By Peter Boone and Simon Johnson
The influential Goldman Sachs economist Jan Hatzius has a new research note out (with Sven Jari Stehn), “Thoughts on the Macroeconomic Impact of Basel III,” arguing that the move to raise capital standards for banks will put a serious crimp in growth in the United States – knocking 1.5 to 2 percent off gross domestic product in the next few years. Their findings are questionable, but in any case we should broaden the discussion to consider exactly how banks like Goldman Sachs affect our macroeconomic dynamics going forward – particularly if they are able to effectively lobby against higher capital. Growth based on risky banking has a tendency to prove illusory.
There are three issues. First, what is the short-term impact of raising capital requirements? Second, how should capital be increased? And third, and perhaps most important, do we really need global banks like Goldman Sachs to operate in their recent “high risk – highly variable returns” mode?
In their note, which is not in the public domain, Mr. Hatzius and Mr. Stehn are willing to acknowledge that raising capital standards can help make banks safer and that this is good for sustained growth over a sufficiently long period of time (think a decade or more), as the Bank for International Settlements suggests. But they make the case that raising capital – at least in the form that this is likely to take place – can slow growth over the next several years.
As they see it, forcing banks to have more capital (as a buffer against losses) relative to assets will increase their cost of capital – leading to higher lending rates and tighter credit standards.
On the merits of this point, Mr. Hatzius and Mr. Stehn unfortunately do not deal with or cite the detailed counter-arguments put forward by Anat Admati, Peter DeMarzo, Martin Hellwig, and Paul Pfeiderer – top finance professors (three at Stanford; Professor Hellwig is at the Max Planck Institute in Bonn), who strongly assert that raising capital requirements would have minimal if any negative effects.
They argue that government subsidies to debt financing of large banks, through tax incentives and implicit guarantees (which relatively penalize equity financing) are inefficient and distortive, and that higher equity levels would reduce this subsidy and lead to fewer distortions in lending decisions. In particular, Admati and colleagues emphasize that because raising capital requirements makes bank equity less risky and banks less prone to collapse, it lowers the rate of return they are required to seek for equity funding.
(The perspective here is that everyone is looking for a combination of risk and return – if you offer a safer asset, it is fine with investors if that pays less return; this, in turn, means that bank executives do not need to seek risk so aggressively.)
Mr. Hatzius and Mr. Stehn also overlook the point made persuasively by Anil Kashyap, David Scharfstein and Jeremy Stein (from the University of Chicago, Harvard, and Harvard, respectively) and their co-authors that even if forcing banks to hit a particular capital-asset ratio could be contractionary (because the banks will dump assets), that will not constrain or limit credit.
The Obama administration has had many missteps with regard to banking – and should have run more demanding stress tests in spring 2009 – but there is no question that when the time came to force banks to raise capital, they used the right approach: they told the banks the precise amount, in dollars, that they needed to obtain from private markets (or receive public capital on terms they would not like – including with salary caps.)
But even if Mr. Hatzius and Mr. Stehn prevail on the short-term issues (at least in the corridors of power) and regulators target capital-asset ratios rather than the more sensible Kashyap-Scharfstein-Stein proposal, so what?
The rationale for the capital increase is that in recent years the financial sector imposed massive losses on the rest of society by the mismanagement of credit. If the big banks have become a machine that provides supernormal returns to employees and creditors while causing frequent losses to taxpayers (through the fiscal costs, measured in terms of the increase in net government debt as a result of the recession), savers (because interest rates are cut to zero by the Federal Reserve’s policy response), and their own shareholders in many instances, then reducing the voracity of this machine is for the general good.
Correlations in economic data that suggest higher credit helps “cause” higher growth are not worth much – if they fail to take into account future losses on “bad credit.” If G.D.P. were adjusted with provisions for future losses, G.D.P. growth would be lower during periods where credit grows very fast.
G.D.P. in the United States in real terms is currently at about the same level now as it was in 2006. (Real G.D.P., annualized, was around $12.9 trillion in the first quarter of 2006 and $13.2 trillion in the second quarter of 2010, although this number is still subject to revision; see Table 3B (on Page 19) in the July 2010 report of the Bureau of Economic Analysis). If you expect G.D.P. growth to be weak for the rest of this year – as does Goldman Sachs – and also factor in population growth, which is about 1 percent a year, very soon we will all realize that G.D.P. per capita has actually declined over the past half decade.
If we had stopped the excessive credit growth in the second half of the 2000s, we would have limited the boom – and also removed a lot of the downside damage.
Highly risky, massive global banks that are – post-Lehman – unambiguously “too big to fail” are absolutely not in the social interest. If we give up some short-term illusory growth as a side effect of curtailing their activities, this is a small price to pay.
A version of this post appears this morning on the NYT.com’s Economix; it is used here with permission. If you would like to republish the entire article, please contact the New York Times.