In 2008, in the midst of what was a financial crisis (and would go on to be known as the Great Recession), the US passes the Emergency Economic Stabilization Act — or as most people called it, “the bank bailout.” Essentially, the government created a fund to purchase $700 billion of toxic assets from banks, injecting capital into the limping financial system and saving many of America’s largest finance companies from failing.
In hindsight, a move like this was probably necessary and averted what could have become an absolute economic catastrophe. But the plan was criticized by economists for being unfair and not as efficient as it should have been. For instance, an investigation has found that the main winners were the large, unsecured creditors of large financial institutions. No doubt, a part of the bailout money was a good investment and paid dividends for the general population — and studies have also pointed out that even an imperfect but timely plan is better than a good, but untimely plan. But probably, not all of it worked for the people.
A new tool could help with that. The AI tool, described in a new paper in Nature Communications, assesses how much money should be invested into a bank during a financial crisis, and which banks should be bailed out — for the people, not for the bank itself.
Banks like AIs, but they might not like this one
As we nervously look at the economic and financial situation around us, bank bailouts could become a hot topic soon enough. With a global recession knocking at our door, a major war still raging in Europe, and a pandemic we seem to just not care about anymore, the economy is walking on broken glass. But should another bailout be on the menu, we may have better tools to prepare.
Researchers from University College London (UCL) have developed a mathematical framework for comparing different bailout strategies in terms of predicted losses to taxpayers. The model looks at the effects of a potential bailout on the bank itself, on other banks in the system, as well as taxpayers’ stakes in the banks.
These are all complex matters. The idea of a bank bailout is to provide the bank with some equity and prevent it from defaulting. This obviously takes money away from taxpayers, but this could be justified if the move prevents further losses. Simply put, not bailing out banks could lead to a domino effect that costs the taxpayers more than the bailout.
“Bank bailouts are controversial governmental decisions, putting taxpayers’ money at risk to avoid a domino effect through the network of claims between financial institutions. Yet very few studies address quantitatively the convenience of government investments in failing banks from the taxpayers’ standpoint.”
The Artificial Intelligence (AI) algorithm was tested by the authors using data from the European Banking Authority on a network of 35 European financial institutions judged to be the most important to the global financial system. The model can be used freely, and can be calibrated to work with other systems if it is fed more data (that is not available freely).
Neofytos Rodosthenous (UCL Mathematics), the corresponding author of the paper, believes the model can be used in practical situations.
“Government bank bailouts are complex decisions that have financial, social and political implications. We believe the AI approach we have developed can be an important tool for governments, helping officials assess specifically financial implications – this means checking if a bailout is in the best interest of taxpayers, or whether it would be better value for money to let the bank fail. Our techniques are freely available for banking authorities to use as tools in their decision-making process.”
The algorithm compares no intervention to different levels of investment in one or several banks at different times. In the case study with data from the European Banking Authority, a government bailout was found to be optimal only if the taxpayers’ stakes in the banks were higher than the threshold value that the model estimated. If this threshold value was changed, the optimal course of action was also changed dramatically.
Of course, it’s not like economists will just trust the tool blindly — but it can be a useful tool in their arsenal to assess the best course of action during a crisis. When a crisis hits, policymakers have to act fast, and sooner or later, the odds are a financial crisis will happen again.
The model can also work with past data and calculate what would have been a better course of action, says co-author Vito Latora from Queen Mary University of London.
“Governments and banking authorities can also use our approach to retrospectively review past crises and gain valuable learnings to inform future actions. One could, for example, review the UK government bailout of the Royal Bank of Scotland (RBS) during the financial crisis of 2007-9 and reflect on how this could potentially be improved (from a financial standpoint) in the future in order to primarily benefit taxpayers.”
This approach of looking at past data can also help policymakers learn how to devise more resilient interventions — and we’ve already seen how important this can be.
Without the 2008 crisis for reference, we likely wouldn’t have been able to weather the economic crisis induced by the pandemic so well, says lead author Daniele Petrone. But to better prepare for future crises, we need robust tools and a firm understanding of the long-term effects of these actions as well.
“Banks have so far weathered the current economic storm triggered by the Covid-19 pandemic. Their resilience has been bolstered by regulatory measures introduced following the global financial crisis of 2007-9 and by accommodating central banks’ monetary policies that have avoided bankruptcies across industries. However, no one can predict the effect on the financial system as central banks reverse previous policies, such as increasing interest rates due to inflation concerns, and so bailouts are still a possibility.”
The study was published in Nature Communications.
