Small Banks and Debit Interchange Reform: Winners or Losers?

The banking industry is making a full court press to get the Durbin Interchange Amendent implementation delayed. Given that a year of delay is worth about $15-16 billion in swipe fees, the banks are quite motivated, and they've managed to get legislation introduced to delay the rule-making so the Fed can study the issue further.  

A potent force in attempts to delay the Durbin Amendment's implementation are community banks and credit unions. While most debit swipe fees go to a handful of very large banks, small banks have been quite concerned about the impact of the Durbin Amendment. And community banks and credit unions carry outsized political influence because of their presence and standing in every community. The local Chambers of Commerce and Rotary Clubs, etc. are filled with the CEOs of community banks and credit unions, not the local branch managers of Bank of America or Citibank.  

But are the small banks right to be so worried about the Durbin Amendment? It's not so clear, and a recent American Banker reader poll indicated that a majority of readers thought that small banks would actually benefit from the Durbin Amendment (I've made a similar argument about credit unions benefitting from the CARD Act.) The Machiavelli of the card industry himself, Andrew Kahr, even opined in the American Banker that small banks would benefit from Durbin. In his words, the claim that Durbin hurts small issuers is "Poppycock." It's pretty hard to argue with Kahr's math, when it observes that it can only help small banks if they get 100bps vs. 30bps for the big banks. So what gives?

Small issuers are rather dependent on debit interchange revenue, in the sense that many of them do not to have huge profit margins, and a serious reduction in debit interchange revenue could threaten their profitability. (I did a study on this looking at credit unions, but I suspect community banks are in much the same boat). Of course, no one ever said that there's a right to make a profit through uncompetitive business practices (although that does seem to be the working assumption of the Wall Street Journal editorial page--are they even aware of the economic crisis caused by banks run wild?), and if banks fail because they can't compete, isn't that just the free market and all that jazz? Or does that argument only apply to nonfinancial firms?

In any case, here's why the Durbin Amendment probably doesn't harm small banks and likely helps them. The Durbin Amendment has two moving parts. Part one caps debit interchange fees at reasonable and proportionate to incremental cost of individual transactions. (The Fed very generously thinks this is in the 7-12 cent range.) Part two mandates that there be at least to networks competing for the routing of each transaction, with the merchant deciding on the routing (presumably selecting the cheaper option). In essence, part one would put a ceiling on fees, and part two would foster price competition beneath the ceiling as networks try to maximize market share. 

The Durbin Amendment contains an exemption for small issuers, defined as issuers with less thna $10 billion in consolidated net assets.  This covers 99% of financial institutions, but only a miniscule share of the dollar volume of the debit market on the logic that small issuers don't have market power and aren't creating the antitrust problems in the debit space. (Not surprisingly, Richard Epstein thinks this is an unconstitutional disparate impact...) The small issuer exemption applies only to part one of the Amendment--the price cap. It does not exempt them from part 2.  In other words, fee caps do not apply to small banks, but general competition does.

The question, then, is what will happen to interchange fee revenue for small issuers? That is largely dependent on two things. First, will there be two-tier or one-size-fits all interchange fees? And second, what will be the impact of routing competition? We know the answer to the first question, I think. It seems pretty clear now that there will be two-tier interchange fees. Visa has announced it will have a separate set of interchange fees for small banks. Several other networks have also indicated this (MasterCard has not, but it's hard to imagine that they won't, as they tend to copy Visa's pricing and, more importantly, can't afford to lose debit market share to Visa). So small banks will not end up under the Durbin interchange rate cap as a de facto matter.

Now that doesn't mean that they'll retain their current level of interchange fees, and that's their real concern. But I think there's a good chance that they will retain the current fee level. The networks have no incentive to lower interchange fees for small issuers--instead, their long-standing incentive to raise fees as part of competition to sign up issuers still remains. So putting aside Part 2 of the Amendment for the moment, I don't think there's much for small issuers to worry about.

But of course we do need to consider part 2.  We don't know exactly how the routing competition will work under the Durbin Amendment rulemaking. It's possible that there will have to be just 2 networks on each card or that there would have to be two networks capable of performing each transaction type on each card (2 PIN/2 signature or 2 PIN/0 sig or 0 PIN/2 sig).   I think the latter reading is the one consistent with the text of the Amendment and the policy goal encapsulated therein, but we don't know what the Fed will do here.

If the Fed opts for a weak reading of the Durbin Amendment, I don't think we'll see very strong price competition beneath the part 1 price cap. If so, there's not much for small banks to worry about overall.  

But if the Fed opts for the strong reading of the Durbin Amendment, then we will have some real price competition beneath the part 1 price cap. The question, then, is whether the networks care more about signing up issuers or about winning the routing on a particular transaction. I don't know what the answer is. If the network doesn't sign up the issuer, it's not even in the game for the transaction routing, but unless it gets the transaction routing, the network gets nothing. My best guess is that these competitive forces more or less blunt each other, meaning that we aren't likely to see major movement in interchange rates for small banks, irrespective of how the Durbin Amendment rulemaking plays out.

Maybe someone has a more advanced model of this, however--I'm all ears.  If I'm wrong about the impact of part 2, perhaps it suggests that the fix is just to exempt small issuers from the routing provisions of part 2.  

What I think this all means, then, is that the Durbin Amendment might actually be a real boon to small banks and credit unions, as it will start to level the playing field in the card space relative to the handful of megabanks that dominate consumer finance. There are lots of economies of scale in the card space, and the small banks are at a disadvantage because of that. But if they end up with higher interchange revenue than big banks as a result of Durbin, that is a step toward undoing the troubling consolidation of financial services around too-big-to-fail institutions. All in all, I think it's quite likely that small banks come out ahead under the Durbin Amendment, which is a reason to support it. 

Final, unconnected observation: nothing in Durbin would permit merchants to discriminate on a bank-by-bank basis. The idea that merchants wouldn't take small banks' debit cards is just silly. It'd violate their agreements with the card networks and would involve either (1) BIN recognition or (2) physical examination of cards at POS by cashiers. The first isn't possible currently, and it's hard to imagine a merchant that wants to bother with the second and tick off customers. 

Comments

[401k]

The Durbin Amendment should have been discussed thoroughly in the article. This will make people understand the benefits and risks of this amendment.

[Hangtime79]

Adam lets do some card math. Do you know why Chase and the money center banks are looking to cap debit card transactions? They will literally make nothing on the transaction and could take a loss.

Let's take the Fed's hard cap of 12 cents on a debit transaction and let's say Chase institutes proposed hard cap of $100 transaction size.

You quote in bps but bps on sales only remains the same when interchange is factored as a % of sales not a flat fee. So in the case of our $100 transaction the issuing banks makes 12 bps. Taking JUST fraud forget about transaction costs the issuer eats between 3 - 6 bps. So you pick up 6 bps on a $100 transaction minus transaction costs.

Now let's set the limit at $200, the issuer's interchange now went to 6 bps versus a 6 bps fraud costs (fraud rarely changes based on transaction size). The issuer now loses money on EVERY transaction. Banks will not eat this cost. We will most likely see "preferred" customers, those with 10K+ balances are higher expempted from this transaction limit practice, but most of those folks use credit cards anyway so its more a throw-in.

Of course as you point out it doesn't hurt smaller banks, but it does nail 90% of the dollar volume or in other terms 70% - 90% of consumers. The story has a little different spin when you say it that way.

With CARD Act (for better or worse), we have higher fees and less credit availability to lower income households, but actually more and better rewards for high income households as the industry shifts from relying on revolver based behavior to transactors where King Amex has always lived. Whether it was the goal or not, we are de-democratizing credit so that only the wealthy can get it.

With Durbin, debit rewards are gone - good chance we will see caps on transaction size across the board on debit. So in effect we have pushed the low income household out of the credit space, over to debit cards, but the debit cards will have their abilities cut therefore we will push these folks even further out into segments such as pre-loaded cards or even back into the underground. Of course, they can go to another bank, but once again we make the individual consumer's life tougher on the chance that it might better our broader policy goals. This is why I have a sneaky suspicion that the debit card caps as they apply to the money centers will fail because the law is making a distinction in treatment that has no basis other then well we don't want them to make money. I would actually like to here your arguments and why you believe this piece of the legislation and the special carve-out for institutions below $10 billion will actually survive a court challenge.

[Adam Levitin]

Hangtime79--Interesting comments. Aren't you forgetting the fraud cost variance in the Durbin Amendment? Doesn't that solve the problem you set forth.

Also, fraud losses in bps are derived from dividing annual fraud losses by annual transaction volumes on a portfolio basis. But when applied to a particular consumer, one can say that consumer X does $5000 dollars in transactions/year and at 6bps, that's $3/annually of fraud costs. So how many debit transactions are necessary to recover that fraud cost over a year? At 12 cents, that's 25 transactions. The size of the individual transactions doesn't matter as long as there's enough transactions. A handful of large transactions can be off-set by lots of small ones. That might mean more flexibility with caps. Or it might just set up a type of overlimit fee source for banks.

Ultimately, given that most debit transactions are for less than $100, I'm not too worried here. I'm also not sure why banks would want to cap debit transactions. Those transactions are going to shift to another payments medium. Some will go to credit, but some will go to checks and some to cash. Is that really what banks want? I think the answer would have to be cap, permit violations of the cap, and charge fees for it.

By exempting the small banks, there is a force that keeps the big banks honest--if JPM or Citi or BoA tries to do caps, they risk losing customers to small banks that don't have caps. I can imagine small banks advertising "non-capped debit cards," just as they might advertise "no-fee ATMs" or "free checking." I think this is actually a really important reason for exempting the small banks. In essence, the merchants are subsidizing the consumer protection benefit of not applying Durbin to all banks.

I think we should be careful about CARD Act claims--it's hard to sort out CARD Act from economic downturn and the Fed's card regulations. We're seeing direct mailings shooting back up, so I'm not so sure that there's a long-term de-democratization of credit.

As far as the legal issues, there is the TCF Financial litigation. It's got two arguments: disparate impact and takings. Both are weak, in my view. Yes, there's a disparate impact on big banks, but that's not unconstitutional necessarily. As long as there's a legitimate regulatory reason for the distinction (not a good one, just one that has some non-malicious basis), it should survive. I'd think the Fed could argue pretty easily that the distinction is reasonable enough given that big banks have market power, small ones don't. The Fed doesn't have to make the case that $10B is the perfect cut-off line. And given the sheer number of small business exemptions (and other <$10B exemptions in Dodd-Frank), I think this one will survive.

The takings claim is more complicated--it comes out of a line of utility regulation cases that say a public utility's profits can be regulated, but it must be permitted to make a reasonable profit (no threshold specified). Debit isn't a stand-alone product, as TCF Financial's CEO stated at the press conference about the lawsuit. It's part of a retail suite that includes the demand deposit account, overdraft, etc. Durbin only regulates one portion of this suite. It doesn't prevent the banks from making a profit (this isn't like regulating total returns on utilities). It's really hard to see this as a takings unless you think that _every_ federal regulation is a takings.

I don't think the TCF litigation will succeed ultimately. But it might score a victory in the District Court just because the forum/judge was so carefully shopped. But the chance that the 8th Circuit would uphold a ruling against the legislation is very small in my opinion.

But if I'm wrong, it also begs the remedy. If the offending provision is the small bank exemption, the remedy, might be to zap the exemption and apply Durbin to all banks. The TCF Financial litigation isn't being fought for small banks. It's being fought first to try and stop the rule-making altogether, but secondarily, to blunt the benefits Durbin might give small banks.

[Hangtime79]

Adam,

Your response deserves a longer and detailed response but I wanted to reply quickly this morning to two points.

Good source of material on Durbin-related research by the Fed's is here: http://edocket.access.gpo.gov/2010/2010-32061.htm

"Aren't you forgetting the fraud cost variance in the Durbin Amendment? Doesn't that solve the problem you set forth."

Durbin refers to Fraud PREVENTION costs, not fraud losses. Issuers are not free to pass fraud losses back through interchange through Durbin only the investments they make in anti-fraud technologies and resources. This becomes more like a utility rate-base case rather than a way to recover higher then average industry fraud loses. Even PIN based transactions according to the Fed have a 3.5 bps $ fraud rate or 3.5 cents per $100 spent almost exclusively taken by the issuer (96%). The question will be is this going to be handled on top of 12 cents or in addition to. Even if its in addition, the Fed pegged the fraud prevention and data security cost at close to blended rate of 1.8 cents per transaction.

"Also, fraud losses in bps are derived from dividing annual fraud losses by annual transaction volumes on a portfolio basis."

To ensure we are all talking the same language the Fed article above quotes its research in both unit and $ rate of debit transaction fraud loss as follows:

# fraud transaction / total transactions = .041% or 41 bps on a # basis
$ fraud transaction / total $ = .094% or 94 bps on $ basis

This tells us something interesting by itself, fraud-based transactions are twice the ticket size of an average debit transaction. Since the avg size of the ticket between signature and PIN doesn't vary dramatically ($37.14 and $41.34 respectively) we can infer the average fraud loss ticket is ~$80.00.

Back to the discussion:
Thus for every $100 in spend in debit transactions both merchants and issuers (consumers are very well protected from this loss and are negligibly impacted) eat 9.4 cents.

According to the Fed's data, within the system as it stands today of that 9.4 cents per $100, issuers eat 57% of the loss or 5.4 cents/ 5.4 bps $.

So let's assume a $100 transaction at 12 cents interchange (will count fraud prevention interchange as a wash since it will go to infrastructure).

12 cents per tran cap - 5.4 cents avg sales loss = 6.6 cents before infrastructure excluding fraud prevention spending.

Now let's move that number to a $200 transaction.
Since $ and # fraud rates stay fairly constant as sales $ volume increases. $200 / $100 x 5.4 bps = 10.8 cents.

$200 transaction
Issuer receives 12 cents per tran cap - 10.8 cents in fraud = 1.2 cents before infrastructure excluding fraud prevention spending.

By lowering the ticket limit to $100, the interchange flat fee of 12 cents becomes a larger proportion of the transaction which scales with fraud and increases profitability.

This is the same reason the networks charge both a % of sales component and cents per transaction. By including both you can effectively regulate the behavior being discussed by ensuring its always profitable for an issuer to take any sale not just small sales if its only transaction based interchange (in this case Durbin with just a transaction fee) or large sales if its just % of the sale (why would an issuer take 1% on a $2 transaction).

Since in the case of Durbin only a per transaction fee was added, the issuer does the logical thing - lowers the limit of spending on any one ticket. In doing so, the issuer has GUARANTEED themselves from losing money as a portfolio in debit, infrastructure costs notwithstanding.

Hopefully this helps your understanding.

[Diane]

I'm one of the many (middle class) consumers that would be affected by debit transaction caps. I use my debit card for everything, it's convenient. I carry very little cash and I do not have credit cards -- and that is my CHOICE. I live largely on a cash basis and well within my salary, and I like it that way. I may be old-school that way, as many of my generation may also be, but it works and I have zero worries about getting ripped off by some bank's credit card schemes. Once again, the banks are whining about not being able to rip off consumers with their endless fees and people like me will pay the price.

These caps would affect everyday transactions. When I buy groceries, the cost, especially with these rising prices, is well over $100, and I use my debit card. For me, and I'm sure many others, buying groceries on a credit card makes little sense. With a $100 cap on debit transactions, I would be forced to one of 2 alternatives, writing a check, or carrying large amounts of cash. The latter can be dangerous in today's world, and the former..well..how do YOU like standing in a grocery checkout line while someone writes a check??? What about prescriptions? Some prescriptions are ridiculously expensive, even WITH health insurance. Again, if you don't want or have a credit card, you're back to carrying a fistful of cash, or writing checks.

And forget making certain purchases on-line, which I do frequently and many have been well above $100. You can't buy on-line with a check or cash, and without a credit card, you're just screwed. Not every on-line store accepts systems like PayPal. I would either have to find the product I want locally.. usually paying higher prices IF I could find it.. or forget the purchase entirely.

The idea of a debit transaction cap is just another bank scheme while they whine about losing money.. boo hoo.. forgive my lack of sympathy. This sort of thing will have a ripple effect of consequences besides pushing some people 'underground'. I know many people who use their debit cards the way I do, and they also do not rely on credit cards and would not want to start either. While I understand the goal behind Durbin, the 'solution' to it that banks are proposing will cause a lot of fiscally responsible people pain.

If the debit card transaction caps were to happen, I will probably just stop using my debit card. It will become merely an ATM card and the banks will lose on all those debit transaction fees entirely. They may get a fee for the ATM transaction (which I do not pay now), but that's all. They make nothing extra on good old CASH transactions!

[Hangtime79]

Diane,

This all goes back to risk/reward. Would you loan someone $250 for 12 cents if you were going to get paid today? Ultimately this is what the debit card issuers are asking themselves and many don't like those odds. Since the issuers are responsible for over half the fraud costs (see my post above) and unlike the retailer there will not be a ton of margin to help offset the losses. I think retailers if they want to continue getting debit business are going to have to either pony up and accept more of the fraud loss if not all of it, otherwise we may go backwards (draconian debit limits) or towards more hybrid like credit card products. As a consequence of all this we may accelerate the adoption of chip and pin technology (a good thing) because their will not be any value on the debit side of signature versus pin use.

When all you have is hammer, everything is a nail.