No, Loiuse Bennetts, “Too Big to Fail” Is Not a Distraction

0 Flares Filament.io 0 Flares ×

American Banker, which generally defends the small guy, has a front page article today by Louise Bennetts wherein she claims that we shouldn’t break up the banks because the case to do so rests on five faulty
assumptions.

The only problem is that these five assumptions are either straw man arguments or simply aren’t faulty. What’s more, Bennetts misses critical reasons why breaking up the banks is a necessity.

Let’s address the five “faulty” assumptions.

Assumption 1: The failure of a TBTF firm will result in a systemic crisis.
Who’s making this claim? (Bennetts doesn’t say.)

We’re not concerned about one TBTF firm going under. We’re concerned about a systemic crisis that causes several TBTF firms to go under at once (see: financial crisis 2007-08). Dodd-Frank may be able to resolve a single firm that goes bankrupt, but we’re totally unconvinced (see some of the quotes here) that the US government will be able to resolve the problem of several TBTF firms going under in tight succession. It is exponentially easier to let smaller banks go under—as happens frequently already without even making headlines.

This is as good of a place as any to reiterate Nassim Taleb’s point about the megabanks from his 2007 work The Black Swan:

The increased concentration among banks seems to have the effect of making financial crisis less likely, but when they happen they are more global in scale and hit us very hard. … True, we now have fewer failures, but when they occur… I shiver at the thought. I will rephrase here: we will have fewer but more severe crises.

The point isn’t that making the banks smaller will completely eliminate systemic risk once and for all. The point is that in a crisis, it’s much easier to deal with systemic risk when that risk isn’t super concentrated and endlessly complex (as it

is in the megabanks). We must counteract the complexity of a global marketplace with simplicity.

Assumption 2: TBTF firms have a funding advantage over non-TBTF firms.
Here Bennetts says that some people are unsure about whether TBTF firms have a funding advantage, but then she says that, yeah, the megabanks do have a funding advantage but it’s the same advantage that all banks over $10 billion get.

Until we see clear reasons that Bloomberg’s claim that the megabanks get an $83 billion funding advantage is wildly off the mark, we’re hesitant to believe that it’s incorrect. Claiming that there’s uncertainty about assertions like Bloomberg’s doesn’t make these assertions wrong. Economists such as Anat Admati and Simon Johnson have repeatedly cited evidence that the TBTF have a funding advantage, as have others. Without serious evidence about why these sources are wrong, we’re hesitant to dismiss them.

Assumption 3: Systemic risk can be addressed at the firm level.
Bennetts says systemic risk can only be addressed at a market level. But this begs the question: Why can’t it be addressed at both the firm level and the market level? If a few firms make loads of stupid decisions (e.g. Citi and Bank of America), can’t we address some of that risk at the firm level? The answer is yes. This assumption isn’t faulty.

Assumption 4: Reinstating Glass-Steagall will address TBTF.
This is simple. Will splitting Citi’s investment banking activities from their commercial banking activities make Citi much, much smaller? Again, the answer is yes, it will.

It doesn’t matter if the primary goal of Glass-Steagall isn’t to address TBTF. Glass-Steagall is one way to address the issue, and one way that we’ve already been through before (and therefore roughly know how it works). We can do it again. It will help diminish TBTF.

Assumption 5: U.S. banks are dangerously and unnecessarily large.
Basically, Bennetts says that because foreign banks are bigger (which is true), US banks aren’t dangerous. But as Simon Johnson asks in the compelling video above, would we be better off if during the financial crisis Citi had $5 trillion, or $10 trillion, or $20 trillion in total assets? The Royal Bank of Scotland was much bigger compared to England’s GDP than Citi was compared to US GDP, and yet what good did the size of RBS do for England when it failed?

Andrew Haldane, executive director of the Bank of England, has published work (which Johnson bases his assertions on above) showing that once a bank grows above $100 billion in assets, any increases in size cease to give a net benefit for society. Just because foreign banks are larger doesn’t mean that US banks aren’t dangerously large.

Conclusion
Despite all her talk of assumptions, Bennetts fails to mention the most important reasons to break up the banks.

Breaking up the big banks will:
1. Decentralize their concentrated lobbying power
2. Diminish the likelihood that we’ll have to bail out bogus megabanks like Citi ever again
3. End the two-tier justice system that allows criminals at megabanks to avoid prosecution

In a sentence, breaking up the banks will better ensure that the United States has a fairer sense of justice. That is all. And that is a lot.

0 Flares Google+ 0 Twitter 0 Facebook 0 Reddit 0 Email -- Filament.io 0 Flares ×

Leave a Reply

wall street banks wall street banks wall street banks wall street banks

Join Us:





Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Google+
0 Flares Google+ 0 Twitter 0 Facebook 0 Reddit 0 Email -- Filament.io 0 Flares ×