired supply-side reduces expected future incomes, which in turn reduces demand today. This lower demand can in turn encourage firms to invest less, which itself also damages long-run productivity growth. This can cause an endogenously generated adverse feedback loop pushing down demand and long-run growth, exacerbating and amplifying the initial shock.



The economic clock has stopped ticking, but the financial clock is still running, creating a mismatch.

A path to recovery

Chapter 3

The severity of the initial recession aside, what can we say about the path to recovery? The current chapter describes how the example of select East Asian economies where the virus is (seemingly!) under control, and the size of offsetting monetary and fiscal policy responses almost everywhere, hold out the hope of an unusually sharp but short global recession. Nevertheless, as the subsequent chapter discusses, the downside risks are manifold.

Select East Asian economies and the light at the end of the tunnel

The current recession is unusually deep, but also likely to be unusually short relative to that depth.

Those looking for light at the end of the tunnel should cast their gaze towards the likes of China, Japan, South Korea, Taiwan and Singapore. These select East Asian economies have shown that strong suppression tactics can bring the virus outbreak under control, while extensive ‘test-trace-isolate’ programmes can control fresh outbreaks. Ten weeks after the peak growth rate of the outbreak in China, daily growth of the infection is now actually negative, as recoveries outstrip new infections. Meanwhile, Japan and South Korea have lowered, if not eliminated, new infection growth (Figure 11). Of course, these economies are also having to re-impose certain suppression measures. There are also question marks over the quality of infection data, which both speak to the downside risks discussed in the next chapter.

Figure 11: Covid-19 confirmed cases, daily infection growth rate


Source: Thomson Reuters Datastream, Aberdeen Standard Investments

Nevertheless, the upshot may be that stringent multi-month lockdown can be avoided in developed market economies as well. Shutdowns in some parts of China started to be eased after three to four weeks, while the lockdown in Wuhan eased from 8 April – 11 weeks after it was imposed. Shutdowns in various European countries are very gradually starting to soften now too, which holds out some hope of a rebound in economic activity later in the year.

Indeed, (very) preliminary data from economies that have partially lifted suppression measures suggests that, while the upfront economic pain has been enormous, a stabilisation, if not yet an actual rebound, in economic activity is possible. For example, the proliferation of daily activity trackers for the Chinese economy is showing some signs of normalisation. Coal consumption is returning closer to something like seasonal norms. And, after plunging in February, the March PMIs for China have returned to around their level from January.

That said, it is important not to confuse a stabilisation in economic activity with a genuine rebound. PMIs only just over the 50 mark are indicative of the former but not the latter. Moreover, the recovery in PMIs could reflect a surge in sentiment and animal spirits from restrictions being lifted rather than a direct increase in activity levels. Meanwhile, Chinese daily activity trackers for proxies of service sector activity – public transport journeys or traffic delays – have shown a less convincing improvement than proxies for industrial activity such as coal consumption.

More broadly, the transition to regular activity after suppression measures are lifted may simply take a long time. Lockdowns are likely to be lifted only partially, perhaps with certain sections of the economy or particular demographic cohorts emerging back into the outside world, while others remain in isolation for longer. Restrictions on activities like foreign travel or large public gatherings are likely to remain in place well after others are lifted. And the large number of, and staggered emergence from, lockdowns internationally is bound to cause extended supply chain disruption everywhere.

Moreover, there are important and difficult questions about whether those Asian economies with a handle on the pandemic really represent good examples for the rest of the world. The ability of Europe and the US to sustain the stringent suppression measures of China, or enact the extensive ‘test, track and isolate’ measures of South Korea, is not clear. Moreover, it is not yet certain that Asian economies will actually maintain a lid on new infections as they relax lockdowns. Early indications hint at a renewed pick up in infections as suppression measures have been eased. The next chapter considers downside scenarios, in which the current recession is much more extended, in further detail.

The speed and scale of policy support

Neither monetary nor fiscal policy will be able to offset the economic disruption from lockdowns. Indeed, policymakers are actively trying to shut down those parts of the economy that rely on human contact. In that context, the language of ‘stimulus’ is not really appropriate to describe the significant policy support we have seen round the world. Instead, monetary and fiscal easing is about helping households and firms bridge the income and cashflow shock that comes from the economic clock stopping while still needing to service debt repayments and other liabilities as the financial clock still ticks.

On the fiscal side, it is worth distinguishing between genuine fiscal loosening and building up of contingent liabilities. In terms of genuine fiscal loosening, there have been one-off cash handouts in the US and Hong Kong; temporary wage subsidies in Australia, the UK and Germany; increasing sick pay benefits in the US and UK; lowering business taxes, tax and social security credits in Spain; and many more policies besides. Globally, the loosening in structural budget deficits adds up to around 3% of GDP.

The building up of contingent liabilities as a percentage of GDP has been larger still. For example, in the Eurozone, these sorts of measures add up to 10% of GDP, led by the credit guarantees for bank loans by the state bank KfW in Germany. In the UK, government loan guarantees are equivalent to 15% of GDP, with a promise that more is available if necessary. Should firms end up defaulting, these liabilities will show up as government debt (rather than deficits), otherwise they will increase debt overhangs for firms receiving support.

Admittedly, the global fiscal policy response in general, and the European policy response in particular, has lacked big-bang cross-border co-ordination. In the Eurozone, fiscal policy measures have largely occurred at the level of individual member states, and have therefore been smaller than they might otherwise have been had they been funded centrally. This is important because the relative inadequacy of Europe’s fiscal response may slow the pace of recovery.

Meanwhile, significant monetary easing has helped smooth the financing of these huge fiscal measures while also providing independent support to the economy in its own right. In the US, the Federal Reserve has cut interest rates to the effective lower bound, announced an unlimited QE programme, and a range of liquidity supporting measures including direct lending to corporates and extending its swap lines to more central banks to ease the global shortage of dollars. Similarly, the Bank of England cut interest rates to the lower bound, restarted a QE programme worth 9% of GDP (bigger than the package following the financial crisis or Brexit referendum), and introduced a number of liquidity measures. After a somewhat hesitant start, the ECB announced an additional EUR 750bn QE programme, purchasing both public and corporate debt over the rest of this year. On top of measures already announced, this adds up to EUR 115bn/month for the rest of this year, well in excess of the peak run-rate of purchases during the post-crisis easing cycle.

Figure 12: Developed Market central bank policy rates


Source: Thomson Reuters Datastream, Aberdeen Standard Investments

Elsewhere, we have seen a wide range of advanced-market central banks (including Australia, Sweden, Canada, and New Zealand) cut rates, start or restart asset purchases, and initiate yield curve control, as they attempt to cushion the effects of local outbreaks. The Bank of Japan also introduced a series of easing measures, doubling ETF and REIT purchases and raising corporate bond buying and commercial paper purchases to JPY 2trn.

These various measures should help to cushion the worst of the downturn, help stop the shock metastasising into long-run damage to the productive potential of the economy, and leave the policy stance extremely accommodative. As and when suppression measures are rolled back, these measures will potentially help the economy to grow rapidly.

An unusually sharp global recession, but short relative to its depth

We use as an example select East Asian economies that are a month or two more advanced in the pandemic and its containment, and the speed and scale of global policy measures introduced. While we think the current global recession will be unusually severe in its depth, it should be relatively short in its duration.

The data already hints in this direction. In Figure 13, comparing the plunge and rebound in the Chinese composite PMI with previous deep economic shocks highlights the unusual short, sharp profile of this contraction. China’s composite PMI spent all of one month - February - below the 50 mark that separates expansion from contraction. But the size of that short fall was enormous, with the PMI falling to a never-before-seen low of 27.5. By contrast, even at the depth of the 2008 financial crisis, Italy’s PMI “only” hit a low of 34.9, albeit the PMI spent nearly two years below the 50 mark. And Greece’s rolling debt crisis of 2009-14 “only” saw the PMI touch a low of 37.7, but it spent a massive four-and-a-half years below 50.

Figure 13: Composite PMIs around recessions


Source: Haver, Aberdeen Standard Investments

Admittedly, our actual GDP forecasts (which come with the obvious caveat that they are forecasts!) suggest a somewhat longer-lived contraction in economic activity and a longer haul back to pre-shock levels of economic output. For example, our G7 GDP forecast incorporates a contraction to GDP over two quarters. However, it is not until eight quarters after the initial shock that the level of GDP is back at its previous level, let alone its previous trend (Figure 14).

Figure 14: Level of G7 GDP from the onset of recession to the quarter in which GDP regains its previous level


Source: Thomson Reuters Datastream, Aberdeen Standard Investments

Figure 15 shows the unusual nature of this recession - if it plays out as we expect - relative to other post-1950 downturns. The current recession is unusually deep, but also likely to be unusually short relative to the size of the GDP contraction, compared with other post-1950 global recessions.

Figure 15: The relationship between length and depth of G7 recessions

Source: Thomson Reuters Datastream, Aberdeen Standard Investments



The current recession is unusually deep, but also likely to be unusually short relative to that depth.

Could things
get worse still?

Chapter 4

For all that our baseline forecasts point to an unprecedented decline in global economic activity, it is not hard to engineer scenarios where the global economy suffers an even worse downturn with a more protracted recession.

Long lasting/rolling containment measures; reinfection of re-opening economies

For example, virus containment may simply take longer than we have factored in. An important piece of epidemiological modelling from Imperial College6 in the UK suggests that strong suppression measures may have to remain in force 2/3rds of the time until a vaccine is widely available. The paper estimates this as 12-18 months away. That would mean that large sectors of the economy remain essentially switched off for a much longer time than we have incorporated into the baseline. Alternatively, a country may suppress the virus internally but, as it opens up again to the rest of the world, could suffer episodes of re-infection due to exposure to regions where suppression has been less successful. Indeed, it is hard to see how any single country can unilaterally open up to the rest of the world while the virus is still uncontained in certain regions. We may already be seeing these dynamics in places like Singapore and Hong Kong.

If households and firms permanently increase savings, it will put downward pressure on the natural rate of interest that cannot be matched by monetary policy easing.

Self-reinforcing default and labour-shedding cycles

Another source of longer-lasting damage would be that the initial shock is so large that it sets off self-reinforcing credit default and labour-shedding cycles. Default and labour-shedding, once they begin, tend to have a lot of auto-correlation. That is, once they start, they don’t stop for a while. For example, when the US unemployment rate has risen more than 1%, it has gone on to rise at least 2.4% (Figure 16). Already, evidence from the incredible spike in initial jobless claims in the US over the last few weeks (Figure 17) and in Universal Benefit claimants in the UK suggests we are seeing an unprecedented increase in unemployment. Putting these valuable employment matches back together will probably be a long and painful process.

Figure 16: US unemployment cycles


Source: Thomson Reuters Datastream, Aberdeen Standard Investments

Figure 17: US initial jobless claims


Source: Oxford Economics, Aberdeen Standard Investments

Additionally, if households and firms permanently increase their savings rate in response to this shock, it will put downward pressure on the natural rate of interest (r*) that cannot be matched by monetary policy easing because of the effective lower bound. Work by the San Francisco Fed7 suggests exactly this outcome following past pandemics: a rise in precautionary savings means that the natural rate of interest declines substantially after a pandemic. It reaches its nadir about 20 years later, with the natural rate 1.5% lower than had the pandemic not taken place. In the long term, that means that economic activity would be lower, as policymakers are unable to adequately stimulate demand.

Stall speed and escape velocity

Finally, our recent research paper ‘Stall Speed and Escape Velocity: Empty Metaphors or Empirical Realities?’ (CEPR, 2019)8 showed that self-reinforcing ‘stall speed’ dynamics are a genuine feature of the business cycle. At low growth rates, economies are simply more vulnerable to negative shocks that can tip them into recession (Figure 26). This is especially pertinent at the current juncture. Even a temporary shock hitting against the backdrop of chronically low growth means heightened risk of economies stalling into a more sustained recession. The lack of conventional monetary policy space confounds this problem, as central banks increasingly struggle to offset the shock. Given the extremely large output gap that this crisis will open up, it is hard to see how disinflationary forces will dominate the inflation outlook. As inflation falls, and central banks are unable to lower the structure of rates much further, real interest rates mechanically increase. This weighs down the recovery even further.

Of course, central banks are not necessarily limited to tools that alter the path of short and long-term rates (such as forward guidance and QE), whose efficacy is curtailed when rates are already low. For example, the central bank could permanently increase money supply to finance easier fiscal policy, which we defined as true helicopter money in a recent note9. It is hard to see how this could fail to stimulate demand and raise the price level. So long as it was conducted within the context of a credible policy framework – ideally a level targeting regime that anchors expectations of the evolution of the price level over the long run – there is no reason this policy would lead to fiscal dominance and a loss of inflation control to the upside. However, the innate conservatism of central bankers and concerns about political encroachment on their operational independence make this policy response unlikely.

Meanwhile, our research found almost no evidence of escape-velocity dynamics in the recovery from recessions. In other words, economies can get into a self-reinforcing slowdown, but do not seem to benefit from self-reinforcing recoveries as they emerge from recessions. Granted, the recovery from this recession will be unlike many in the past. The gradual turning back on of parts of the economy should deliver a continuous impetuous to growth. However, this provides little comfort to those looking for a V-shaped profile with a very rapid recovery. We are unlikely to experience anything like escape velocity when the economy does recover.

Indeed more broadly, the nature of this recession and (eventual) recovery highlights the difficulty of understanding patterns in business-cycle dynamics. With global growth slowing in 2019, before falling into recession this year, a naïve parsing of the data may make this episode appear to have the characteristic form of a ‘stall’. That is slowing growth begetting further slower growth, before dipping into a recession. Of course, this is not really the dynamic that has played out this time. Rather, the economy has been hit by an unforecastable shock from the pandemic and subsequent lockdowns. So empirically, this may ‘look’ like an incidence of stalling but it is really a different type of recession. This demonstrates the need to understand the specific shocks that are driving each idiosyncratic business cycle. Nonetheless, it is clear that low growth and low policy rates across the world meant the global economy was more vulnerable to any particular shock. This increased vulnerability that comes from lower trend growth will be a recurring feature of the global economy in the future.



If households and firms permanently increase savings, it will put downward pressure on the natural rate of interest that cannot be matched by monetary policy easing.

The long-term

Chapter 5

What of the long-term consequences of this recession? All significant and disruptive recessions have long-lived consequences, and this one is likely to be no different.

We see the current crisis as deepening the central geostrategic cleavage of the modern world, between the US and China.

The European Sovereign Debt Crisis forced drastic innovation in the tools of the European Central Bank. The Great Recession birthed secular stagnation, but also the recent uprising of political populism. Going further back, the Volker Shock brought developed market inflation under control. And the Great Depression ultimately expanded the role and intervention of the state in the economy via the New Deal but also the Smoot-Hawley tariffs. The current downturn is likely to be no different in having a range of long-term repercussions.

But the consequences are likely to be unique, reflecting the uniqueness of the shock. This chapter is a brief first pass at sketching out some of these consequences. It is unavoidably speculative at this stage. We will continue to do more work over time, delving into these developments.

An altered balance between monetary and fiscal policy

The extraordinary lengths to which fiscal policy has been deployed to fight this crisis, and the implicit absorption of much of the associated debt-issuance by central bank balance sheets, may mark a turning point in the relative roles of monetary and fiscal policy.

In extremis, this crisis may be the end of the road for monetary policy as the primary tool of macroeconomic management. Monetary policy could now play a supportive and second-fiddle role to activist fiscal policy. The central banks’ role could be in financing fiscal aims, rather than acting independently to set the price level.

But we need not extrapolate quite so heavily to see an altered balance between monetary and fiscal policy emerging from this crisis. For example, it could be that monetary policy remains an independent bureaucratic function that targets inflation via short-term policy rates in ‘normal’ periods. But, when in crisis-fighting periods with limited room to lower real rates but plenty of scope for quantitative easing and even helicopter money, it kicks into gear as the funding mechanism for very activist fiscal policy. This isn’t full fiscal dominance as such (which technically means fiscal, not monetary, policy setting the price level). However, it would represent an altered balance between monetary and fiscal policymakers in fighting crises and providing macroeconomic stimulus, while remaining within the current framework of independent central banks.

To some extent, these were trends already underway in a number of economies. The enormous role monetary policymakers had played in the Great Financial Crisis meant they were already under increasing scrutiny, and even threats to independence, before the current shock. Meanwhile, central bankers themselves had been rethinking the appropriate role and target of monetary policy through various strategy reviews. In addition, they had been lobbying fiscal policymakers to take a more proactive role in counter-cyclical macroeconomic management. Macroprudential measures are also likely to become more important to central banks as they try to ensure greater resilience to future shocks. The turbulence of the current recession may just be accelerating trends that were already underway.

Deepening the US-China cleavage

We see the current crisis as deepening the central geostrategic cleavage of the modern world, between the US and China. For example, the US administration has already labelled Covid-19 “the Chinese virus”, originating in China’s wet markets and adventurous eating habits. On the other hand, we could interpret the Chinese response to the virus as testament to its centralised political system and huge state capacity. It instigated large-scale and apparently highly effective lockdowns and other public health measures in short order. China has taken the opportunity to increase its soft power during this crisis too, supplying medical equipment and expert advice to some European and African countries.

Meanwhile, the US public health response has been slow and uncoordinated. It has exposed the limitations of a large federal democracy in dealing centrally with an emergency of this scale. But should treatments for the disease and a vaccine ultimately arise from the deep wellspring of US pharmaceutical research, the crisis may ultimately be testament to Western innovation.

Altering the course of globalisation

In the short-term, the current recession clearly involves a collapse in global trade volumes and the movement of people and capital across borders. This is both directly, as broad economic activity contracts, and indirectly, as restrictions on the movement of goods and people have clicked into place. In a sense, this is directly unwinding globalisation.

Even once the most severe limits on movement are lifted and economic activity recovers, this recession will likely have long-term consequences for globalisation. For example, firms may shorten supply chains more permanently, in order to build resilience to any future large-scale disruptions. Having had their fragility laid bare, just-in-time delivery models are likely to become less pervasive. People’s desire to holiday or study abroad may take a hit. And trade barriers that protect national industries and build economic self-reliance are only likely to continue building.

On the other hand, the crisis could give extra impetus to the globalisation of information – that is the free-flow of data across borders. Remote working will likely increase and countries may pool resources to find medical solutions to the Covid-19 virus.

Again, these are trends that were, to some extent, already underway. Indeed, trends that our previous work building a globalisation index and charting the changing nature of globalisation had already identified10. But they are set to get additional impetus from the impact of this recession.

Figure 18: ASIRI Globalisation Index

Source: Aberdeen Standard Investments

Increasing the size of the state

In many countries, the state is likely to emerge from this crisis bigger than it entered. Public spending will be much higher, as deficits balloon and the size of the economy itself shrinks, and taxes may have to rise. More money will be spent on healthcare and the social security net. The state will have temporarily paid the wages of vast swathes of the workforce, under short-time work schemes and other such programmes. Large parts of industry may have big equity stakes from the state. Or they may at least have benefited from subsidised lending schemes underpinned by the state. And governments are likely to demand some quid pro quo for their support offered to firms through additional oversight and regulation. Moreover, we expect the expansion of the state to be ‘sticky’. That is, politicians may not wish to give up the increase in powers and reach gained during this recession.

This is not an acceleration of a trend already underway. Rather, something of a reversal of the atrophying of the role of the state in several major Western economies over the past few decades. What we might broadly call ‘neoliberalism’ has seen the state retreat from various aspects of individuals’ lives. From welfare systems to economic infrastructure provision, to product and labour market regulation. Outriders of this reversal were already present in the political discourse of many Western economies in recent years, of course. The US presidential campaigns of Bernie Sanders, the industrial plan of erstwhile Conservative This crisis may be an accelerant for the rise of the surveillance state in particular. One way in which Asian economies are instigating successful ‘test-trace-isolate’ regimens is via location tracking of individuals. Even Israel has passed an emergency law to use mobile phone data to track people infected with Covid-19. Widespread state tracking of citizens’ movements, undertaken initially under the auspices of infection control, may become permanent features of policing and law enforcement.

A boost or a burden for productivity growth?

As the structure of the economy evolves in the post-crisis period, it is likely that this will have a significant impact on productivity growth. Firms may practice less efficient inventory management and sourcing processes as they seek to reduce their vulnerability to future shocks that temporarily disrupt cross-border supply changes. This is likely to weigh on productivity growth at the margin as more resources are devoted to inventory management and a preference for short supply chains means comparative advantage becomes a less important driver of the location of production. Furthermore, those sectors which are likely to emerge larger following this crisis, including the public sector and healthcare, tend to have lower productivity growth. Meanwhile, agglomeration effects, where locating economic activity in close proximity to other similar activity tends to have positive spill overs on productivity, may become less powerful if there is a movement away from high-density cities and towards greater working from home. Finally, generous funding schemes by governments might make it easier for so-called zombie firms to stay in business, thereby blunting the process of creative destruction. This could slow the reallocation of capital in the economy and reduce potential growth.

Against this, it is plausible that firms use this period of labour shedding to invest in capital, which is a pattern we have seen in previous downturns and might be accelerated this time because capital does not face the same health risks as a human labour force. As such, the economy may become slightly less labour intensive and slightly more capital intensive at the margin (a development which the death of workers themselves only plays into, unfortunately), which would typically push up productivity growth, all else being equal.

Online everything

Enforced social distancing has increased the use of online tools – whether it is working remotely, ordering groceries online, or streaming video and music content. These are not new developments of course. But certain forms of digital penetration are still surprisingly low in some economies and among some population cohorts. For example, only 5% of Italians or Greeks “sometimes” or “usually” work from home (Figure 19). And less than 10% of the UK’s grocery shopping by value is usually done online and then delivered by van. Expect the on-lining of our everyday lives to be permanently altered by the current recession and associated suppression measures.

Figure 19: Share of the population working from home, 2018


Source: Eurostat, Aberdeen Standard Investments

Hastening or impeding the low carbon transition?

The current recession almost certainly marks a local maximum in carbon emissions and pollutants that lower air quality. For example, data from European Space Agency satellites shows the reduction in nitrogen dioxide emissions over European industrial clusters relative to the same month last year. The widespread shuttering of economic activity cannot but have lowered anthropogenic emissions.

Whether this marks a more permanent turning point in the energy transition is less obvious. For example, the postponement of COP26 is probably not good news for international efforts to fight climate change. And to the extent that economic stimulus measures sacrifice environmental protections – as is already occurring in the US where vehicle emissions standards have been rolled back – this is not helpful for climate change mitigation efforts either. Meanwhile, the collapse in the oil price reinforces demand for fossil-fuel-led development, as it is now more cost effective in an already difficult economic environment.

But for governments, there is an opportunity to make the recovery environmentally sustainable through investment in clean energy infrastructure and technology solutions, as well as harnessing new stakes in companies to influence corporate behaviour. More fundamentally, the widespread shuttering of economic activity to achieve a policy aim – in this case contain the spread of Covid-19 – has surely pushed the Overton Window vis-à-vis the economic sacrifices polities are willing to absorb tackling the threat of climate change. There is certainly a difference between sacrifices to avert a short-term disaster like a pandemic that politicians of all stripes see as being in their interests, and the slower burning crisis of climate change that is still heavily politically contested in certain countries. But it has at least entered the realms of ‘thinkability’ that we might accept large upfront economic pain to achieve bigger policy and humanitarian aims. Either way, expect climate activists to take this argument up once the current recession ends.



We see the current crisis as deepening the central geostrategic cleavage of the modern world, between the US and China.

Investment implications

Chapter 6

These structural changes to the economic and political environment will also have important implications for drivers of the return of various assets. Broadly speaking, we would expect returns to be lower on average commensurate with a lower growth economy, but pockets of opportunity will open up as capital needs to be reallocated and put to work in the newly structured economy.

The risks of both a deflationary trap and a regime of higher inflation have both increased in probability.

Natural rate of interest

The equilibrium, or natural rate, of interest is likely to fall even lower, continuing the secular trend characterising the interest rate environment for decades. This is because nominal growth is likely to be much lower due to the combination of lower potential growth and the fact that inflation is likely to be extremely low for a long period as the economy clears the huge output gaps that this crisis opens up. Moreover, investors’ perception of risk in the economy will probably increase, which simultaneously pushes up the risk premium for risk assets and makes risk-free assets like government bonds more attractive and so pushes down on the natural rate of interest. Finally, if there is a permanent increase in precautionary savings as households and firms look to build stronger liquidity buffers to protect them in future crises, this will also push down on the natural rate.

Inflation expectations and inflation risk premia

While the most likely outcome of the crisis is a sustained period of low inflation, the opposing tail risks of a deflationary trap and of much higher inflation have probably both increased in probability. The deflationary trap would develop if policymakers were insufficiently robust and creative in the steps they take to smooth this crisis and then create the conditions for a rapid recovery in demand. The risk of higher inflation comes from the prospect of fiscal dominance, when monetary policy becomes subordinate to the need to finance large government deficits, which eventually generates higher and higher inflation as interest rates are held systematically below the natural rate. Given the increased risk of both these scenarios, it is natural to expect that the inflation risk component of bond prices and asset prices more generally will increase.

Equity returns

In equilibrium, the higher risk premium demanded by investors to hold those assets will push up expected returns. Against that, lower nominal growth will reduce the expected future cashflow of many companies, while changes in the management practice of some firms, perhaps due to greater governments ownership, will tend to make shareholder return a less pressing priority – both of which would push down expected returns. Overall, the return on equities is likely to be somewhat lower than pre-crisis expectations.

However, from where we stand right now, the steep correction in equity market prices means that expected returns are boosted by attractive valuations. This is especially true if markets tend to overshoot when responding to incoming information in part due to behavioural biases and in part to provide the steep discounts required to incentivise investors to once again hold riskier assets.

More specifically, and perhaps perversely given the impaired growth environment stemming from this crisis, stocks that offer genuine growth may become even more attractive versus their ‘value’ counterparts. They also benefit from a duration effect, in that growth stocks tend to have earnings profiles that are discounted over a much longer term horizon, the value of which mechanically increase as discount rates fall. Moreover, the disappearance of dividends payments undermines much of the classic case for value stocks.



The risks of both a deflationary trap and a regime of higher inflation have both increased in probability.

The views and conclusions expressed in this communication are for general interest only and should not be taken as investment advice or as an invitation to purchase or sell any specific security.

Any data contained herein which is attributed to a third party ("Third Party Data") is the property of (a) third party supplier(s) (the "Owner") and is licensed for use by Standard Life Aberdeen**. Third Party Data may not be copied or distributed. Third Party Data is provided "as is" and is not warranted to be accurate, complete or timely.

To the extent permitted by applicable law, none of the Owner, Standard Life Aberdeen** or any other third party (including any third party involved in providing and/or compiling Third Party Data) shall have any liability for Third Party Data or for any use made of Third Party Data. Past performance is no guarantee of future results. Neither the Owner nor any other third party sponsors, endorses or promotes the fund or product to which Third Party Data relates.

**Standard Life Aberdeen means the relevant member of Standard Life Aberdeen group, being Standard Life Aberdeen plc together with its subsidiaries, subsidiary undertakings and associated companies (whether direct or indirect) from time to time.

Risk warning

Risk Warning

The value of investments, and the income from them, can go down as well as up and you may get back less than the amount invested.