Quick Takes
Happy Friday, Fintech Takers!
I hope your week has been wonderful.
Mine has been a bit frustrating. I was unwillingly assigned the role of Walter Fielding in my own version of The Money Pit; broken appliances, stuck garage doors, snow tires rolling this way and that.
Not life or death, thankfully, but annoying nonetheless.
It even seeped over into my professional life.
Yesterday, we attempted to host a virtual event on how lenders are managing in this moment of extreme uncertainty. We were all set for a great, highly interactive session when the system we use to host events unexpectedly malfunctioned.
For anyone who was impacted, my sincere apologies. We are actively working on rescheduling the event and will reach out to you with the updated day and time as soon as we have it!
- Alex
Quick TakesCopy anchor linkCopied
It was an overwhelming week and a lot happened, so we’re going to keep today’s newsletter simple.
Three quick takes on three fintech news stories.
#1: Fair LendingCopy anchor linkCopied
What happened?
The CFPB finalized a new rule:
The Trump administration’s CFPB finalized a rule rolling back fair lending enforcement by eliminating a tool used to weed out unintended bias and making it harder to bring claims that lenders disczouraged borrowers based on race, gender, or other characteristics.
The Consumer Financial Protection Bureau will no longer use disparate impact—statistical analyses for identifying discriminatory lending practices—under the Equal Credit Opportunity Act, the agency said in its final rule posted Tuesday.
Lenders will face discrimination claims only in instances where there is evidence, such as emails and other documents, showing intentional bias, according to the rule.
So what?
OK, first, I feel the need to point out how absurd it is to have a federal agency tasked with protecting financial services customers that only bestirs itself to write a rule when either A.) the crypto lobby gets mad about something (see open banking) or B.) it discovers something within its purview that uses the terms “diversity,” “equity,” or “inclusion” … or synonyms of those terms, like “fair.”
Here’s Acting Director Vought:
One Hallmark of the second Trump administration has been the eradication of discriminatory raced based policies that have permeated every aspect of government under the banner of ‘diversity equity and inclusion’ … The administration of fair lending laws is no exception.
I just … I can’t … I mean, what?
Disparate impact isn’t a discriminatory race based policy!
Here’s how disparate impact testing actually works:
- Lenders and their regulators analyze lending outcomes across protected classes — groups that fair lending laws protect from discrimination, such as race, sex, age, national origin, and religion — looking at who were approved or denied, pricing and terms offered, loan amounts extended, and product steering.
- If there is a statistically significant difference in outcomes that a lender's underwriting policies produce for a protected class compared to a control group, the lender must demonstrate that the policy causing the disparity serves a legitimate business necessity (typically tied to creditworthiness, safety and soundness, or regulatory requirements) and that the policy is narrowly tailored to that purpose.
- Even where business necessity is established, the lender (and regulators) must then search for a less discriminatory alternative (LDA), which is a different policy, model, or underwriting approach that would achieve substantially the same business objective with a smaller disparate impact.
- The lender implements the chosen policy (or its less discriminatory alternative), documents the analysis and rationale, and continues ongoing monitoring.
I can’t stress this point enough: disparate impact testing is a good thing!
It acknowledges the reality that, today, credit decisions are mostly made by machines and that bias can and frequently does creep, unintentionally, into the decision making processes that those machines use. This bias is very difficult to detect when constructing those processes because bias rarely shows up as a single obvious variable. It hides in correlations — ZIP codes that track with race or employment patterns that track with sex and age — and gets distributed across hundreds of model features in ways that are individually defensible but collectively harmful. You can build a model with the best intentions, exclude every protected characteristic, and still produce a system that systematically disadvantages the groups fair lending law was written to protect.
The only reliable way to catch it is to test outcomes after the fact.
And when lenders and regulators catch a problem, after the fact, the default approach is to search for a less discriminatory alternative for achieving the same business objectives that the lender is pursuing. That’s a great remediation process that fairly balances lenders’ business goals with the goal that we all have (the Trump Administration included … a white Christian man falls into three different protected classes!) to ensure equal access to fairly-priced loans.
I feel like I’ve been asking these questions a lot over the last 18 months, but here it goes again: Why are we doing this? Who asked for this?!?
#2: Creator Card Copy anchor linkCopied
What happened?
Visa is partnering with TikTok:
Visa has joined forces with TikTok to launch a debit card specifically designed for content creators in the UK.
During a TikTok Live, a creator is rewarded through virtual gifts which can then be converted into "diamonds," which in turn can be exchanged for real income.
This income typically arrives in bursts, rather than as a regular monthly salary, meaning short delays between earnings settling in their account which can make it harder to cover everyday costs or reinvest in their business, says Visa.
According to a survey carried out by the payments giant of creators on multiple social media platforms, 86% of creator-run businesses are self-funded and 49% experience late payments, highlighting the ongoing cash flow challenges they face.
The new Creator Card gives users access to tools designed to better separate business and personal finances. The card and business account enable users to access their earnings faster and spend money straight away.
So what?
The product that Visa and TikTok have built here seems pretty straightforward. The payout process for creators through TikTok is quite long (mostly due to fraud and compliance risk reasons), and by pushing money directly to a debit card, Visa and TikTok can shorten the process significantly. From a backend functionality perspective, my guess is that Visa and/or the unnamed issuing bank are advancing the funds and taking on some of that fraud/compliance risk (possibly with TikTok sharing some of it) while the payments are processed.
What’s more interesting to me is just how complex TikTok’s process is for viewers paying creators during their TikTok livestreams. Believe it or not, this is how it works:
- Viewers buy “Coins” with real money. Roughly 1 coin = $0.015, with small bulk discounts. Coins exist only inside TikTok.
- Viewers spend Coins on gifts during live streams. Each gift has a fixed coin cost; a “Rose” is 1 coin, a “Galaxy” is 1,000, a “Lion” is 29,999, a “Universe” is 34,999. Viewers tap to send them during a creator's livestream.
- TikTok converts gifts into “Diamonds” for the creator, which is where the platform takes its cut. The take rate is roughly 50%, though it may be more when processing fees are factored in. Each diamond is worth about $0.005 when cashed out.
- Creators withdraw Diamonds as cash. Usually to PayPal or a linked bank account. The minimum withdrawal is usually around $10 worth of diamonds, though some regions set it higher (around $100).
My goodness, that is complicated!
And there are some valid reasons for it. The biggest one is that TikTok doesn’t want to be regulated as a money transmitter. If TikTok let viewers send real dollars directly to creators, it would trigger an enormous compliance burden: money transmitter licenses in every U.S. state, FinCEN registration, and KYC/AML on every transaction. Coins let TikTok recharacterize the whole thing. Viewers aren't sending money to creators, they're buying virtual goods from TikTok (the same maneuver behind Twitch's Bits and every mobile game's gem economy). Coins also help TikTok manage the Apple tax, since Apple takes a 30% cut on in-app digital goods purchases, bundling payment into a coin purchase means Apple takes its cut once, and everything downstream is "spending virtual goods," which Apple doesn't touch. And then there's the psychology: virtual currencies consistently drive higher spending than real money. Spending 1,000 coins on a Galaxy feels like playing a game; spending $15 to tip a streamer feels like a financial decision (and a bad one at that, when you factor in TikTok’s 50% take rate).
All of this makes me think back to the Biden-era CFPB and its brief fascination with video games. I mocked it at the time, but the more I’ve thought about it, the more I’ve come to agree that these digital goods and currencies ecosystems have become more complicated and banking-like than old people like me fully appreciate.
They probably do deserve more scrutiny.
#3: PerpsCopy anchor linkCopied
What happened?
Kalshi and Polymarket are showing us who they are:
Kalshi and Polymarket are getting closer to launching new products that will take them deeper into crypto trading with so-called perpetual futures.
Kalshi plans to launch the product, tied to the prices of cryptocurrencies, in the coming weeks, according to a person familiar with the plans who was not authorized to discuss non-public information. The products will be Kalshi’s first foray beyond the event contracts it offers on its prediction markets.
Shortly after the first news reports on Kalshi’s plans, Polymarket announced on social media that it too is launching perpetuals.
“We price the future,” the company wrote in a social media post on Tuesday. “Now you can lever it.”
So what?
“We price the future. Now you can lever it.”
I’m sorry. Give me a second.
[Slams head against desk a few times]
OK, I’m back.
Let me be very clear: The last thing that the average prediction market user needs is leverage.
In case you don’t know, a perp (perpetual future) is a bet on the price of something (usually crypto) that never expires. You put down a small amount of money to control a much bigger position, hold it as long as you want, and close it whenever. Because the contract never settles, the platform charges a small fee every few hours (a "funding rate") to keep the perp's price tethered to the real asset.
Put simply, it's a leveraged, indefinite price bet with a meter running against you the entire time you hold it. The consequence, as with all leveraged bets, is that it can move against you quickly. A 10% move against a 10x position wipes you out.
However, the prize for the prediction markets is enormous. Last year the global trading volume of perpetual futures on crypto exchanges was estimated at more than $80 trillion, nearly all of it offshore, and the CFTC has signaled it wants to pull that volume onshore under its oversight. Today, Kalshi and Polymarket make money on prediction-market volume that spikes around elections and sporting events and craters between them. Perps offer something their current business can't: continuous, 24/7 volume in any market condition. And the economics are structurally better. Instead of earning thin spreads on discrete events that resolve and disappear, perps generate funding-rate fees every few hours, liquidation fees when leveraged positions blow up, and trading fees from users who churn positions dozens of times a day.
The public policy argument for prediction markets is that they are tools that aggregate information and tell us something true about the world. That civic-goods framing (which I wrote more about here) is how they earned their regulatory latitude and their billion-dollar valuations.
This move to offer perps completely undercuts that argument. Perps aggregate no information; a leveraged bet on Bitcoin's price just tracks the spot market that already exists. This is raw directional speculation, and the business model depends on users losing: platforms earn on funding rates, liquidation fees, and the churn of users who trade constantly, which means they have every incentive to help people trade more, hold longer, and lever higher, not less.
I already knew that prediction markets were just unregulated casinos. This is simply more proof.
MORE QUESTIONS TO PONDER TOGETHER
Big news for the endlessly curious (yes, you): I’m collecting your fintech questions on a rolling basis.
What’s keeping you up at night? What great mysteries in financial services beg to be unraveled? Think of it this way, if a stranger is a friend you just haven't met yet, your question is a Fintech Takes conversation waiting to happen.
One that could headline a Friday newsletter or be answered in an upcoming Fintech Office Hours event.
Drop your question here, whenever inspiration strikes!
WHERE I'LL BE
✈️ After Hours: AI in Banking | April 29 | NYCCopy anchor linkCopied
Kiah Haslett and I are going to be hosting an event with the marvelous folks at Team8 next week! It’s about AI and how it is changing banks’ build, buy, and partner decisions. Great topic. Great venue. Space is limited, but let me know if you’re interested!
🏀 Fintech Takes 3v3 Classic | April 30 | NYCCopy anchor linkCopied
There’s also going to be basketball at New York Fintech Week! Come hoop with us!
🏔️ Fintech Frontier Summit | May 31 - June 3 | Superior, MTCopy anchor linkCopied
Talk about a great topic and a great venue. We’ll be talking about bank - fintech partnerships at a ranch in the mountains in northern Montana. In June. Fuck yes. Space is very limited, but let me know if you apply, and I’ll put in a good word!
✈️ Open Banker Salon | June 5 | Washington, D.C.Copy anchor linkCopied
OK, I’m cheating a bit on this one. I won’t be attending this event (too close to my son’s birthday) but I really wish I was. John, Ashwin, Casey, and the rest of the Open Banker team have done a great job pulling together a compelling group of speakers and topics, and the format (a salon!) sounds exactly like what we need more of in D.C.
💻 *Bonus: The Marketing Channel 1 in 4 Americans Walk Past Every Week (May 13th, live — with Grocery TV's Marlow Nickell, The Marketing Millennials' Daniel Murray, & me) Copy anchor linkCopied
Fintech brands don't have branches. And I think physical retail is a more interesting answer to that asymmetry than most realize. Join us May 13th to get into how.
*This rec is brought to you by one of our fantastic brand partners.
Thanks for the read! Let me know what you thought by replying back to this email.
— Alex
Join 5,000+ other finance and fintech leaders in the Fintech Takes Network
The conversation doesn't have to stop here
Keep learning and connecting in the Fintech Takes Network
EVENTS | FEED | LIBRARY | DIRECTORY
