The world has seldom been worse-equipped to fight a recession. Yet it has never had fewer recessions to fight. That makes the next global downturn difficult to imagine but it will most likely be a traumatic and unlooked-for event, more like the sudden outbreak of a new disease than the annual onset of flu.
Recession alarms sounded loudly this summer thanks to an inversion of the US yield curve, which means rates on shorter-term debt are higher than those on longer-term ones. That came as the current economic expansion broke records and we experienced a global slide in measures of manufacturing sentiment.
With interest rates trapped at zero or close to it across much of the industrialised world, central banks have little room to respond. Any slump in economic activity is a frightening prospect.
Possibly, this bout of global weakness will be allowed to spiral into a US and global downturn. Yet that would defy the pattern of recessions in recent decades, which has changed both in cause and frequency. They are rarer and, when they do begin, the spark is more often a financial meltdown than sliding business activity or an economic overheat.
Recession predictors have always had a terrible record but in the past, stopped-clock doom-mongers only had to wait a few years for their moment of glory. Now redemption can take decades. At 122 months, the current US expansion is the longest on record, just ahead of the period from 1991 to 2001.
Former US Federal Reserve chair Janet Yellen argues it is a "myth" that economic expansions die of old age. Longer expansions mean recessions have become rarer. In the 35 years after the second world war, there were eight US recessions; in the 35 years following that, there were four. Trends in other developed countries are similar.
In the 1990s and 2000s, economists used to talk about a "great moderation", the idea that output had become significantly less volatile since the early 1980s.
But that proved embarrassing after the 2008 financial crisis and the phrase went away. The evidence since then, however, suggests the phenomenon is real. The volatility of quarterly growth in the US is now lower than it has ever been.
Several explanations have been put forward. One is the decline of inventory cycles in manufacturing: the use of just-in-time techniques has reduced the need to adjust stockpiles.
Another is the economy's structural shift towards services. Demand for spinning classes, management consultancy, karaoke and plastic surgery may fluctuate with the overall economy, but unlike steel mills and automobile factories, they have no great cyclical dynamics of their own.
A third and probably more important explanation is the success of central banks in stabilising inflation and averting boom and bust cycles.
Most US recessions in the postwar years began when the economy got too hot, inflation picked up and the Fed increased interest rates to choke off demand.
It is now decades since that last occurred. More recently, central banks have struggled with a lack of inflation, not too much.
When policymakers do not fear inflation, they can provide stimulus to offset a slowdown.
Indeed the risk is rather that they lack the ammunition to counter real shocks such as Brexit or US President Donald Trump's tariffs on China. But they are certainly trying. The Fed is cutting interest rates, mirrored by India, Thailand, the Philippines and New Zealand among others in August alone. Fiscal stimulus is on the way in South Korea and the UK. China is set to do more.
This is where many recession predictions go wrong. When the threat is obvious, the policy response arrives in advance. A bet that the current global slowdown will turn into global recession is in part a bet that these current stimulus efforts will fail.
It is easy to follow this argument into a fallacy: policymakers will act to prevent a recession, so no serious recession is possible. In the early 2000s there was plenty of hubris along these lines — but 2008-09 was a brutal corrective.
That recession began in the financial sector, with subprime loans and housing, just as the recession of 2001 began with the bursting of the dotcom bubble.
The business cycle may have damped but the financial cycle of Hyman Minsky is still with us. Finance is the most probable source of the next downturn, when it comes.
If real economic fluctuations are hard to predict, timing a financial crisis is next to impossible. Analysts may spot imbalances in markets but they can only guess at when and how they will unwind. There are some signs we are late in the current cycle. Asset valuations are high by historic standards. Private debt has risen a lot in China and some peripheral economies.
The Trump administration is rolling back financial regulation. It all increases risk. But the signs of stress that often precede a crisis — wild ebullience or rising defaults — are not obvious.
None of this means a recession will not occur — unpredictability is the whole point — but based on the pattern of recent downturns, there is little sign that one is already under way.
What is certain is that an economic slump under present conditions would be exceptionally hard to fight. That is likely to be the pattern for future recessions: rare and terrifying.
Written by: Robin Harding
© Financial Times