[abalone]

This is the same naive analysis everyone makes when they first look at the payments system. "Look at all that money. 2-3% on every transaction. A $500B tax. LOOK AT ALL THAT MONEY."

The reality is this: Most of that money gets passed back to consumers via rewards, benefits and consumer protections.

It's not a tax so much as an incentive for consumers to keep using their cards. And so it is considerably harder to come up with an alternative that is appealing to merchants without taking anything away from consumers.

For example, the interchange on a Visa rewards card for a typical brick-and-mortar retailer is about 1.5 - 1.65%. (Processors mark that up, but that's the "wholesale" fee that goes to the card issuer.) But many rewards cards pay out at least 1% cashback, on top of other benefits. That leaves a much smaller margin to compete over.

And remember, a new competitive option faces massive rollout and adoption costs that the entrenched system does not. Even the acts of changing behavior, upgrading POS systems and training staff are adoption costs. So your new alternative has to offer significant benefits for both merchants and consumers. Significant enough to overcome adoption costs.

Oh, and there's one more thing: debit card interchange just got regulated down to almost nothing (0.05% + 21 cents) by the Durbin amendment. So there already is an alternative, low-fee option that merchants can steer consumers toward and that they already support fully. So that pretty much takes out the opportunity to offer a lower-fee, lower-consumer-benefit option. That already exists now.

That leaves what? A higher-consumer-benefit option? Why would merchants adopt that? A same-benefit option but at a lower cost? But how much lower would the cost be while still matching 1-2% cashback reward programs and whatnot?

But.. but.. LOOK AT ALL THAT MONEY. :-)

(Btw, the digital cash for micropayments and garage sales and whatnot does sound interesting to me. My criticism is limited to the project of competing with Visa/Mastercard/banks.)

[jval]

If card companies are paying it all back to consumers, then what are their shareholders ending up with?

I agree with you that not all the $500Bn is going directly into the pockets of shareholders, but the reality is that there is a huge transaction cost in taking a clip and then passing part of it back to a consumer. However much is lost in the process, it might not be $500Bn but it is definitely a lot of money, and it is unnecessary.

There are a lot of arguments as to why merchants won't adopt Bitcoin for payments, the main one being that they actually need most of the features of modern finance that these companies charge for. The fees though are definitely an argument for Bitcoin, and not against.

[saalweachter]

Actually, where are all of these numbers coming from in the first place? I know I've heard a lot about the "credit card tax", but I can't seem to pull the numbers off the public data on these companies.

Visa has $10.4 billion in revenue off of processing $4.4 trillion in transactions; that seems to make the credit card tax a mere 0.2%, which is off by an order of magnitude from the conventional-wisdom "credit card tax". Where does this mis-match come from? Is the revenue hidden, and Visa is taking in a few hundred billion in revenue? Or is the revenue potential just much smaller than conventional wisdom says? (eg, the 2-3% and $X trillion come from different classes of transactions, and shouldn't be combined.)

[patio11]

Visa is not the only party receiving money on the transaction: the processor/merchant bank and issuing bank also both receive cuts which dwarf Visa's.

[saalweachter]

Hmmm, that is definitely a part of the mystery.

According to their annual reports, Chase, Bank of America, Citicorp, and Wells Fargo had a combined $15.3 billion in card services revenue/card fees.

If you add that to the $7.4 billion in revenue from Mastercard, $10.4 billion in revenue from Visa, and $27 billion in non-interest revenue from American Express (of which $17 billion was "discount fees", which I think means cash-back), you're up to $60 billion dollars.

I feel comfortable going from $60 billion to $100 billion just extrapolating from the top companies to the rest of the industry.

Any guesses on where the extra $400 billion is going?

[abalone]

It's not about whether merchants will adopt bitcoin. They'd be happy to take digital cash, or any method that pays out less to consumers.

The question is whether consumers will adopt it.. for more than the cases where anonymity is paramount. As it stands cards offer much better protections and benefits than cash, which is where most of the high fees go.

Not 100% of it -- networks and banks still do make money on the system. But the opportunity is much smaller than the gross processing fee would suggest. Between one and two orders of magnitude smaller. And when compared with the adoption and rollout costs of a new system, it's far less of a compelling case than cd's essay might suggest.

[cdixon]

My view is that the whole system was designed for a different era and not for the web. The problems include 1) consumer having to fill out a form at each merchant, 2) consumer having to decide if she trusts the merchant (1 & 2 lead to the strategic asset of "cards on file"), 3) the bank trying to determine with probabilistic algorithms whether it was really me paying, 4) the security of the whole system is very flawed. 5) all the money spent on marketing these services (including rewards) when instead the services should be baked into the internet.

That said, I agree with you that the digital cash / new behaviors are the most interesting part of Bitcoin. Just much harder to explain.

[abalone]

I agree with all those points, but the key question is whether the virtual currency community can come up with something significantly better to offset the considerable adoption costs.

Another consideration: whether the entrenched players can simply make minor adjustments to counter that threat. For example, take your first point, filling out forms. It used to be you always had to do it. Then contactless payments took off in Asia. There was all this buzz about NFC and how it was going to disrupt the system and mobile carriers were going to get involved and so forth. So what did Visa/Mastercard do? They just relaxed the rules so that swipes under a certain amount don't require a signature. Poof, there goes the opportunity, because there's no way the cost of an NFC rollout is worth the extremely minor difference between a swipe and a tap. (Go ask Google Wallet.)

And so it goes for most of the technical issues, I think. Trust and security too.. you could make it easier on the merchant, but at the same time harder for consumers to recover funds if they "did something wrong" like installed malware.

Excited about the new frontiers into digital cash though. And also scared too. The biggest use case for bitcoin right now is not payments. It's international money transfers that skirt capital flight controls. And, for us non-libertarians, it ought to be a big concern. It's a profoundly anti-democratic force. It gives the wealthy minorities (esp. in third world countries) a veto power over policy. Don't like some new environmental or safety laws, or higher taxes to fund public education and health? Just transfer capital out of the country and watch it wither.

[XorNot]

More importantly: NFC happened anyway.

Pretty much all my local merchants have machines that accept NFC payment taps - to the extent that the Commonwealth Bank has an app for the Galaxy S4 which will emulate a debit card tap and be accepted by any machine.

The only time I end up keying a pin or signing is usually when the staff themselves don't realize that I can just tap my card (the machines have big NFC icons on them, but they don't offer the machine to me to tap).

[abalone]

You might consider that the reason staff don't realize you can tap is that nobody uses it, because you can just swipe without a PIN for small purchases. I am not saying this is superior, just observing an example of how even changing consumer behavior is difficult when the value add is minor.

[XorNot]

No in this case it's because they don't present the reader to you - they punch in the sale, then ask you to hand them the card and swipe it.

Also in Australia I've not seen swipe without PIN used very much - a swipe usually requires a pin, whereas tap does not (and is also a lot faster).

[abalone]

Australia has different rules. In the US at least, the PIN is not required for small purchases.

It's interesting you mention Australia because they've taken a different approach to the fee issue. They regulated it away. Cards cost merchants very little. Didn't need bitcoin to do it, just passed a law.

Also interesting: it did not result in a lowering of consumer prices. The merchants kept the windfall.

[dlubarov]

If you compare interchange with rewards rates, they seem comparable, but most merchants pay much more than interchange. Direct relationships with issuers isn't feasible, even for the largest merchants. There are network fees and usually other middleman processors involved.

Visa and First Data each make >$10B/year revenue, which is already 0.2% of US GDP, and none of that money goes toward cardholder rewards.

[abalone]

Interchange is by far the biggest component. The second biggest part is the processor markup. The card network's cut (e.g. Visa itself) is extremely small in comparison.

Just to define these terms for our fair readers:

Interchange is the cut that goes to the bank that issued the consumer's card. That is standard and you can see Visa's rates here: http://usa.visa.com/merchants/operations/interchange_rates.h... Processors are the entities that connect merchants to the card network. (These are often also banks, or another org working with a bank.) One of the most important things they do apart from that technical function is vouch for their merchants. They guarantee that consumers are protected from bad behavior on the part of their merchants, like selling counterfeit goods or running tons of fraudulent cards. So that's why it's often a pain to get a merchant account, and why easier-to-get merchant accounts have higher rates (because they have more fraud).

So, yes, processor markup varies. But we don't need a new currency system to drive competition in the processor space. That's orthogonal. That could happen 100% on top of the current system.

So I'm not sure what else you're referring to as "network fees". Visa/Mastercard's slice? Visa's net income last year was $3B, which is not small, but hardly a major share of GDP. There's also decades of worldwide expansion costs (incl. massive marketing spend) that's gone into that.

[pfisch]

It seems like you need a new security system to reduce fraud which will reduce fees, and that is what bitcoin offers potentially

[XorNot]

Yeah that must be why people aren't losing Bitcoin and having it stolen every second week...

[pfisch]

That has nothing to do with the security mechanisms around bitcoin and everything to do with noobs and the infancy of the technology.

[XorNot]

It has quite a bit to do with the security mechanisms around Bitcoin, when compared to the security mechanisms around modern transaction processing.

Bitcoin advocates like to bang on about hyperinflation which isn't happening, meanwhile they're being robbed blind.

[rictic]

Aren't rewards also subsidized by interest rates charged on those cardholders who carry a balance at the end of the month? My understanding was that that's a very significant income stream, and that's missing from your comparison of ~1% rewards to the 1.5-1.65% interchange.

[abalone]

It's complicated. Rewards cards tend to be less competitive on APR, so they are less attractive to consumers who carry large balances. The wealthiest consumers who spend the most often have the best rewards programs and carry no balance. But you're right, that's a revenue stream that factors into it. There are other sources too, like penalty fees.

At the same time, some cards "magically" do more than 1%.. they do 2% or even 3% in some cases. You can guess where those funds come from.

But don't miss the bigger point here. The card issuer market is a competitive one. Regardless of where the revenue comes from, competition pressures issuers to hand more of that value to the consumer. It makes it a less lucrative business than the raw processing fee might suggest.

[kybernetikos]

> Total earnings for the year 2011 for the entire credit card industry were $18.5 billion, which was up slightly from the 13.6 billion earned in 2010.

http://www.bcsalliance.com/creditcard_profits.html

18.6 billion is still quite a bit of money to extract for the service of letting people spend their own money.

All those extra benefits should be unbundled and it'll be interesting to see how many consumers purchase them when their costs aren't being subsidized by cash buyers.