Simon Johnson, the former chief economist at the International Monetary Fund, is the co-author of "13 Bankers."
If Goldman Sachs were to hit a hypothetical financial rock, would it be allowed to fail -- to go bankrupt as did Lehman -- or would it and its creditors be bailed out?
I posed this question on Sunday to four experts (Erik Berglof, Claudio Borio, Garry Schinasi and Andrew Sheng) from various international organizations at the Institute for New Economic Thinking Conference, known as INET, in Bretton Woods, N.H. -- and to a room full of people who are close to policy thinking both in the United States and in Europe. (Let me note, in the spirit of disclosure, that I'm on the advisory board of INET, which covered my travel expenses to the conference.)
In both the public interactions (you can view the video and in private conversations, my interpretation of what was said and not said was unambiguous: Goldman Sachs would be bailed out (again).
This is very bad news - although admittedly not at all surprising.Why wouldn't policy makers allow Goldman Sachs to fail?
The simple answer is that it is too big. Goldman's balance sheet fluctuates at around $900 billion; about one and a half times the size of Lehman when it failed. All sensible proposals to reduce the size of firms like Goldman - including the Brown-Kaufman amendment to Dodd-Frank financial legislation - have been defeated, and regulators show no interest in tackling Goldman's size directly.
The largest financial institution allowed to go bankrupt post-Lehman was CIT Group, which was about an $80 billion financial institution. Some people thought CIT should be bailed out; fortunately they did not prevail, and CIT restructured its debts in November and December 2009 without any discernible disruptive effect on the economy.
Supposedly, Dodd-Frank expanded the resolution powers of the Federal Deposit Insurance Corporation so it could handle the orderly wind-down of a firm like Goldman, imposing losses on creditors as appropriate, without having to go through regular corporate bankruptcy. (After more than two years and more than $1 billion in legal fees, Lehman's debts are still not fully sorted out.)
At a news conference at INET on Friday evening - which I attended - Lawrence H. Summers, the former head of the National Economic Council, emphasized the importance of this resolution authority.
But the resolution authority would not helpful in the case of Goldman Sachs, a global bank that operates on a vast scale across borders. Such a case would require a cross-border resolution authority, meaning some form of commitment among governments. As this does not exist and will not exist in the foreseeable future, Goldman is, as a practical matter, essentially exempt from resolution.
If a bank like Goldman were in trouble, there remain the same unappealing options that existed for Lehman in September 2008 - either to let it fail outright or to provide some form of unsavory bailout.
The market knows this and most people - including everyone I've spoken to in the last year or so - regards Goldman and other big banks as implicitly backed by the full faith and credit of the United States Treasury.
This lowers Goldman's cost of funds, allows it to borrow more, and encourages Goldman executives - as well as the people running JPMorgan Chase, Citigroup and other large bank-holding companies - to become even larger.
No one I talked with at the INET conference even tried to persuade me to the contrary.
Given that this is the case, the only reasonable way forward is to follow the lead of Prof. Anat Admati and her colleagues in pressing hard for much higher capital requirements for Goldman and all other big banks. If they have more capital, they are more able to absorb losses - this would make both their equity and their debt safer.
Professor Admati was at the same INET session (her video is on the same page) and made the case that Basel III does not go far enough in terms of requiring financial institutions to have more capital.
Claudio Borio from the Bank for International Settlements argued strongly that requiring countercyclical capital buffers - that would go up in good times and down in bad times - could help stabilize financial systems.
But when pressed by Professor Admati on the numbers, he fell back on defending the current plans, which look likely to raise capital requirements to no more than 10 percent of Tier 1 capital (a measure of banks' equity and other loss-absorbing liabilities relative to risk-weighted assets).
Given that financial institutions in the United States lost 7 percent of risk-weighted assets during this cycle - and next time could be even worse - the Basel III numbers are in no way reassuring. Tier 1 capital at the level proposed by Basel III is simply not sufficient.
Even among smart and dedicated public servants, there is a disconcerting tendency to believe bankers when the latter claim that "equity is expensive" - meaning that higher capital requirements would have a significant negative social cost, like lowering growth.
But the industry's work on this topic - produced by the Institute of International Finance last summer - has been completely debunked by the Admati team.
Intellectually speaking, the bankers have no clothes. Unfortunately, the officials in charge of making policy on this issue are still unwilling to think through the implications. Capital requirements need to be much higher.