Credit card rewards function as a tax on those without credit cards and those with credit cards but without the ability or inclination to keep up with the panoply of options. When I use my American Express card to buy a train ticket, the train company donates money (via a circuitous route) to American Express and American Express rewards me with British Airways points. Everyone who buys their ticket with a debit card is subsidising my next vacation. I wonder if their bots will be so generous?
Credit Card Rewards Reward Who?
There are many observers who are uncomfortable with the way that payment cards work in practice. There is, for example, the criticism that card issuers explicitly charge the rest of the economy for the work involved in recruiting and rewarding the most desirable customers. The Commerce Commission of New Zealand recently published a consultation paper on the costs to businesses and consumers of card payments which makes this point. They summarise the situation clearly: where there is no surcharging, consumers who choose to pay through lower cost card payment options (eg, PIN debit or cash), subsidise other card users through higher retail prices and where there is a blanket surcharge (which is typically around 2% in New Zealand), debit card users subsidise credit card users.
This subsidy is significant. Sumit Agarwal of the National University of Singapore, Andrea Presbitero of the IMF, and André Silva and Carlo Wix of the Federal Reserve find that American credit card reward programmes redistribute around $15bn a year from the less to the more “sophisticated” consumers. Now, as it happens, financial sophistication (which the researchers approximate with credit ratings) also correlates with education, income and race. Which means that high-school graduates, the poor and ethnic minorities are the least likely to earn credit-card rewards.
Even though I might be classified as a sophisticated consumer, I am not sure that I do my best to harvest rewards from the unsophisticated. It just takes too much time and too much brainpower to work out on a per-transaction basis whether I should use the card the gets me reward points, or the cashback card or the card that get me frequent flyer miles when I am standing at the check out. For one thing, working out the “foreign exchange” rates for Points and Pounds and Avios means having access to real-time data and continuous monitoring of redemption.
(I hate it when you get one of those e-mails offering frequent flyer miles or hotel points for buying a case of wine or whatever. It can easily take half an hour to work whether a “special offer” to buy points for cash makes sense and it can easily take a day to work our whether bank account A without miles is a better deal thank bank B with miles.)
Now, however, we are shifting from a world of AI agents (who suggest things for you to do) to a world of agentic AIs (who act of your behalf). Imagine what business will look like in the new age of “agentic commerce”, an age where AI is deputized to make just this kind of decision of behalf of a consumer. Want to earn rewards on a purchase? Let your agent decide how do pay! Payments are so boring! This is just the kind of tedious “work” that most people will be only too happy to hand over to bot. Smart shopping assistants are not new, but they being transformed by AI. What will happen to loyalty and rewards programmes when consumers always make the right choice to optimise their rewards, always get to use their points in the best way and so?
If you think this is a marginal issue, by the way, let me remind you just how important these reward schemes are, using frequent flyer programmes as the example. The three big US schemes – American’s AAdvantage, Delta’s SkyMiles and United’s MileagePlus – generated almost $16 billion of revenue in 2022, equivalent to 11% of their parent companies’ total revenue.These schemes are great for the airlines because the points cost them nothing until they are redeemed, and often cost them nothing at all because the points are lost, forgotten or expired. The Financial Times found that Wall Street lenders valued the major airlines’ frequent flyer programs more highly than the airlines themselves.
Does the value of these schemes go up or down when agents always make the best decision on behalf of consumers. I genuinely do not know, so if you have done the calculations please do share the results!
How Will Credit Card Rewards Change?
Is obvious that with new payment methods come new fraud vectors, and there are many issues that must be resolved before consumers can really take advantage of the new technology to ditch the mundane and free their minds for more value-adding distractions. There are also serious issues of liability that will need to be resolved. If you didn’t mean to pay for something, is that the agent’s responsibility, the merchant’s, your bank’s, or yours? There are a great many security issues to deal with as well. Can someone trick your agent into making fraudulent purchases, and if they do, who is on the hook? There are also authorisation questions to answer: How can agents sit in authentication flows like liveness verification, two-factor authentication and so on?
Actually, I think that we already have all of the cryptographic techniques necessary to address these issues so we can start to think about how the sector will be reshaped by the arrival of agents. These agents could help consumers in maximizing their credit card rewards and reducing costs, which might in turn pressure issuers to offer more competitive rewards programs. What’s more, as consumers become more adept at leveraging those more competitive reward schemes this could lead to more lucrative rewards programs, driving up issuers costs. If the agents help consumers to avoid fees (and interest payments) while rewards costs continue to increase, this will mean business models have to change business models, maybe focusing more on transaction fees charged to merchants or refining tiered or premium services.
As far as I can tell, most card issuers have no strategy for consumers becoming an order of magnitude smarter overnight. This is going to be an interesting year.
Read the full article here
