The AI ​​tool predicts when the bank should be bailed out

Artificial intelligence

Credit: Pixabay/CC0 Public Domain

An artificial intelligence tool developed by researchers at UCL and Queen Mary University of London can help governments decide whether or not to bail out a bank in crisis by predicting whether the intervention will save taxpayers money in the long term.

The AI ​​tool is described in a new paper in Nature Communicationsdoes not evaluate only if a rescue is the best strategy for taxpayers, but it also indicates how much to invest in a bank, and which bank or banks to bail out at what time.

The algorithm was tested by the authors using data from the European Banking Authority on a network of 35 European financial institutions deemed most important to them. global financial systembut can also be used and calibrated by national banks using detailed proprietary data that is not publicly available.

Dr Neofytos Rodosthenous (UCL Mathematics), corresponding author of the paper, said: “State bank bailouts are complex decisions with financial, social and political implications. We believe that the AI ​​approach we have developed can be an important tool for governments, helping officials assess the financial implications for their Specifically – this means checking whether a bailout is in the best interests of the taxpayer, or whether it would be better for the money if the bank was let to fail. Our techniques are freely available for banking authorities to use as tools in their decision-making process.”

Co-author Professor Vito Latura (Queen Mary University of London) added: “Governments and banking authorities can also use our approach to retrospectively review past crises and gain valuable information to guide future action. One could, for example, review the UK Government bailing out the Royal Bank of Scotland (RBS) during the 2007-2009 financial crisis and thinking about how this can be improved (from a financial point of view) in the future in order to benefit taxpayers in the first place.”

In a bank bailout, government investment in a bank increases the equity of the bank and reduces the risk of default. This short-run cost may be justified to the taxpayer if it results in lower taxpayer losses in the long-run—that is, it prevents bank defaults that do more damage to government finances.

In their study, the researchers created a mathematical framework to compare different bailout strategies in terms of expected losses for taxpayers. Factors considered include how long the financial crisis is expected to last, each bank’s probability of default and the impact of default on other banks in the network, as well as taxpayer stakes in the banks.

Using a mathematical control process, called the Markov decision process, the researchers included in this framework the effect of government intervention at any given time.

They then developed a custom AI algorithm to evaluate optimal rescue strategies, comparing non-intervention with different types of intervention—that is, different levels of investment in one or many banks—at different points in time during the crisis. AI technology is needed because modeling such a system is very complex, as the future behavior of all banks in the system can be infinite.

in their Case Study Using data from the European Banking Authority, they showed that a government bailout would be perfect only if taxpayer stakes in banks were greater than some critical threshold value, determined by the model. The optimal policy changed drastically once the loss percentage crossed this threshold.

Moreover, it has been shown that a government bailout tends to be more favorable the greater the distress of the network (defined in terms of a percentage reduction in banks’ equity), the longer the crisis lasts and the greater the banks’ exposure to other banks (i.e. the amount they lent to other banks, Thus, they would have lost if these banks failed.)

The researchers also found that once a bank gets bailed out, the best strategy for taxpayers is if the government continues to invest in that bank to prevent default. This could result in the bailed bank having no incentive to hedge against risk, which could lead to increased risk.

Lead author Dr Daniele Petroni said: “Banks have so far weathered the current economic storm caused by the COVID-19 pandemic. Their resilience has been enhanced by regulatory measures introduced in the aftermath of the global economic crisis. financial crisis 2007-2009 and by accommodating the monetary policies of central banks avoided bankruptcies across industries. However, no one can predict the impact on the financial system as well central banks Reversing previous policies, such as raising interest rates due to inflation fears, bailouts are still possible.”

more information:
An artificial intelligence approach to managing systemic financial risk through bank bailouts by taxpayers, Nature Communications (2022). DOI: 10.1038/s41467-022-34102-1

the quote: AI Tool Predicts When the Bank Will Be Bailed (2022, November 17), Retrieved November 17, 2022 from https://phys.org/news/2022-11-ai-tool-bank-bailed.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without written permission. The content is provided for informational purposes only.

Leave a Comment