Call it the financial meltdown forecaster. The team of economists who last year demonstrated that a small number of companies wield disproportional power over the global economy has now produced a simple visual tool that can monitor financial stability in real-time.
Like a weather forecast, DebtRank could monitor global financial activity for telltale signs of impending disaster. Its designers say it could anticipate and so help prevent global economic crises like the events of 2008, from which the world is still reeling.
Importantly, it’s not just size that matters when it comes to how much risk a troubled bank poses to the financial system. Even a relatively small loss at a firm intimately connected to many other banks could cause a ripple effect that threatens the whole system.
To test DebtRank, Stefano Battiston of the Swiss Federal Institute in Zurich used previously confidential data on the 2008 crisis gathered by the US Federal Reserve. The analysis identifies a number of firms that were particularly precarious: had just one of those defaulted on its obligations, it could have resulted in the depletion of more than 70 per cent of the wealth in the entire network. During the crisis, coordinated – and very expensive – intervention prevented such an outcome.
Regulators like Andy Haldane at the Bank of England have called for tools that would provide regulators with a clearer understanding of the state of the global financial system in real time.
DebtRank is an answer to such calls. Battiston says that armed with tools like this, regulators could theoretically spot potential problems as they occur and intervene before the entire system is at risk of collapses. They could also model the impacts of interventions ahead of time, allowing them to experiment until an optimal solution is decided upon.
The previously confidential data included detailed daily balance sheets for 407 institutions that between them received bailout funds worth $1.2 trillion from the Federal Reserve. The data covers 1000 days from before, during and after the peak of the crisis, from August 2007 to June 2010.
Battiston and his colleagues focused on the 22 banks that collectively received three-quarters of that bailout money, and which turned out to be intimately linked to one another in a closely-knit network. Almost all were members of the “super-entity” found at the heart of the global economy by last year’s study.
From the data, Battiston and his colleagues calculated a daily DebtRank value, ranging from 0 to 1, for each bank. The value reflects the likelihood that a bank will default, and how much this would damage the creditworthiness of other banks and companies in the network.
Spiral to collapse
A DebtRank value of 0 means that if a bank defaults, it poses no risk whatsoever to other members of the network. Theoretically, if a bank with a value of 1 were to fail, it would obliterate the economic value of the entire network.
Battiston represents DebtRank values pictorially within a spiral. Companies ranked 0 sit on the outside of the spiral, and a company ranked 1 would occupy dead centre. As a company’s DebtRank value rises, it moves steadily towards the centre of the spiral, posing ever greater threats to the entire system as it does so.
Representations of the 22 companies before and at the peak of the crisis show visually how valuable the tool could be to regulators. In August 2007, well before the crisis, all 22 banks sit around the edge of the spiral (see image 1), with DebtRank values averaging just 0.08. By the crisis, the average had risen to 0.52, sending banks plunging towards the centre (see image 2).
At that point, several of the individual institutions could have wiped out more than 70 per cent of the total value of the network on their own, had they failed.
The model also demonstrated how shocks to the system can be rapidly propagated between the closely connected banks. If all 22 simultaneously suffered a 10 per cent loss in the value of their assets, as was seen in the sub-prime mortgage crisis, the losses to the entire network could have caused a the system to crash.
Battiston says that his team is collaborating with the European Central Bank and a selection of national banks in Europe to experiment with DebtRank and other systems designed to avoid financial meltdown. To create DebtRank registers, regulators would need to receive daily or weekly updates of all transactions, debt profiles, detailed balance sheets and other assets.
The problem, he says, is that most global financial transactions are confidential, and any regulation that does occur is only at national level. “The police are local, but the transactions international,” he says. “There’s no authority keeping track at international level,” he says.
If such a system was installed globally, the data the regulator receives could be kept confidential, and highly-connected banks at risk of defaulting could be spotted early and warned privately to fix the situation, to solve the problem without panicking the markets.
The study is a good proof of principle and its methodology looks promising, says Simone Giansante at the University of Bath in the UK. Though he cautions that Battiston’s study has only looked at two types of transactions, and says a more comprehensive tool would be better – assuming the researchers could obtain the relevant data.
Journal reference: Nature Scientific Reports, DOI: 10.1038/srep0054
Additional reporting by Michael Marshall
- Sorry, there are no polls available at the moment.