Small banks are good for consumers

By Felix Salmon
April 1, 2009

Kevin Drum has a few practical issues with my idea about capping bank size at $300 billion:

What happens when you have a whole bunch of banks all operating at their maximum allowed size? Do they keep taking in money but just sitting on it? Of course not. Do they essentially shut down, not taking any new customers? What about natural growth among existing depositors? (For that matter, what about natural asset growth?)

Even more important, what happens when banks can’t can’t compete with each other by growing? What would they compete on? My guess is that they’d they’d compete on keeping the biggest, most profitable customers and would pretty much lose interest in smaller customers. So small depositors would find themselves increasingly unwelcome, paying higher fees and penalties, having a harder time securing loans, and so forth. After all, what incentive would a capped bank have to treat small depositors decently if they don’t don’t really want them in the first place?

The first thing that a bank does when it starts bumping up against the cap is to start selling off its assets: that’s why I’m a fan of old-fashioned securitization, where banks genuinely move assets off their balance sheet. Kevin’s worried that securitization only applies to assets and not to liabilities, and he wonders what will happen to bank deposits; I’m less worried about that. But fine, if deposits grow too large, then maybe banks can be allowed to exceed the asset cap so long as they invest excess deposits only in Treasury bills. (Or, to put it another way, if you sell off a bunch of loans, you have to invest the proceeds in Treasuries if your deposit base is greater than the asset cap.)

There are other things banks can do to move excess deposits off their balance sheets: they can move big deposits into externally-managed money-market funds, for instance.

As for banks competing with each other by growing, I simply don’t believe that they do. That was the point of Mike’s post, at Rortybomb: when banks get big they’re less competitive, not more competitive. And I think that Kevin is just wrong when he thinks that big customers are more profitable than small customers. Check this out, on the subject of Mexico’s biggest microfinance lender, Compartamos:

Compartamos, which had 1.16 million clients and MXN5.73 billion in loans at the end of December, saw its net profit rise nearly 28% to MXN1.12 billion in 2008, giving the bank a return on average equity of 44%, compared with an average for Mexico’s banking sector of 12.5%.

Yes, that’s a 44% return on equity in 2008. There aren’t many banks which can say that.

If there was a cap on bank size, then small depositors would actually become more attractive to banks: you get a much higher return on assets when you’re charging a $25 fee to someone with $500 in the bank than you do when they have a lot more.

So I’m not worried about consumers being hurt by caps on bank size. Quite the opposite, in fact: I’m pretty sure they’d benefit. Just look at credit unions: they’re nearly always small, and they’re nearly always great for consumers. We should take that model and run with it.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

people talk about counter cyclical capital requirement etc. couldn’t we have capital (and other) requirements that are related to bank size? That would act as a break upon size (ensuring bigger banks are less risk for instance – and banks would want to stay small to escape more onerous regulatory requirements), rather than imposing a crude measure of size which banks would instantly just work out ways to get around. Although I guess size based regulation would also require crude measures the banks could just work around.

Posted by Luis Enrique | Report as abusive

P.S. can you ask Reuters to give use temporary comment editing facility (say, 5 mins after posting) – lots of typos in my comment above, I apologise.

Posted by Luis Enrique | Report as abusive

In college, 30 years ago, I took an econ class based in part on Schumacher’s book Small is Beautiful. At the time I thought he was out of his mind. I now know he was brilliant.

I think we have banks too big too fail, too big to manage, too big for any economies of scale. They have to go. Is a possible solution to this the elimination of interstate banking? Interstate banking dates to the mid-1980s and was seen as a way out of the Thrift debacle (in part). Seems like it just lead us to where we are today. Time to unwind.

Posted by scott | Report as abusive

Felix, I have been a long follower and really like your blog. I found it particularly difficult to post comments on Portfolio, so I welcome this change. Thank you.

I agree with your argument except that I would make the limit based on a percentage of the overall assets of all the banks in the country. Other additional equations might help. But basing it on a hard number doesn’t account for things like inflation/deflation that would necessitate the inefficient task of revisting the law to reset the number.

I agree that we need to start limiting the size of our banks, but I’m really concerned about losing the benefits of the big institutions.

Mike downplays the convenience factor, but I really do value the ability to find a Citibank on every corner here in NYC, and to be able to access my account free when I’m in Chicago.

Also, don’t the big banks gain in operational efficiencies with their scale? I feel like it would be hard to get the same quality of website / technology / 24 hour service with a small bank.

Posted by ab | Report as abusive

Completely agree that big banks are dangerous (to all stakeholders except senior management who get a zero cost call option on upside performance of a colossal amount of capital: notionals count – only one CEO whether $100mn of capital or $100bn of capital, not to mention the leverage…) But I digress. What I don’t understand (well I sort of do, but that just makes it more appalling imo) is why all these other stakeholders put up with it. You shouldn’t have to mandate smaller bank size, it should naturally happen via optimization. The one obvious thing that was broken and needs to be fixed (and may have mitigated the excesses of the past) is the total mispricing of systemic risk insurance. You wanna be Citigroup or RBS; or even Bear Stearns (by virtue of a big derivatives book) fine, pay for the ‘too big to fail’ insurance. You could create a market in this – as long as none of the underwriters were themselves too big to fail and collateralized their positions with a haircut – to help the gov’t price it. I would imagine that as a function of balance sheet size and business complexity, the price of this insurance would resemble an exponential function. And wouldn’t you know it, if these giant financial mastodons had to pay this fair price for this insurance, growing the balance sheet to infinity would suddenly look less clever, even for the Larry Moe & Curly’s in the executive suite. It would be an ideal and very effective damping mechanism imo and avoid the need to make arbitrary decisions on how big is too big.

What we have here (he said glibly, with the luxury of brazen oversimplification) is just (another) tragedy of the commons. The banks over-exploited a resource (gov’t insurance) that was underpriced (at zero.) We were running the financial equivalent of a nuclear reactor without cooling rods. We shouldn’t be surprised we got a meltdown.

(btw would be great if Reuters could use Disqus in comments)

Would that we could have Kevin’s problem of an excess of deposits!

The core dilemma of banking over the past four decades has been the dramatic decline in the importance of deposits during the steady erosion via disintermediation of the banks’ privileged position. They lost to the capital markets on the assets side, played follow the leader into prop trading and debt issues, and bulked up their liabilities with funding from the markets. When the margins got too skinny on that, they turned to fee-based origination business. But it turns out that with SIVs and CDSs etc, they didn’t actually shift risk away from themselves, just produced a lot of contingent liabilities in return for those fees, which are now coming back to haunt them.

I’d like to see the banks go back to a much greater reliance on deposits on their balance sheets — not all the way to “narrow banking”, but a major step in that direction. There’d still be an important role for the capital markets. Use old-fashioned, plain-vanilla securitization for their capital markets business. As Felix suggests, securitization is a highly useful and, for the financial intermediaries, safe way of matching borrower and investor time, liquidity and risk preferences. The banks wouldn’t have to rely so much on wholesale funding for their liabilities.

However, the shadow banking system, which doesn’t have to pay FDIC premiums on their liabilities, now has a built-in competitive advantage over depository institutions that don’t resort to wholesale funding. That’s because we’ve discovered to our horror that it’s just as possible to have a run on the credit markets as runs on banks. So the government, without a formal FDIC system, still has to foot the bill.

In the long run, none of the proposals to limit the banks’ size or activities will make much difference if we don’t also charge the shadow banking system some sort of fee for the liabiltiies they create. The smaller banks Felix hopes to emerge won’t be able to compete given their higher cost of funds.

I don’t think we can leave it up to the markets to price in the higher risk that non-guaranteed capital markets obligations represent vis a vis insured deposits. During bubbles, when the hunt for yield is on, any market pricing of risk differentials is going to disappear. So some sort of capital charge or a transaction fee on debt creation seems to me to be an essential precondition of reducing the size of the financial sector and encouraging “reintermediation”.

I’m damned, however, as to how to design such a charge. And it would have to be an international initiative or we’d just be shifting the shadow banking system even further off-shore than it already is and encouraging even more regulatory avoidance schemes.