By IESE Insight
Everyone remembers Lehman Brothers’ collapse in September 2008. But it’s worth fixing our attention on what happened just two days later: The U.S. government decided to spend $170 billion to save the giant insurer American International Group (AIG) from a similar fate.
Why was AIG saved? U.S. regulators said that AIG’s bailout was necessary, in part, because of its complex ties with banks around the world. So, the insurer’s failure could have led to contagion and systemic risk. On top of “too big to fail,” regulators were looking at “too interconnected to fail.”
Indeed, AIG’s many counterparties benefitted to the tune of billions of dollars from the U.S. bailout, albeit through the backdoor. First, domestically, the likes of Goldman Sachs ($12.9B) and Merrill Lynch ($6.8B) were compensated. Then, Societe Generale ($11.9B), Deutsche Bank ($11.8B) and Barclays ($8.5B) counted among the biggest winners overseas, as reported in the press. The “backdoor bailouts” helped these financial institutions weather the 2008 global financial crisis, but they also reinforced the incentives to maintain and even maximize the financial world’s interconnectedness.
Is banking regulation doing something wrong if all these interconnections put the system at greater risk? That is a multibillion-dollar question posited by Tim Eisert and IESE’s Christian Eufinger.
In their paper, the co-authors show through a theoretical model how banks benefit from being highly interconnected — and point out that it’s not ideal for the greater economic good. Essentially, banks have learned to exploit backdoor bailouts, realizing that if they extend themselves more on the interbank network, they are more likely to tap into government rescues in their own country and/or elsewhere. So, banks can ultimately boost their own returns by channeling their funds through other banks before these funds are invested in real assets, although they are increasing systemic risk in the meantime. Understanding how this works could help regulators formulate better policies to lower systemic risk and prevent future crises.
Shining a Light on the Interbank Network
To better understand the shadowy world of interbank dealings, the authors model the case of two banks and show that a high probability of a bailout incentivizes them to mitigate their investment risk through the interbank market, extending credit to one another. First, explicit or implicit public guarantees provide incentives to take on credit which a bank can then use to make investments, especially if it’s likely that their government will settle its creditors’ claims even if the investment fails. And second, banks can hedge further by lending funds to an intermediary bank rather than investing directly in an asset. If it’s probable that this other bank will be bailed out, when necessary, this offers extra protection, transferring part of the risk onto the governments involved.
In short, the authors’ model shows that banks increase their returns by channeling funds through lending to and borrowing from other government-protected banks before investing in assets. This makes the interbank network larger and more complex — and more vulnerable to contagion.
Furthermore, the authors say, this is a self-reinforcing mechanism. Banks with a higher probability of being rescued become more attractive as intermediary investment options for other banks, which means they become even more interconnected. The more interconnected a bank is, the more likely it will be bailed out if it collapses, and so goes the vicious circle.
What to Do About Perverse Incentives
The authors conclude that excessive interbank exposure increases bankruptcy costs and systemic risks. It’s a case of perverse incentives.
So what can be done? Simply put, they recommend that governments reduce banks’ incentives to become too interconnected, but this can be complicated. Options include introducing transaction taxes, limiting bailouts geographically, or introducing no-bailout policies. But for regulation to work, governments would have to take a coordinated approach to lowering bailout expectations. Their model shows that if one country removes its guarantees, banks subsequently have higher incentives to look for backdoor bailouts elsewhere and increase their international interconnectedness. Another avenue for reform is to make interbank activities more transparent, through a global clearing house.
Global collaboration isn’t easy to arrange, but financial systems could be sounder for it. For the 10th anniversary of the collapse of Lehman Brothers — and bailout of AIG — it’s food for thought.
Methodology, Very Briefly
The authors build theoretical models with different bank lending and borrowing structures and different government bailout probabilities. They determine the circumstances under which banks are incentivized to increase their exposure to interbank networks, due to government guarantees.
Eufinger gratefully acknowledges financial support from the Ministry of Economy and Competitiveness [ECO2015-63711-P (MINECO/FEDER, UE)].