Fear is stalking the financial world once more. The thin shoots of optimism that poked up as stock markets beat their pre-crash peaks are withering amid concern about everything from the United States's quantitative easing to the financial health of China.
Global markets have been on the slide since the last week of May, leaving investors barely ahead of where they were in January. Yet why now - and why so much ? After all, concern about these issues has been swilling around for months.
But get this: just before the slide began, a team of researchers in Switzerland posted the following cryptic note: "Analysis by the FCO indicates that the US stock market index S&P 500 is growing at an unsustainable rate (8.3 per cent gain in four weeks) and will soon correct".
And so it did. Just a matter of hours later, the index and financial markets in general began their descent.
So who are these financial soothsayers? They are a team of mathematicians and physicists led by Prof Didier Sornette of the ETH Zurich university in Switzerland, who together staff the Financial Crisis Observatory. They aim to do for financial markets what meteorologists do for the weather.
They claim to be able to detect the tell-tale signs of market "bubbles" in financial data, and even to predict when they are likely to pop.
The idea of predicting market upheavals by such means is hardly new. Back in the 1930s, a Kansas-born accountant named Ralph Elliott claimed to have found a way of predicting the market through waves of price movements caused by the ebb and flow of optimism among investors.
In October 1987, the American market analyst, Robert Prechter, made headlines by using these "Elliott Wave" methods to warn clients to prepare for a huge collapse in share prices - which duly manifested itself in the notorious Black Monday crash that year.
The trouble with such stories is that they are just that: stories. It is actually easy to predict anything - just make the same prediction repeatedly, and ignore all the times it turns out to be wrong. A stopped clock is still correct twice a day.
And in economics, there has never been a shortage of Nostradamuses forecasting meltdowns. As the Nobel Prize-winning economist, Paul Samuelson, joked back in the 1960s, "Wall Street indexes predicted nine out of the last five recessions".
But Prof Sornette and his team are determined to change the flaky image of financial forecasting. Their May 21 forecast was not some one-off lucky guess, but the latest in a slew of impressively accurate warnings over the past 15 years.
They also have serious credentials: Prof Sornette is renowned for his work in fields as varied as geophysics, materials science and brain research.
Most important of all, their techniques have a solid foundation. Rather than relying on patterns dredged from data - a notorious source of unreliable "laws" - the FCO team uses techniques already proven in other fields plagued by the risk of sudden failure.
At their core is the concept of "self-organisation", the way in which anything from a heap of sand to the Earth's crust will steadily become more unstable until it suddenly re-organises itself into a more stable state again.
Since the late 1980s, Prof Sornette has been studying this phenomenon, developing diagnostics of systems becoming unstable - and ways of predicting when they will re-organise themselves.
It turns out that the instability builds through a series of ever-faster waves, which manifests as so-called "super-exponential" growth.
The idea of exponential growth is familiar enough. A population that grows exponentially at a constant 50 per cent a year shows its characteristics. After one year the population is 1.5 times its original size, 2.3 times after two years and so on, ballooning to more than 57 times its original size after just a decade.
Impressive - but nowhere near as impressive as super-exponential growth, where the rate of growth itself grows over time: 50 per cent the first year, say, then 60 per cent after year two and so on. After a decade of that, the same population would explode to more than 700 times its original size.
Prof Sornette and his colleagues have found ways of detecting such super-exponential growth in financial markets, and of predicting when the resulting bubble is likely to burst. These are now the key "instruments" used by the FCO.
The team's claims have generated considerable scepticism. At his TED talk this month, Prof Sornette recounted how he warned hedge fund managers in 2007 of a major correction in the Chinese stock market by the end of the year.
His prediction was dismissed on the ground that the Chinese government would never permit such an embarrassing event in the run-up to the Beijing Olympics. Three weeks later, the markets began a collapse, ending in a 70 per cent loss by the year-end.
Such is the scepticism about market forecasts that when Prof Sornette warned of another, smaller correction in 2009, he was again dismissed - on the ground that the Chinese government must have learnt its lesson. That prediction, too, proved correct, but this time Prof Sornette found himself blamed for having triggered the collapse by publishing his forecast.
He and his colleagues have responded by setting up an international repository for their predictions, which are date-stamped and kept tamper-proof and secret until after the event. Over time, the evidence that they are on to something will become overwhelming - if, indeed, they are.
But this is just the start. Prof Sornette believes the theory behind the FCO's predictions raises the prospect of not just predicting bubbles, but also preventing them.
Let us hope he is right, and that the world economy doesn't blow up before he gets the chance to save it.
Robert Matthews is visiting reader in science at Aston University, Birmingham, England
