• Fintech Brainfood
  • Posts
  • Fintech 🧠 Food - Nubank, Klarna, CBDC regulation and why Hyper-personalization is the zenith of banking.

Fintech 🧠 Food - Nubank, Klarna, CBDC regulation and why Hyper-personalization is the zenith of banking.

Hey everyone 👋, thanks for coming back to Brainfood, where I take the week's biggest events and try to get under the skin of what's happening in fintech. If you're reading this and haven't signed up, join the 6,315 others by clicking below, and to the regular readers, thank you. 🙏

Could stablecoins be Africa’s new savings vehicle? On the 11:FS Blog, Gwera takes a look at what’s happening with Defi in Africa, and when you look at currency devaluation, and the nature of cooperative finance in Africa, things get interesting.

Weekly Rant 📣

Hyper-personalisation could be more than a buzzword.

Historically only banks were allowed to create financial products, but the embedded finance revolution has opened that up to

  1. Non-finance Brands 

  2. Finance Brands

Non-finance brands, especially Vertical SaaS platforms, embed finance at the point of need. Shopify gives you lending and banking where you manage your online store; accounting platforms are bringing together your payments and banking. Even platforms to run gyms like Mindbody are embedding finance to add more value for their users and capture more revenue.

Finance brands are solving consumer and business problems building on new Payments processors and Banking-as-a-Service providers.  Whether it's the movement to "get paid early" started by Chime or the broader earned wage access we see from companies like Branch, consumer innovation has been astonishing. The increased time to market and reduced cost to launch has also led to the digital community banks like Daylight, Greenwood, and Cheese that solve real social problems for their communities.  

Entrepreneurs and builders are doing amazing things within the constraints of the financial system. But there are limits.

There are limits to what's possible to create because everything is ultimately constrained by how financial products are manufactured, distributed, and configured. 

The classic example here is the Apple Card. While Apple has created a stunning product with strong cashback incentives for its users, it is ultimately limited by the risk appetite of the underlying bank (in this case, Goldman Sachs). No matter how modern the Goldman Sachs systems are, they are predicated on an old-fashioned model of how products get manufactured, distributed, and configured.  

(I'm picking on Goldman here, but this could be Citi, JPMC, or even one of the smaller partner banks like Evolve, Meta, or Bancorp).

There are also two other potential financial product creators.

  1. Creators and artists

  2. The consumer / SMB customer

Imagine what artists or creators could do if we removed the shackles of how products are manufactured and configured? My good friend (and fintech nerd) Will White had a great story about the zenith of banking. Beyonce bank.

Imagine if, during a concert, Beyonce could give everyone a 50% instant rebate on merchandise, but only in that venue and only while the show is happening.  Creator-led hyper-personalization could unlock some fascinating experiments only limited by their imagination.

And what about consumers and businesses themselves?

Imagine if you could give each customer the ability to have their own "product sliders," maybe one customer cares more about savings rate and is willing to take a lower credit limit to get that. Perhaps another customer wants their rewards in Bitcoin? Rather than launch new products for each customer, why can't the customer change that setting?  Customer-led hyper-personalization could be even more exciting than creator-led.

To do that, you'd have to really re-think how products are manufactured, distributed, and configured.

How did products get manufactured historically?

Financial products have a special legal status; they tend to be highly regulated, meaning that only regulated financial institutions can create those products.  (Note, "financial institution" is a broader term that could mean banks, but also regulated lenders, brokers, insurance companies, and more).

This means no matter who you're buying the financial product from, ultimately, there's a regulated entity somewhere behind the scenes who has manufactured some form of financial product that you're now using.

When you're buying the Apple Card, you're subject to the risk policies of the underlying bank. When you're using a BaaS provider, you're subject to the capabilities offered by their underlying banks.

When you look under the hood at a bank, technology real estate, at the very core, is their "core banking platform" and likely a general ledger. This ledger records the balance of each customer. So Alice has $10 and then makes a $5 payment; now Alice has $5.

These core banking platforms have welded together several capabilities.

  1. The recording of balances

  2. The movement of money (over rails like ACH, Wire, Cards)

  3. Regulatory reporting

  4. The creation and configuration of financial products

This fourth point is the kicker.  Because banks have a mental model of what a financial product is and the limits of risk appetite, so does their software.

Banks define financial products and the parameters by historic norms. The definition of a deposit account hasn't really changed for decades—the same for a loan, the same for mortgages. These financial products tend to be very fixed in time and space.

Thankfully BaaS providers have abstracted a lot of the pain of the underlying bank. But BaaS is still limited by how the financial product itself gets manufactured, distributed, and configured.

Why does this limit BaaS?

Because everything sits on the regulated financial institution, everything ultimately gets limited by the financial institution. If you wanted to create a financial product instantly for an esoteric use case, that would have to fit inside the bank's risk appetite.  

BaaS providers have done a great job productizing one or many banks' risk appetites to reduce the cost and time to market. Still, the BaaS provider could never own that risk unless they became the bank (e.g., Solaris or Griffin). 

But what about modern core banking software?

If the bank just upgraded their core banking software to something more modern, could we have hyper-personalization? If you believe the marketing, then yes, we would. But here's the kicker part-deux. Even the most current core banking platforms share the bank's definition of what a financial product is.

Modern core banking systems have three approaches to hyper-personalization.

  1. CRM-driven

  2. Data-driven

  3. Declarative (e.g. Smart Contracts)

The CRM-driven model is to layer a modern customer relationship management system over the top of the core. This CRM comes pre-packaged with products, workflows, and no-code tools for business users to edit and manage products. These solutions are for many banks an upgrade, but still, a) Only the bank itself can define the products and b) they rely on the decades-old model of what a financial product is.

The data-driven model says to feed everything into a data-lake, use the best modern machine learning to learn about customers, and present the best real-time, contextual product offers. This is certainly an upgrade for many banks on where they are today, but it's also not new. Capital One has done this well for decades. It doesn't solve the actual creation of the financial product, at low cost, within risk appetite.  So againa) Only the bank itself can define the products and b) they rely on the decades-old model of what a financial product is (although this is somewhat enhanced).

The declarative approach is interesting. With a smart contract, in theory, you can create any financial product as a financial institution (provided you know Python). But it is limited to the financial institution itself and requires significant skill (and imagination). And once again…a) Only the bank itself can define the products and b) they rely on the decades-old model of what a financial product is (although this is somewhat enhanced).

Implementing any of these approaches and just better APIs would make the life of BaaS providers and fintech companies so much better as a first step. A bank with great APIs has a huge competitive advantage.

But the real opportunity is to empower brands, creators, and even the customer to hyper-personalize and configure their own products. And that means we have to fundamentally re-think the architecture.

So you're telling me we can't have nice things?

Well, maybe we could if we changed where we start from. The definition of financial products. If we change that, we change how we architect the software.

The core insight is that banks are basically state machines. To make everything configurable in a bank, you need to model as much as possible of that bank as the state machine: every vendor, every system, every customer touchpoint, all of it.  

Modular or configurable banking has existed for more than a decade, but so much of this begins at the traditional mental model of what a financial product is today.  Therefore it enables its customers (the banks) to construct financial products the old way, and then layer on personalization later.

Instead, if you start with the insight that banks are state machines, the interesting question becomes, what is the most efficient way to see as much as possible of what's happening in that bank and construct financial products from what you see? The answer is rarely "replace the core."

Back to Beyonce for a second...

To create that one-off experience 

  • What tool would Beyonce's team need? 

  • How would they configure that product? 

  • How would the bank ensure it is compliant?

  • How do you make that economical when the product is thrown away?

You need a bank that thinks differently about how financial products are manufactured, distributed, and configured. You need a different way of thinking about a financial product and the architecture that gets you there.


4 Fintech Companies 💸

1. Kikoff - The credit builder app  

  • Kikoff users sign up to receive a one-time $500 credit limit spent within the app.  Items in their store start at $10, and Kikoff reports every payment a user makes to the credit bureaux. The initial credit account product is free, but Kikoff also allows users to pay for an installment product and a credit card at $9 per month.

  • Only being able to buy products from within the app's own store is clever. Despite lending $500 to someone without a credit score, they know exactly how the money got spent.  Kikoff has an interesting wedge product, and users may move up over time into the higher revenue-generating products. Neat.

2. Retrypay - Payments Orchestration for e-commerce

  • Retrypay aggregates many payment processors (e.g., Paypal, Adyen, and Stripe), allowing e-commerce store owners to use a single platform to manage them all. This can help with global reach, accessing more payment types, and resilience (e.g., if one payment processor rejects a payment, retry).  Retrypay also allows the e-commerce store to offer benefits like "let your users pay locally" to avoid international payment fees.

  • Payment Orchestration is a massive trend in the past few months, but the proposition is compelling.  The alternative to using a payments orchestrator is to build this capability internally.It will be interesting to see if the big players considered "processors" like Stripe or Paypal acquire some of this orchestration capability. It could get weird but interesting in the next few years at the checkout.

3. Snowball Wealth - Student Loan Planning

  • Snowball wealth aggregates student loans and provides insights like how long it will take to pay down loans and how much it will cost.  Snowball users can then plan ways to save money monthly or pay down their loans sooner.  Snowball makes money through referral fees when users re-finance their loans.

  • I really like the angle of fixing student loans as the beginning of building wealth.  Snowball potentially has a unique wedge product into either re-selling other financial products or becoming a savings & investments platform over time.

4. Kibanda TopUp - Food Supply Chain Management for Africas restaurants

  • With Kibanda, restaurant owners can order staples like grains, rice, or beef via the app or SMS and receive them the next day.  Half of the Africas population eats from informal restaurants daily, and during COVID-19 food supply has been fragile. 

  • By digitizing the ordering and payment of goods, Kibanda can begin to grow these informal restaurants into businesses. In Malaysia, we've seen simple apps like Bukukas move from simple stock management to payments to payroll and ultimately towards banking. Kibanda could do similar things.

Oh hello, enjoying Brainfood? Why not go ahead and hit subscribe, then it will conveniently land in your inbox every week :)

Things to know 👀

1. Fintech is 🔥 as Nubank raises $750m and Klarna $640m

  • Nubank, the Brazilian digital bank with 40m customers, has raised $750m to fuel growth in Brazil and Mexico.  $500m of the new investment is from Berkshire Hathaway, Warren Buffet's famous "value investing" vehicle.  

  • 🤔 My Analysis: The Berkshire investment is interesting for two reasons. 1) It's a recognition that digital-only challengers have arrived and represent "value." 2) The Brazilian market represents a massive fintech opportunity. There are dynamics to LATAM banking that make this make sense (vs. say, Chime). Interest rates are higher, and incumbent competition had been so much weaker historically.  Still, you have to hand it to Nubank; the execution has been flawless.

  • Klarna, the European BNPL provider, has raised fresh funding at a $40bn valuation from investors, including Softbank. The funding comes amid "massive growth" in the US market, with more than 18m users in the past year.  Klarna may use the new funding to fuel an acquisition spree as it heads towards IPO.

  • 🤔 My Analysis: BNPL is rocket fuel.  In Q1 earnings, PayPal saw 70% of users making a repeat transaction in 6 months (h/t JohnStCapital). It is a win for the merchant (increased conversion), and consumers love it.  Klarna, Afterpay, Affirm, and now PayPal are in a foot race for expansion and growth. Meanwhile, banks are trying to launch "installment credit cards," lol.

  • 🤔 My Analysis: Klarna and Nubank now bigger my market cap than Barclays, Burger King, or Delta Airlines.

  • China's e-yuan, its central bank digital currency, has now been trialed with over 4 million transactions and $314m spent so far.  But according to a survey of more than 3,000 Chinese by the economist, there are several barriers to its adoption.  

  • Issued raised include technical literacy, not trusting the security of the technology, and concerns about privacy (!). However, by far the most significant issue raised was the quality of the product.  Alipay and Tencent have such a massive utility and user experience lead it may be impossible for the top-down, government-sponsored product to gain traction.

  • 🤔 My Analysis: You can build it, and they won't come if the product is shit.  This is a lesson for some central banks, who seem sure that if they build a CBDC, it will outcompete private stablecoins and prevent the risk of the e-yuan dominating.

  • 🤔 My Analysis: Beijing sees its e-yuan reducing the reliance on the tech giants and the US-centric dollar system. The ECB similarly, is looking to launch its digital wallet and a digital Euro. The end of the dollar dominance creates a foot race for the new global reserve. It’s going to be an interesting couple of decades.

  • 🤔 My Analysis: Contrast this approach with the Bank of England paper on stablecoins TL;DR, mass adoption of stablecoins might create better payments but increase the cost of credit (and make commercial bank cost of capital even higher, sucks to be a bank). The Bank of England proposes to regulate algorithmic stablecoins (e.g., MakerDAO), somewhat like banks themselves (with capital requirements, but perhaps less stringent).  This document's posture is that stablecoins are inevitable, so what are the trade-offs as we think about ways to regulateThe exec summary is worth a read. BOE does nuance so well.

  • 🤔 My Analysis: The trade-off for central banks appears to be either create a complex, multi-layered currency system that can be embraced by the public and private sector. Or, create a complex and multi-layered policy framework that embraces many types of stablecoins.

Quick hits

Good Reads 📚

  • JSC takes us through a brief history of finance and why financial incumbents are so inefficient.  Fintech 1.0 came along and made the user experience much better (e.g., Square, Chime). Fintech 2.0 is the embedded finance movement (as new BaaS providers abstract the pain of the infrastructure and lead to every company becoming a fintech company).

  • Fintech 3.0 will be the re-architecture and rebuilding of the finance infrastructure itself, led by Defi.  Defi will make the infrastructure itself composable, programmable, transparent, and accessible. Already stablecoins can be sent anywhere globally, instantly, by anyone with compatible software, and Defi lending protocols like Aave create actual peer-to-peer lending.

  • 🤔 My Analysis: TL;DR, Fintech 1.0 = Better UX. Fintech 2.0 = Embedded Finance. Fintech 3.0 = Defi.

  • 🤔 My Analysis: I love the terminology Fintech 3.0 because I massively agree, Defi is the logical evolution of Fintech, but it's also 5 years (at least) from the mainstream.

  • 🤔 My Analysis: I really like the simplicity of the Jesse Walden mental models for Defi and NFTs: Defi = money legos, NFT = media legos. Another good way to think about Defi. If Robinhood built some cool feature, there’s no way Wealthsimple could use it. But in Defi, if Uniswap builds something, and can all be copied and even remixed (and that’s exactly what happened and led to the creation of Sushiswap).

  • 🤔 My Analysis: There is definitely a weird tribal "fintech vs. Defi" thing happening on Twitter that I ranted about a few weeks ago. I'd encourage skeptics to take a read of the JSC piece; it's balanced and well-argued.

Other great reads I didn't have time to do justice to

Tweets of the week 🕊

That's all, folks. 👋

Remember, if you're enjoying this content, please do tell all your fintech friends to check it out and hit the subscribe button :)