I have been pondering transaction costs lately. Seems like this is an area where intervention could have massive impact: these costs are effectively a tax that incurs deadweight loss so according to prevailing economic theory if we can reduce it the whole economy should be catalysed.
IIUC governments already introduce price caps for certain elements of this market but I think we could be doing more? Like mandating incumbents to implement open standards and subsidising their competitors?
Dunno though I'm not very informed about the topic, just suddenly discovered recently how much of my income is just funding a fairly small number of payment providers.
The EU introduced price caps on interchange in 2015 capping interchange on credit cards at 0.3% and debit cards at 0.2%.
That’s almost an order-of-magnitude lower than interchange fees in the US, which are usually around the 1%-2% level.
This is big part of the reason you don’t see high value reward cards or cashback program in the EU. There simply isn’t the interchange revenue to fund them, which is good thing, because high interchange + cashback creates a nasty regressive tax that transfers vast amounts on wealth from those with poor credit ratings (usually the least well off society) to those with high credit ratings (usually the most well off in society).
"because high interchange + cashback creates a nasty regressive tax"
There's an interchange surcharge for rewards cards. Stripe bundles this all together to make the pricing simpler.. but most merchants use traditional merchant accounts, and they are charged based on the type of card used.
Doesn’t really change anything. VISA and Mastercard rules also prevent merchants from discriminating based on card type, either you except all VISA/Mastercard cards, or you except none.
So the exact manner in which they’re billed to merchants is kinda irrelevant, the cost ends up being spread over all consumers, regardless of their ability to recover the money via cashback or rewards.
Prices aren't set based on costs. Prices are set at the maximum the market will bear. If you've ever set prices for a product, you would see this yourself. Price*Volume=Revenue
Consumers dont care if you can pay for the inputs to your product. They obtain a certain value from your product and have a certain expectation of what the price should be, and if your costs are above that, consumers are happy to watch you go out of business. In fact, inability to increase prices over time while inflation increases the cost of inputs, is one of the ways the economy pushes marginal business out of the market.
> If you've ever set prices for a product, you would see this yourself. Price*Volume=Revenue
If your market is completely elastic, then sure. But only hypothetical markets (and perhaps some fungible goods) are completely elastic. Most markets don’t have a perfect linear correlation between price and demand.
The other things you’ve completely ignored, is the fact that interchange impacts all sellers equally. So as a consumer you can’t avoid the interchange tax by shopping around, which in turn will increase a consumer’s willingness to pay higher prices.
This may come as a surprise to you, but consumers don’t have a perfect way to calculating the value of something. Ironically, the perceived value of a product is highly influenced by both its price, and the price of its competitors (with Veblen goods being the extreme). So if you introduce a flat 2% price increase across all sellers, then it’s reasonable to expect consumers value perception to increase by a similar amount.
"But only hypothetical markets (and perhaps some fungible goods) are completely elastic."
That's exactly my point.. You can't change the price without changing the volume.. and that's why you cant pass on costs. The item is already priced at the optimal point in the price-demand curve. Moving the price based on cost increases, only results in lost revenue. The only companies that are able to do that are the ones that haven't priced their goods correctly to being with.
"then it’s reasonable to expect consumers value perception to increase by a similar amount."
You seriously think consumers know when the prices of the inputs for an item change? And then they adjust their perceptions of the item based on their detailed knowledge of how the product is made? The average consumer doesn't know how anything is made or what any of the costs of an item are. Go ask some random people what they think a credit card costs merchants... nearly guaranteed you only get wrong answers.
> That's exactly my point.. You can't change the price without changing the volume.. and that's why you cant pass on costs.
In an elastic market. This is not true in non-elastic markets. Obvious example is the price of ERs in the US. Pricing doesn’t impact volume in a meaningful way, because people don’t decide get injured based on how much an ER costs.
> The item is already priced at the optimal point in the price-demand curve.
This is a bold claim. You’re gonna have to provide some evidence that every item in world is somehow priced at the optimal optimal point on the price-demand curve.
> The only companies that are able to do that are the ones that haven't priced their goods correctly to being with.
You are correct. But pricing at the optimal points is the rare exception. Most items aren’t priced at the optimal point, because discoing the optimal point is incredibly difficult, and not a requirement for a profitable business.
> You seriously think consumers know when the prices of the inputs for an item change?
No, that’s quite obviously not what I wrote. I think consumers perception of value increases as prices increase.
You seem to have a great deal of difficulty identifying the differences between theoretical perfect behaviour, and reality. Very few things in the real world closely match their theoretically perfect behaviour, so theirs inefficient and market gaps everywhere. Something things like interchange can capitalise on. If you’re gonna come back with another response about how this doesn’t work in a perfect market, then I recommend you don’t waste your time, I’m not interested in debating if reality and theory match.
> either you except all VISA/Mastercard cards, or you except none.
There are many companies discriminating on card type; not accepting debit cards, not accepting business cards (which have higher fees for the merchant), not accepting credit cards, not accepting foreign cards, not accepting certain BIN ranges.
This breaks the Visa Mastercard rule. For which merchants can and will be fined.
Obviously smaller merchants can get away with it, because nobody cares enough. But get big enough, and an issuer is going get upset and start demanding the Visa/Mastercard enforce their rules.
Except when companies outright openly discriminate any form of payment other than cash with a surcharge that directly benefits the card networks because those players are huge and they don't want to miss out on this free money.
I'm mostly referencing Just Eat Takeaway dot com which have increased their online payment fees to 3% now (at least in CH) which is ridiculous because there's no way such a large player has such high fees with any acquirer.
> nasty regressive tax that transfers vast amounts on wealth from
As long as you're not a well off person with an AMEX card that regularly makes purchases outside of your AMEX card currency.
AMEX is nuts in that all non-home currency transactions go via USD.
So, for example, if you have a EUR AMEX and you make a GBP purchase, your conversion will go GBP -> USD -> EUR with AMEX taking a cut at each step along the way.
> This is big part of the reason you don’t see high value reward cards or cashback program in the EU.
Interestingly my US 1.5%-3% (regular, also 5% special) cashback cards keep paying out on transactions here in the EU. I've always wondered who is losing out in that equation.
You probably pay for it with the exchange rate spread. Your card may say no foreign transaction fees but they're already taking a 3% profit margin on the currency conversion.
I expected this, but I have tested this theory many times, and my credit card issuing bank or even debit card issuing bank offer exchange rates extremely close to the spot rate.
For credit card, I have tried this with Bank of America, Chase, and AmEx. For debit card, I tried with Schwab. And in all cases, I checked the currency exchange rate, and it was very close to the spot rate.
Where you do get hammered is by international merchants, who ask if you want to pay in local currency or USD (if you are American I guess). You should always choose to pay in local currency (meaning the merchant requests the payment in local currency from your card), because your credit card issuing bank will give you a much better currency conversion rate than the merchant will.
If you say you want to pay in USD, then the merchant will calculate a USD price that is much higher than the local currency price.
This. There's so much funny business that goes on with FX rates that it makes me almost pro-crypto.
Even the people who claim to offer raw spot market rates (ie. Interactive Brokers), in fact do not offer that. I would not be surprised in Visa & Mastercard do the same--although they claim otherwise.
Come to think of it, would Visa/MC be doing the conversion? I believe that's the issuing bank's job. But they offer the rewards, not Visa/MC, so it doesn't change the original answer much.
Visa/MC are doing the conversion, and that does makes sense to me given their central position in the transactions. There are (basically crooked) originating (I think that's the right word) banks/payment processors that convert it for you, with a 3% plus spread. Paypal is the most egregious in this category.
The rewards are funded by the banks via interest income. 3% cashback cannot be covered by interchange fees, for instance.
Visa/MC do the conversion, but it’s not unusual for banks to tack on extra fees. Additionally banks can choose to settle in multiple currencies, and handle the FX themselves.
VISA/MC do offer the spot rate. But the idea of a spot rate is a little iffy anyway, as it assumes that there’s a counterparty willing to sell/buy at that price, and at the volume you want to transact at. When you’re operating at MC/VISA scale, that isn’t a given, and neither of them are interested in taking on FX risk. So their spot prices will be close, but likely not identical to spot prices you might find at other FX providers. Over the long term though, I would expect any delta to balance out.
When I took my Schwab debit card to Europe, the exchange rates that I got were identical to the ones posted on Visa's website, suggesting that it's Visa doing the ForEx conversions.
A lot of banks do charge a percentage fee on top of Visa/MC's exchange rate, but my bank does not.
Schwab and Fidelity debit cards are the only ones that do this AFAIK. In addition, you get unlimited ATM fee rebates, even outside the US. I’ve stopped taking cash home and instead withdraw from my Schwab account :)
The answer is in the original Stripe article: The merchant pays much higher fees when you pay with your US credit card than when you pay with a EEA credit card because the interchange limits only apply to purchases within the EEA made with a "standard" EEA credit card (corporate cards are also exempt, which is why Stripe is charging higher fees for them as well).
Last I looked at the numbers most transactions were covered by the Durbin Amendment. Interchange fees for exempt single message (PIN) transactions had also fallen to basically the cap. It was really only exempt double message (signature) transactions that were still expensive.
How do credit cards provide 60 day interest free loan with only 0.3% charge to the merchants? I can see low interchange rates for debit cards like the US the interchange for debit card for major banks is capped at 0.05%.
You can doubt it if you want, but the Credit Reference Agencies make a big fraction of their money because this is in fact possible.
Say I'm Big Bank, and I want to launch a card I'm calling "Bonus Credit". I want to target customers who mostly will pay me, eventually, but aren't so flush that they can pay their full balance every month (e.g. they buy their summer break in January, pay it off over the next 3-4 months). I call Jumbo National Credit and I offer them £25 per sign-up if they give me customers who have no super-negative credit events in the past say, 12 months but are say 75-95% on recent payment history. Jumbo present this as a "Pre-qualified" instant Credit Card option for customers who match the profile, on their free web site. Click now, you are guaranteed to be accepted. Customers don't even realise they're reading an advert.
You may have noticed that although 30 years ago "Credit score" was something you needed to go out of your way to discover, today you can get your score on a web site - sign up today to see it instantly. Why are these companies telling you something that was once an important trade secret ? Because these sites are now effectively marketing for 3rd party credit products. AIUI Experian invented this, I worked with people who were in a small rogue group at Experian which proposed instead of fighting the government to keep it secret they should turn access into a revenue source, they were laughed at internally more or less up until the record profits rolled in and then suddenly it was champagne and bonuses.
No, not in Germany. I still believe revolving credit cards aren't something targeted at people who'll miss paying their balance in full in most of continental Europe, and issuers can't really base their profitability on that alone.
Ironically enough that's why a company like Klarna became so big over here, the whole "buy now pay later" concept is completely new to most low-income customers.
There is often a small annual fee connected with having a credit card (I think I pay on the order of $25 a year). Also effective interest rates are in the 15-25% range for most cards if you don't pay it off in full, which I'm sure quickly adds up.
0.3% over 60 days (in practice more like 15 to 45, only commercial cards have longer billing cycles) was really not that bad during a decade of near zero interest rates.
And now that banks will finally make a bit more money on mortgages and other kinds of loans, I assume they don't mind the increasing cost of their customers' credit balance.
There are lots of other ways credit card companies can make money:
- Higher interchange fees when you purchase something outside of the EEA
- Currency conversion when you buy something not in EUR
- High interest payments when you do not pay off your balance in full on time
On average, it will also not be a 60 day interest free loan: the "average" transaction will happen in the middle of the billing cycle, so that gives you 15 days until the invoice is sent, and many people I know pay their invoices much earlier than after the 30 days payment term.
Key part of the phrase "high value". Some US cards give back ~2% on all purchases.
The best of those appears to be 1% on purchases at six specific retailers or 0.1% on purchases in the EU/1% outside of the EU. Many of them don't even give cash back, they give their own cryptocurrency that they made up out of thin air.
This is the answer. In the US the business essentially funds the rewards these cards provide because they can charge whatever interchange fee that card has decided.
Stripe provides value-added 'premium' service which competes with merchant banking solutions offered by most EU banks, and Stripe does not have a dominant position. But in general, yes, EU does place a LOT of attention to these deadweight transaction costs, and has had large interventions - for example, if you look at the price listed in this very article, the reason why 'international cards' have 3.25% fee and 'standard EU cards' have a 1.5% fee is primarily caused by EU-imposed changes to interchange fees which get paid by the acquiring/merchant institution (including Stripe) to the issuing bank.
My years in fintech have taught me that it'll be very difficult to get these charges down without legislation. What they are is, essentially, insurance.
For Visa/MC, "running the network" at cost is possible on much less money, but the network involves a lot of elements along the chain that can get transactions reverted. The "insurance" (1-3% of tx) pays for the legal-adjacent issues related to handling those transactions being contested.
Go to a restaurant, use a US terminal? Great, the restaurant owner can modify the transactions after it's been authorized and long after you're gone, "because tip". There's zero checks on this, it's a matter of "it works because most people don't do it". So when it happens, sometimes the payer notices and issues a dispute and the dispute management is part of the network. This is where a lot of costs go.
Anyway, the network is ridiculously bad. The fraud/aml checks are not usually shared among payment providers, because they can just milk each other instead by selling their checks. No wonder it's hard to get the transaction fees down.
And yeah, it's a duopoly. The solution by the way isn't to directly try to build a global visa/mc competitor; with the moat, that's impossible.
Rather, it's to build on top of what European countries are doing. EU countries have built their local competitors (Bancontact, EPS, BLIK, iDEAL, Sofort, ...). Those have lower costs, and so would any locally-targeted provider because they each have to deal with less risk and complexity. Any aggregator (like Stripe is, by the way) can take payments for all of them and push people away from card payments. The problem with that very last part is a UX issue, people like paying by card.
It's a difficult problem. What makes it especially difficult IMO is that once you're down this path and become successful, it takes some very specific, very early business choices in order to be able to turn down the mountains of cash that show up to your doorstep in the form of "align your fees with the rest of the industry".
Yup exactly, but you need to take that strategy and apply it beyond europe. For this you need local networks in the US as well for example. Right now, there aren't really any; some crypto based local networks but that's all.
But the US is undergoing some serious banking network updates this year and that likely means some new local networks are being built. This idea will look very different by 2024 and maybe even be feasible.
Unfortunately EPI has stalled, I don't know if we'll ever see MC/Visa cards co-branded with a pan-European scheme.
However, the free market approach seems to work for mobile payments solutions such as the extremely popular ones in nordic countries. They have started to build an interoperable and federated network in Europe and beyond. For instance, there's a real demand from businesses that want to let Chinese or Indian tourists pay with AliPay or UPI.
Don't get me wrong, I think DSP is one of the greatest thing the EU has ever done, but it looks like physical cards just need to die before we can see a true competition.
You're referring to the launch of FedNow, right? It's interesting because most of the analysis I've seen keeps assuming it'll be more B2B and that people will just continue to use P2P apps. However, I don't see any reason said apps won't move to FedNow themselves. The underlying ACH should be fairly easy to swap out with FedNow, right? Venmo and CashApp have only free money (in the form of lower fees and faster settlement) and better security to gain, it feels like.
Yeah FedNow. It will be b2b but this means an easier time to launch fintechs in that space that make use of it. Consumers will benefit, the iteration time is just a bit longer.
I wish those transaction cost would be made visible.
Imo the seller should not care which card the client use, if it's more expensive to use a "premium" card then it should be made visible to the client and he should be the one to pay the overhead.
Yes this is all that needs to happen. The EU should mandate that sellers can pass transaction costs on to consumers if they want (but no more than transaction costs, looking at you airlines).
> EU should mandate that sellers can pass transaction costs on to consumers
EU actually mandates the opposite. It is illegal to pass the transaction costs to the consumer. The price you see on an item has the be the final price you end up paying.
In many online shops in EU you can see "Administrative fee" if your order is too small for example. So while companies are not allowed to do "Credit card fee" they still do it indirectly in one way or another, those that don't impose extra fee usually have higher delivery fee.
Brazil has a similar rule, but the price you see on an item is the largest one you may end up paying. So, the seller can still reduce it if you choose a cheaper payment means, and many of them do that.
I love them for it. When I don’t live in Europe it’s infuriating to see the price levitate after every checkout step. Thankfully it’s still not as bad as it used to be like 10-15 years ago.
One influence in the setting of this rule was that payment card processors had been throwing their weight around requiring that merchants not pass on cost, which was fairly clearly an abuse of their position of power. (For most in-person business, which was largely what was being dealt with at the time, cash is essentially costless, and so for some small businesses being forced to swallow payment card fees was genuinely an unreasonable burden.)
There is no US sales tax. You're thinking of state and local sales tax, which in most states (maybe all that charge a sales tax, I don't know) generally added on top of the sticker price. I wish it were included (when possible), because it can be frustrating -- especially to those of us who live in states that do not have a sales tax.
For taxes, in US eCommerce transactions at least, that generally isn't possible until checkout anyway you need to know where you are shipping the order to apply correct taxes. For brick and mortar shopping, I agree, taxes should just be included in the label price.
A breakdown into what part of the price is set by the business and what part is set by the government seems to fit your provided definition of “visibility”.
I don't see transaction fees as fundamentally different than any other cost to the business. Do consumers also need to know how much the business pays in electricity, or the rent the business pays?
On the other hand, transaction networks clearly add massive value, so it’s very likely that net of fees (without which the network may not exist at all) these networks still add value.
So it’s more a question of finding the optimal equilibrium rather than viewing fees as “deadweight”.
“Dead weight” might be too harsh a term, but I think “rent seeking” is spot on. And it’s generally not a good idea to let rent seekers extract value in proportion to the damage they could potentially cause.
Absolutely. And for the colloquial English phrase "dead weight" that would correspond to the portion of these fees which go above and beyond the cost of running the (high-utility) transaction network. Reducing the rent-seeking on this network also reduces both the colloquial and technical meanings of "deadweight".
But for the economic term "deadweight", additional deadweight loss can be reduced by anyone who can provide similar functionality to the current interchange network but at lower cost than the current network. This wouldn't reduce the colloquial "deadweight" because in theory 100% of the fees could already be going towards necessarily funding the existing network. But it would still reduce the economic deadweight of 1-4% tax on every transaction.
I don't see a reason something as critical as transaction networks should remain privatized, it should be operated by society (via democratic government), not private interests.
We don't have private corporations operate highways or the federal mail system. They may operate (heavily regulated) airlines, but not the traffic control that manages their interactions.
Imagine having to pay a "takeoff clearance fee" or a toll station at every highway on-ramp.
Private highways are very common in Europe. So is mail delivery by private companies (Deutsche Post is a big one).
In Portugal, the mail company is private and 90% of the highways are operated by private companies. And this is a country where the Socialist party has been in power for 20 of the last 30 years.
That is highly surprising to me. How do you have 90% privatized highways without toll booths everywhere? Without a universal funding source like taxes how is construction and repair funded?
I assumed it was mostly to pay for chargebacks and fraud. What I think is crazy is basically anything < $10 doesn't need chargebacks, fraud protection, etc, and the fee is inordinately burdensome for those transactions.
On a $1 itunes song it would have been an >30% cut, you can forget about selling any small digital thing worth less than that.
I agree, but that's saying because credit card suck in way X, they should suck in way Y as well to accommodate. Stolen credit card information shouldn't be a thing, and people shouldn't be passing security critical 20+4 digit codes around (think oauth).
Why wouldn't small items have chargeback costs? Do credit cards ban chargebacks for them? If someone buys something cheap, it doesn't come, and the credit card company refuses to allow a chargeback, that's going to make some customers mad.
Would you really file the chargeback paperwork for a $1 purchase? Even if you're scammed, it's at the level where you say "I'll be careful next time" and move on.
* Money. This doesn't matter much too me, but it might for people who make a lot less than me. The comment I replied to was about anything less than $10, not about $1 specifically.
* Preventing future scams. By doing a chargeback against scammers, you're making it more likely they'll get banned. So you're helping to prevent other people from getting scammed in the future. It's a good deed. When I see phishing sites, I report them. Why wouldn't I report a scam?
* Revenge. Some people want to see the bad guys get punished.
I would not call it revenge but yes, if you do nothing because it is low value or not important enough for your time, bad people will do bad things more because they are not punished. I try to report spam to the registrar and hoster and I report fraud to the bank for that reason.
Solving coordination problems through voting. Something as fundamental as the payment system should have people competing on making the underlying processes more efficient, not building the biggest network lock in effects.
By raising prices, they are encouraging competition. Eventually the prices will reach a point where people stop using the product. Until then, I guess people can try to get some laws passed to force their favorite company to do what they want. Of course, more laws usually raise the price of entry for competitors.
> IIUC governments already introduce price caps for certain elements of this market but I think we could be doing more? Like mandating incumbents to implement open standards and subsidising their competitors?
This is a very dangerous path to go down. Direct price control by governments usually leads to unforeseen consequences and often to disaster. Mandating open standards is probably a good idea.
UK/EEA businesses who are paying out in USD to a US-domiciled bank account will now incur a 1% fee, with a minimum fee of US$2.50.
My business does most of it's sales in USD and pays out to a US bank account, I don't understand this fee at all. Just because they can I guess? Stripe is really becoming the new PayPal. Will definitely be exploring options to move to Adyen or continue moving more customers over to Paddle as we've found Stripe to be increasingly frustrating to deal with.
Now we're looking at 3.25% + £0.20 for the card payment in the US, 0.5% for Billing to handle subscriptions, 1% to payout, 0.4% if you want invoices, 0.5% if you need to handle sales tax. Already at 5.65% + £0.20 - that's without any of the paid radar tools.
I thought just the base fee was high but these rates are insane. Where I live, the taxes we pay are less than this. It's not like stripe needs this money to build roads, they are just moving digits.
> Dispute fees (also known as chargebacks) will increase from €15 to €20. Due to costs for managing dispute evidence submissions (regardless of outcome), we'll no longer refund this fee if the customer's bank resolves the dispute in your favor.
> If you or your users have an account located in an EEA country that has not adopted the Euro, here are the fixed fees of €20 in local currencies: Bulgaria: ЛВ40; Czech Republic: 550Kč; Denmark: 200kr; Hungary: 7,000Ft; Liechtenstein: 20CHF, Poland: 90zł; Romania: 100LEU, Sweden: 200kr.
200 DKK is almost 27 EUR. This seems a bit expensive considering that you don't get refunded if you're in the right anymore.
That merchants hold almost 100% of the liability for fraud even if they do everything "right" to vet a transaction is crazy to me. The other parties (issuing bank, payment processor, payment network) are in a much better position to combat and detect fraud, since they see ALL the transactions. But the incentive to do better is low, since the merchant not only holds the bag, but sends them a penalty fee.
Disputes are not always fraud on customer side and it’s really not the job of the bank to figure out who is wrong. Imagine two wallets disputing the fate of a coin on a bazaar - does it really make sense? E.g. if content of the package is stolen at warehouse or something else happens, where you think you sent the product and customer got nothing, customer may dispute your point of view and request chargeback. There’s no easy universal answer here, so it’s better to look at it in a constructive way and factor the risks into the price. Of course, bank or payment processor can offer an insurance product in this case and include this risk in price of transaction instead of collecting the fees. But this is just a game with financial models and putting labels.
Agreed, though the other 3 parties know a lot more about the customer, past disputes, and so on.
Also, the argument applies for any kind of chargeback. Again, the other three parties combined see all transactions and have lots of context. And, if they required things like ip address, shipping address (they only ask for billing), they could really improve fraud detection. But they don't...there is an incentive problem.
This is convenient for merchants to offload fraud prevention to banks by sharing more data, but then this should be a service for which you pay and a section in your privacy policy, so that your customers know which data is shared with the bank and why (I‘d say that would be too much data). Fraud levels are different for different merchants, so bundling this service with core product would be unnecessary tax for those businesses which rarely deal with fraud.
It makes taking cards via Stripe completely untenable as anyone can dispute a legit charge and you're still out of pocket - imagine being charged 20 EUR for a 3 eur order even if the customer actually ordered it and just felt like being a dick
You just refund 3€ order and move on. This is the reason why marketplaces implement sophisticated solutions for returns — sometimes it is cheaper to let a few customers to get their product for free than to spend ton of money on returns, refunds and legal disputes. You can always reflect this risk in your prices.
That's fine if the customer actually asks for a refund in the first place. All the chargebacks I've handled through work (admittedly only in single digits), the customer just started a chargeback with no prior communications.
It's a pretty frustrating process as we've got no issues giving refunds in most cases.
My comment was specifically where the customer is being a dick, ie; not asking for a refund, or being refunded and then disputing it anyway (in which case you're now 20+3+fees down which is even worse)
They're claiming that it costs them 20 euros to manage evidence that the merchant uploads? What does this process entail, other than forwarding that evidence on to the customer's bank?
I find the dispute fees to be high already, since a customer can initiate a dispute with near-zero effort, but then I have to run around and gather evidence showing that their claim (for which they have to provide zero evidence) is not true. If someone has a subscription with me and it's renewed for several years, and then they decide they don't want it anymore (but don't tell us), they can initiate a "fraud" claim with their bank, which means that I have to prove it wasn't someone else who used their card — with their email address — to set up the account years ago.
I have usually been able to email the customer, ask if/when they asked to terminate their account, and submit that as evidence that no request was made prior to the fraud allegation.
But there should be a higher standard for the initial claim (some evidence should be required, which would at least ensure that claims are correctly categorized as "third party fraud" versus "I forgot this was a subscription").
It says something when traditional banks are basically the same price now and are almost as good in features (and many other card processors are cheaper, especially if eg you stick to a processor in the EU for EU cards) - I'm not sure why anyone would pick Stripe anymore, especially as they have just as many "hidden" costs as others (eg fraud stuff)
If you're working with COBOL records then changing the number of characters of any field, or adding or removing a field is pretty much impossible.
Which leads to funny questions at development time like "how many characters do we reserve in the customer record for the third child of the second wife of customer X, when they're remarried, this third child is not a child of the customer but there does exist an alimentation 'agreement' between the customer and their third wife". You must make this decision knowing that whatever your answer ... it can never be changed again.
>500kb per record. And, of course, mostly it just has first and last name, address, birthdate and balance, nothing else.
Well... yeah, but no one on their sane mind would build the backend of a mobile app or of a web app in COBOL.
Of course you can have all your COBOL and your mainframes down below rolling and churning like it's 1975, but the authentication of an app is layers up above in a totally distinct system. Or this or you're crazy.
You will find there is often a big difference between logging into the account and making any change or transaction. Getting the password right might give you account details and some basic transaction history, but it won't allow you to transfer money without 2FA.
True, but you assume 2FA is implemented correctly and doesn’t have a bypass. What I’m saying is that they’re often poor at what they do, so I don’t put excessive trust in them.
For quite a while they've been trying to emulate fintech style ease, and to some degree they have made it (although most still insist on their own hosted gateways, which is fine as SAQ D compliance is a real chore) - they do things like OOB webhooks etc, like Stripe/Paypal/whoever do, so really the flow and integration is pretty much the same
edit: my experience only applies to UK (which were better and cheaper way before EU) and EU banks, the US is a minefield that actually is better avoided by just paying higher fees to a popular processor rather than deal with banks who think its the 1800s)
Stripe stopped being competitive when they began keeping all of the transaction fees, including a percentage of the entire transaction, when issuing refunds. A customer changed their mind... well you just lost a lot of money.
Imagine you sold a product on the internet, you paid Fedex to ship it. Your customer decides they don't want your product anymore. You refund them their money, but then you have to refund them the shipping cost. Not just the cost to ship it to them, but they want you to handle the return for free. Yeah, that costs money. Transaction fees are not free. Stripe doesn't own the credit card network.
No that is incorrect. The credit card network returns most of the fees. They absolutely do not keep a percentage of the full amount. The credit card network might keep $0.3, but Stripe decided to keep the full 3-4%.
I agree but to be fair, that was always an expense to them. They just took it on the chin and used it as a competitive advantage/point of differentiation and it helped them grow. Especially makes sense as they initially targeted markets where software trials were being sold.
The carding ecosystem is going down. In the last two years, I saw a surprising uptick in the number of businesses that will charge a 2-3% fee if you pay with card. This is from the US to Europe to SEA.
My guess is that most commerce will start offloading this cost to customers especially as new payment methods become available (ie: wallets)
Lest government impose cards monopoly and force transactions with them, I think we have reached/are close to the peak of mastercard/visa.
Iirc in the US at least the gov’t finally struck down the “must charge same as cash” requirement that the credit card companies were forcing on everyone.
And inflation has given them incentive to start actually doing it - at least two restaurants here have a cash discount now.
> Dispute fees (also known as chargebacks) will increase from €15 to €20. We’ll also no longer refund this fee if the customer’s bank resolves the dispute in your favor, due to the costs Stripe incurs for managing dispute evidence submissions (regardless of the outcome).
Is Stripe testing dispute fees for all chargebacks out in EU before rolling out in the US?
1) This is gonna be bad for NGOs
2) So now people can just DDOS an org with chargebacks to put them in the red?
I wondered if it's a practice run for the US, although it's possible the process for "managing dispute evidence submissions" is more involved in the EU. They seem to do next to nothing in the US — just forward the docs you upload to the customer bank, I assume.
I agree with 2 — this would be really bad if a business were targeted with chargebacks. There would need to be some defensive mechanism that kicks in and retroactively refunds all these fees in such cases.
I also wonder if this is just a way to pump up their margins in advance of an IPO.
We had someone try to use hundreds of stolen credit cards on our nonprofit's donation form to test them. What a nightmare! I really hope these fees don't get charged to nonprofits.
It’s interesting how these fees slowly inch up. It’s a slippery slope argument or more like a glacier. But once the market stabilizes, I think there’s lots of pressure to eat these fees over time.
I wonder if everyone will end up with the apple model of only getting 70% and letting some huge conglomerate fight over fees, or just vertically integrate.
I’m not a fan of government regulation but this is an area where it may be worth having some digital cash mediated exchange where the transaction fees are meant to be absurdly low, like 1% or 5 cents whichever is lower. It would provide all the insurances of cash, so none, but would be a strong financial infrastructure that helps consumers and business.
This has been my hope for crypto since 2009, but the fees have always been higher than visa for consumer purchases.
I don't know how Stripe can justify these pricing, except for "oh we can extract more money, let's do that!". In other (non-Western) countries pricing for local payment systems can be less than half of what Stripe charges, e. g. Alipay (one of the biggest Chinese payment services) charges about 0.55% merchant fee without per-transaction cost.
Even in the EEA if you only care about payments, nothing stops you from going to a cheaper alternative. For instance https://www.stancer.com/en/pricing
And I don't know the rates big players negotiate with Adyen but it's probably similar.
Yeah but it's not the same economy of scale, you'll reckon. Alipay is also completely systemic in China and cant exactly be considered a private entreprise anymore: they're "more famous than Jesus" in their own way.
I barely ever heard of Stripe as a European, and now that I live in China, it's almost mandatory to have alipay.
Why don't Stripe customers shift to Alipay? I don't know anything about the service, but that seems like a pretty straightforward win, if they're as similar as you suggest.
The key part of the parent comment was 'local payment systems' - that rate is for Alipay payments from China-issued payment instruments, it's not for accepting Visa or Mastercard. It wouldn't be enough for Stripe customers (merchants) to switch to a cheaper payment system, the actual consumers would have to do it as well.
Because countries where both work are not completely overlapping. It'd be very interesting to compare Stripe market share in countries where there are strong competitors.
That only applies to the card networks not merchants or processors, but there is also a dual rate system that the EU are obsessed with where intra EU is much cheaper, they mandated the same for voice calls too, it's very dumb and just insulates the EU further (which is their plan it seems)
It doesn't apply to the card scheme, it applies to interchange, i.e. the part that goes back to the card issuer.
I don't see how that is dumb, the EU cannot do anything about issuers outside of it, and PSPs have to reflect the huge difference in interchange fees one way or another.
Correct, merchants do not charge interchange, banks/networks do, and its dumb because it increases the costs of doing business in the EU (which as I said, makes it more insular)
Fees are not the only cost of business, compliance is too - and things like if you want to enjoy lower intra-EU costs then you need to be in the EU which is an added cost as a non-EU business, etc - it's not as simple as "fees more eu bad"
You're kind of agreeing at this point, despite what the EU believe they are not self sufficient and require international business - the only people who lose out by that not happening are citizens and there are already enough reasons doing business (or even buying) stuff from outside the EU is a tedious and expensive chore
That is the case. Stripe will be charging the American businesses with American customers 2.9% + 30¢, and their European businesses with European customers 1.5% + 25¢.
Everyone is increasing prices in Europe. Inflation in the UK is crazy. I'm standing in my local coffee place and a basic Americano went up £0.50 overnight
It's crazy and a lot of gouging is happening (not so much these smaller companies. They do have it tough. The owner does drive an expensive BMW though)
> premium cards [are] commercial, corporate, or business cards issued by Visa and Mastercard [and attract more expensive fees, even compared to Amex]
Does anyone know more about that? Why are they more expensive? Is there a different underlying cost structure, or simply price discrimination by some intermediary?
There's more information on "premium cards" here[0], but it doesn't explain the price difference or why a separate category is needed.
About 85% of payment fees are taken by the bank that issues the credit cards (this portion is called interchange). The remaining 15% is split between the card network and the merchant acquirer (in this case Stripe).
Stripe collects some amount of of money from their merchant customers for each transaction. They could set their take rate to be whatever they’d like it to be, this is and unregulated part of the market and is the reason you hear about crazy 10%+ fees for porn, gambling, etc. websites.
Stripe then sends a subset of that fee (interchange plus a network fee) to Visa. Visa has a public ratebook saying exactly how much Stripe needs to send them for a ton of different kinds of transactions.
Visa keeps their fee and passes the interchange to the issuing bank and the transaction is settled.
Issuing banks feel like they “earn” interchange by acquiring customers and taking on their credit risk.
Now it’s important to remember that until very recently, Visa was wholly owned by the banks. This is because Visa is basically a “don’t shoot the messenger” actor who acts on the behalf of banks while taking a relatively small part of the pie for themselves.
Now, on to premium cards. One of the things banks have started to do recently is say “Hey, we acquire really good customers and we give them airline miles so they use their cards way more than they would otherwise. We should be compensated for that!”
The same thing happens for corporate cards, except worse since most regulations on interchange target consumer cards.
What that looks like in practice is new classes of Visa cards (Google: Visa Infinite) that have higher interchange rates (because the banks do all the hard work of having rich customers). Because Visa is a “neutral party,” they can get away with “if you want to accept any Visa cards, you have to accept them all.”
Therefore, cash/debit consumers continue to subsidize lavish vacations and corporate spend for the wealthy, because of course they do.
Because they give more benefits (insurance, lounge access, cashback etc...) and someone has to pay for this (i.e. the merchants).
For example R̶e̶v̶o̶l̶u̶t̶ ̶M̶e̶t̶a̶l̶ premium cards and corporate cards which offers benefits such as lounge access to pay-to-use lounges, travel insurance, cashback etc.. are mostly designed by the schemes in parternship with the issuing banks.
Also note that mostly in the US and Europe AMEX works in another way than Visa and MasterCard. Where V and MC are in a 4 party model (Issuing Bank that gets interchange, Processor (like Stripe) that processes the payment and gets a cut, Merchant that pays the merchant discount rate (or processing fee) and the customer that got their card from their issuing bank getting some benefits) Amex mostly cuts out the issuing bank and issues the cards to the customers directly.
Hence because they don't have to share the interchange with the banks they offer way better benefits on their cards (airmiles, centurion lounges, travel benefits, extensive insurance) but also at a high cost for the merchant because these are really premium cards.
Some consumer banks do issue AMEX cards so they will get their interchange.
In Asia there's also Amex Debit cards, I think definitely less popular in the rest of the world.
Something to note, also a lot of people in Europe think Visa or MasterCard == a Credit Card, this is not true. A credit card is an actual revolving credit instrument issued to the consumer by a credit institution such as a bank that then uses Visa or MasterCard rails.
A card that you get from Lloyds or HSBC when you open an account in the UK might say Visa or MasterCard but it's on a prepaid or debit programme with the scheme but using Visa and MasterCard payment rails. The interchange on these (pre-brexit) and still in Europe is capped by law.
Furthermore a lot of EU, Latam and Asian countries have their local payment rails which run at way lower costs than Visa and MasterCard to the merchants (Wechat, Alipay, Giropay, Payconiq etc.. to name a few).
One last thing, I do think Open Banking in the UK and Europe, especially when Variable Recurring Payments are coming will be a game changer. The banks have been lobbying to delay most of the UX improvements for the users because they just make too much money as an issuer to get interchange.
> The banks have been lobbying to delay most of the UX improvements for the users because they just make too much money as an issuer to get interchange.
To quote Hanlon’s razor “never attribute to malice that which is adequately explained by stupidity."
Having been involved in these conversations, I can tell you with some confidence, that banks are far more worried about being unable to deliver the technical work on time, than they’re worrying about interchange.
There will be fees to pay for OpenBanking Variable Recurring Payments, it’s not going to be free for merchants to use. Which makes sense, contrary to popular opinion, running a bank account, and transacting, costs quite a bit of money. Every Faster Payment you send costs your bank a couple of pennies, not much on an individual level, but it sure adds up.
Having been in some of the same conversations, although I'm a big believer in Hanlon's razor, if someone suggests that for an OpenBanking payment the "authentication" should be the bank calling the user and having them listend to a 45 second message to then key in a confirmation code it's no longer stupidity, it's creating artificial barriers. The same bank can perfectly do quicker ACS with OTP or tap to confirm using a push notification but for OB PISP it requires a phone call...
You should see some of the fraud cases that happen via open banking and faster payments.
It’s an extremely difficult problem to deal with because there is no dispute mechanism in the faster payment flow, and Faster Payments are not interested in adding one. As a consequence, once the payment is authorise, the moneys gone, and getting back is almost impossible.
Fraudsters have been taking huge advantage of this, and regulators are demanding that banks protect customers from these scams, or eat the cost of reimbursement.
Having worked on this specific problem, I can say that calling the customer for authentication is one of the most effective ways to prevent these scams. The call normally forces the customer to hang up on the scammer, which is incredibly powerful because it removes the primary source of pressure on the customer, and gives them space to think. Most customers then realise they’re being scammed, and stop there.
You may say that there’s other ways of warning customers, like in app notifications etc. Well I’ve tired all that, they’re not effective, regardless of how big, scary and red you make them. Even when tuned so the false positive rate is almost zero, so most customers only every see them when they’re actually about to be scammed, they still don’t work. Reason why they’re not effective, customer mentions what they see to the scammer on the phone, scammer explains it away and pressures the customer to continue.
This isn’t to say there isn’t some better balance, or that the banks aren’t being lazy. But the fraud question is serious one, and a very difficult one to answer. Simply ignoring it when discussing Open Banking is either naive, or intellectually dishonest (I’m not claiming that you’re intellectually dishonest, but there are plenty of people who like to gloss over the fraud issues, or just victim blame).
I'm very happy that the banks take fraud seriously, even if it delays innovation because the current chargeback flow is horribly broken for everyone involved except the schemes.
I'm not saying Revolut Metal is one of these, I'm just saying the card schemes create different segments for cards that have their own benefits, for these benefits the cost is actually mostly going to either the user paying a membership fee or the MDR paid by the merchant to cover the cost. The benefits are not just a freebie given by the issuer or scheme, it's a marketint cost that someone has to pay for.
re Open Banking, it's still a system where everyone but the actual customer gets useful access to financial services (ie; I can't get an API from a bank as the account holder, but I can give API access to a third party) - unfortunately that isn't mandated as part of the system and for me makes it completely useless.
Corporate cards, or cards with more "features" (lounge access, insurance etc.) tend to have yearly fees associated with them, at least in Australia. I was under the impression the customer covers the cost of these extras, not merchants.
The Revolut Metal (personal, not revolut business) example might not be the best one (see also my other reply) however the reason I mention is still the same, someone has to pay for the benefits these premium cards offer, the interchange on these is higher.
You attribute it to perks, which is not correct, or Amex and Diners Club would be more expensive. It's because interchange isn't regulated for commercial cards.
These fees are usually waived based on spend and often just require a simple call to the issuing bank. Even 400 usd won’t cover hundreds of thousands of airmiles gained by spend.
Commercial interchange isn't regulated in Europe. The card networks have increased their interchange rates for commercial and business cards in many European countries.
> or why a separate category is needed
These fees have been moved into a separate category to contain increases to cards where network costs have increased significantly. (Rather than apply them more broadly.)
Assuming this is related to Stripe trying to decrease the effect of the next funding devaluation. [0]
Hard to imagine they are still in a cashflow burn despite being around for so long and seemingly capturing a large marketshare of payments. Perhaps its primarily for insiders to take money off the table before an IPO lockup, with a not-so-great 6 month post IPO forecast?
the beginning of the end of stripe. Instead of attracting more customers and building the ground for a healthy startup ecosystem, stripe chose the shortsighted 'let's just charge more'. this immediately makes other alternatives like adyen more attractive...bad move and I don't get how patrick collison who roams around here let this happen. the 2.9%+0.25 was already allowing for healthy margins, and was higher than competitors.
Square is as well. I think there are several players which is good because if you design your payment systems to support multiple integrations you can switch between them without a ton of friction (source: last company switched a few times based on the terms they negotiated).
Its crazy there isn't more competition. I was hoping that would be one thing cryptos would be good for.
One thing I think regulators could do is force cc fees to be added to the price instead of hidden. Eg a $10 item costs $10, $10.20, $10.40 depending on which payment method you use.
> I was hoping that would be one thing cryptos would be good for.
It already is. While still not many, there are companies that will pass the savings realized from avoiding traditional payment processors fees by giving you discounts if you pay them in crypto (e.g. mullvad).
> One thing I think regulators could do is force cc fees to be added to the price instead of hidden.
I agree. Increasing competition by nullifying this part of merchant pgw agreements (standard in most agreements) would definitely be better than the status quo. Another option is to simply impose limits on cc fees and ban non-compliant processors but I believe your method is better. Give the customer the information and power to choose.
I thought they would be benefitting from everyone else raising their prices. Isn't this double dipping?
Edit: and isn't it quite likely it's easy for companies just to pass this charge straight on to customers, given the climate? Maybe that was Stripe's intention with the timing
I know this is in the EU, but do yourself and the businesses you love a favor. Pay with CA$H. The value you receive from local small businesses is worth the time it takes to avoid the tax on our economy.
I dunno, the 25 cents surcharge payment processors take on my €10+ payment isn't really worth that much of a hassle to me. Ads have encouraged me to pay by card for years.
I don't know why Stripe is taking a percentage on top of the normal processing fee, I'm guessing it's to deal with percentage differences for people paying using credit cards. There are many cheaper alternatives out there if Stripe is getting too expensive.
Interesting, I did not know that internet payments are so expensive. In Poland (UE) max interchange is 0,3%. Blue Media (Polish payment gate) cost 1,19%. 1.9%-3,25% + €0.25 looks expensive.
Well as it so happens I'm only in the first day or so of integrating Stripe in to my subscription-based side project in the EEA. What should I use instead? Molliepay?
Since these charges are from the gross, it all depends on your profit margins. If they're >50%, sure, it's not a big thing. If it's 5%, that means that 0.5% extra on Stripe eats 10% of your profit.
So Stripe still has some edge vs PayPal in all but non-EU cards. Not sure if they offer a microtransaction optimized fee schedule, if not, PP is far better there.
«Dispute fees (also known as chargebacks) will increase from €15 to €20. We’ll also no longer refund this fee if the customer’s bank resolves the dispute in your favor, due to the costs Stripe incurs for managing dispute evidence submissions (regardless of the outcome).»
That’s very interesting. Stripe is so greedy and dishonest.
How can Europeans use Stripe when they can't use Google Fonts and Google Analytics on the grounds that sending even a single IP packet to a server under the control of a US company violates the GDPR?
One aspect of this is that data processing which is necessary to do what the customer explicitly requested is permitted (GDPR article 6.1.b). Making the payment for your purchase is such a thing and this processing is lawful, but doing analytics or tracking who viewed the site through fonts is not, and generally requires explicit opt-in consent.
Is that how it works? I thought as long as the user gives explicit consent to that happening it's fine. Most checkouts I have seen have a checkbox you need to check before being redirected to stripe/paypal and the likes which asks for explicit consent to the privacy policy rules.
It's not even slightly how it works, and your interpretation is correct, so long as the customer has consented then you're fine. The issue with Google Fonts and Analytics is that the customer has very rarely consented to tracking (particularly with fonts), and in a great many cases even where there is a "I consent to tracking cookies" dialog the cookies were dropped before it was displayed.
With Google Font decisions have been about consent. With Google Analytics it's been about the transfer of data to a US org when a US court can issue a FISA letter. Basically, a privacy lawyer from the Netherlands has been hammering Google in court repeatedly.
I find them to be good indicators of how bad a website/service is. The harder it is to dismiss the dialog while rejecting everything, the worse the service tends to be in other metrics. I think of those dialogs as big bright banners which advertise loud and clear how disrespectful a website is of its users. Noticing the pattern means I waste considerably less time on websites which aren’t worth it.
Me too, I wish the companies would obey the law without putting in the consent banner. It's not like there's any law obliging them to put in the consent banners. They freely chose to piss us all off even though they had the alternative of behaving like perfectly profitable pre-internet businesses and ... not tracking us.
Yes, I can't believe we've created a system where those are the norm. But we also have real world cities plastered with billboards, so it's not like everything is clean, perfect and rational in our worlds.
The issue is not related to consent but to "safeguards".
GDPR requires that user data is guarded in certain ways, and European courts have ruled that no data can be sufficiently safeguarded if it is under control of a US company.
Google "analytics gdpr safeguards" or see here for example:
The issue is, giving a US org data is considered a transfer of data to that the US since a US court can force a US org to turn over the data no matter where in the world it is located. US data laws are not compliant with GDPR. It's basically just down to the spy laws they have, for FISA letters I believe, without those they would be fine.
No. Technically it's illegal to provide such service, not receive it.
Fun fact, this very website is not GDPR compliant. I've never ever seen a cookie notice on HN, which is legally required.
Of course, none of this is relevant, because AFAIK HN has no physical presence in the EU, but still, this site is not GDPR compliant, which just outlines how stupid GDPR actually is.
Cookie notices are required if you for example use them to pass data to third parties.
A login cookie that is just for auth is not that. It is specifically requested by the user and implies that the user's data is managed on that site and can be evicted.
Same with analytics. The problem is not that you're doing analytics. It's that the user doesn't know that your doing it and that you're passing on that data to a third party.
Would have assumed that anonymous logging is fine? Generating usage data? Error logs? As long as it’s not sold, private data or used to identify the user?
Highly doubt that kind of gathering is a problem. If it were you could close 95% of the web.
As I said, PII. So if your only PII is storing the IP one-way hashed and salted, keeping the salt only for a day as like (I think) Plausible does, it’s probably/possibly okay.
Besides that, intent also matters. For example, we had to start logging IPs for every newsletter change, or you can log IPs in your access log for security reasons without consent. Logging the same IP into your analytics tool becomes an issue.
> I've never ever seen a cookie notice on HN, which is legally required.
For clarity: it isn't always required. Only if you have third party (tracking) cookies.
I don't know about HN. But its perfectly possible to have analytics, ads and other functional cookies, without pestering your users with cookie popups.
Again: cookie popups and concent-banners aren't required. They are only required if you have "invasive" tracking in place.
e.g. I've worked on web-apps that were tracked by a selfhosted matomo, by plausible or some other tracking, that did not have any GTM or other tag-managers, that had no ads or only ads which were served from their own domain and without any 3rd party trackers, lacked all the GAFAM-pixels, had their fonts and other assets self-hosted (or on a simple, non-tracking CDN) and so on. Non of these needed any form of banner, popup or wizard.
I don't think you are right. Right now, the only cookie present is my session cookie. This should fall under obvious reason to have a cookie (it's not a tracking cookie). The sets of required and of all cookies are identical. Therefore, a dialog like this changes nothing.
> I've never ever seen a cookie notice on HN, which is legally required.
What for? The GDPR does not require consent for purely functional cookies, and the only data I see stored is my account cookie. So the existing privacy policy should cover them.
It's illegal for you to provide a service that isn't compliant. So if you use Stripe, therefore transfer data to a non-compliant country, you would be illegally providing a service that isn't complaint.
Also, the cookie notice is not GDPR it's a separate law. And if you only use functional cookies such as login cookies no notice is required.
IIUC governments already introduce price caps for certain elements of this market but I think we could be doing more? Like mandating incumbents to implement open standards and subsidising their competitors?
Dunno though I'm not very informed about the topic, just suddenly discovered recently how much of my income is just funding a fairly small number of payment providers.