site banner

Friday Fun Thread for January 9, 2026

Be advised: this thread is not for serious in-depth discussion of weighty topics (we have a link for that), this thread is not for anything Culture War related. This thread is for Fun. You got jokes? Share 'em. You got silly questions? Ask 'em.

2
Jump in the discussion.

No email address required.

I think I know what's going on here. This quote - and the data - comes from paper dating from 2013. And indeed, if you look at Figure 3b in my link, that was the case up to about 2017. When it changed, and rewards expenses started to exceed transaction income, and have exceeded it since. This also matches my own experience - a while ago, 2%+ no fee cashback cards either did not exist or were a rarity that required a lot of hoops to jump through. Now they are commonplace. As you can see in the graph, the rewards expenses went from ~3.4% in 2013 to about 4.5% in 2022, while the transaction margins decreased.

The article discusses (and refutes) the idea that rewards beneficiaries are "rich" and interest payers are "poor", but neither I nor thread-starter made such claim (it's not the fault of the article, obviously). In fact, both categories may be rich, or poor, it's irrelevant - the discussion about whether tx margins or interest is the main source of revenue does not require any specific income distribution among either category.

The article says:

I'll say, as an aside, this paper dates back to 2014, but it's the best available in field, in my personal humble opinion, and I don't believe the conclusions of it have materially changed.

Given what I have seen in my link, I must question this opinion and claim that while the conclusions of the article may have been warranted given the data from 2013-2014, the situation did materially change. At least a claim from the Fed to that effect strongly indicates it did, and one needs much more than an offhand "I believe" to counter that. Maybe the conclusions of the article - which differ from the initial claim - are still warranted, but I do not think that the old data in the article supports what you purport it to support anymore.