• Fintech Brainfood
  • Posts
  • Fintech 🧠 Food - 25th April 2021 - Apple Card Family, Stripe Issuing & What brands need to know before embedding finance

Fintech 🧠 Food - 25th April 2021 - Apple Card Family, Stripe Issuing & What brands need to know before embedding finance

Hey everyone 👋, thanks for coming back for more Brainfood where I take the biggest events of the week and try to get under the skin of what's going on. If you're reading this and haven't signed up, join the 5,512 others by clicking below, and to the regular readers, thank you 🙏

Weekly Rant 📣

What brands need to know before embedding finance

Wharton Fintech invited me to a panel this week on embedded finance, which came packed with insights including:

  • Regulation: Providers can help with regulation, but you're always going to have compliance responsibilities to some degree.

  • Regulation: When you start, you will be a target for criminals and fraud (especially as an entrepreneur / young business)

  • Regulation: Deeply understanding regulation becomes a significant competitive advantage for fintech products.

  • Lending: If you're going to be lending, banks can be your advocate with the regulator.

  • Lending: If you're going to be lending using AI, you're going to need a lot of data.

  • Lending: The debit card focussed propositions tend to focus on a community or problem (e.g., get paid early or underrepresented communities). The emerging credit card propositions seem much more focused on cashback, rewards, and status.

Massive shout out to Dan Henry, CEO of Greenlight, Dan Kimerling of Deciens Capital, Giles Cade, CEO of Cross River Bank, and the ludicrously talented Cokie Hastosis Lasanga Consulting.

There are two ideas I wanted to unpack here. Regulation and lending, both big themes as we push into the later stages of 2021, where everything is regtech and embedded lending.

Understanding regulation as your competitive advantage 📈

The insight that made Stripe so successful for me was the invention of the payment facilitator (PayFac) category. Stripe is essentially one big merchant and makes its customers "sub-merchants." By being the merchant, Stripe productized all of the compliance requirements of being a merchant online. This meant that anyone who wanted to accept payments online had guard rails in place by the APIs.  The simplicity of the Stripe APIs was not only a feature for time to market and ease of use; it was productizing compliance.  

In the embedded finance landscape, the use of regulatory understanding as a competitive advantage plays out a couple of ways.

Firstly, smaller banks can advocate for brands or entrepreneurs to help regulators understand something new or different.  Giles at Cross River gave the example of helping the regulators understand using AI for lending.

Secondly, new BaaS providers are increasingly productizing compliance in exciting ways. The BaaS providers like Synapse, Bond, and Unit create guard rails for what initial financial products can do. They also work directly with entrepreneurs and the underlying bank partners to help solve compliance and engineering problems.  

Thirdly, the new wave of regtech companies (like Alloy, Comply Advantage, and Hummingbird) simplifies and automates the challenge of running a compliance function.  Many banks had teams of people, software from the 90s, paper processes, and committees; by contrast, these new providers have streamlined KYC / AML and the SAR process dramatically.

If you want to embed finance, you need to understand the interplay of these three concepts.  The modern regtech companies provide incredible tools, but your team has to use them. The BaaS providers can help you do innovative things and dramatically improve time to market, but your team and brand are the front end to the customer. The banks can be your advocate to the regulator, but if you fuck up, they'll also have to shut you down.

Lending is amazing, but lending is hard. ⚠

In recent weeks I've covered several startups doing innovative things with credit models. Tomocredit stands out for its $100 advance credit line and credit builder. It works on a 7-day cycle, where the user has to pay back in 7 days. Tomo reports this to the credit bureaux if the user pays back and will gradually increase their credit line.  By compressing the time to understand who is creditworthy and who isn't, Tomo has increased the speed at which it can collect data to build its lending models.

Some months ago, I also covered this excellent blog by Affirm and how they used data to build their credit model.  

In lending, if your credit model is going to be unique, your competitive advantage is quickly getting to a lot of data.  The keys to success are.

Finding ways to compress the learning time for AI.  Tomocredit realized its users typically get paid weekly, so paying back weekly made sense. By advancing $100, they can write off those who don't pay back as a marketing expense.  (This is much harder with mortgages, for instance, but to an entrepreneur, that's an exciting challenge)

Working with lots of data.  Affirm used a significant amount of data to prove you could be more accurate at modeling repayment behavior by simply looking at the current state of repayments. Models would try to look at payment behavior historically (e.g., did someone previously pay in 30 days, but now they paid in 60 days). Affirm's conclusion was the only thing that mattered was, of all the repayments ever, how did payers behave at 30 days. This new approach proved to be more accurate than the traditional model over time.

Learning from people who have done lending before.  The hard-fought experience won by people who have run lending programs is beneficial no matter how good you are at engineering. Affirm had a baseline of traditional lending models before they introduced their new AI model. Balancing what has worked historically with the new models will be critical.  

When you're lending, you have so many requirements to manage, like how do you know your model isn't discriminatory to minorities? 

Entrepreneurs build to solve problems and often don't build to manage the unintended consequences.   

There is excellent support out there and lessons from those who have been there before. Mixing that with entrepreneurial spark is the key to building the products of tomorrow.


4 Fintech companies 💸

1. Knox Financial - The platform for landlords

  • Knox helps people identify valuable properties to rent, become a landlord and manage their property over time.  Knox brings together cash flow management and bill payments with services like tax preparation and tenant management.

  • Landlords are a high-value niche.  Becoming a landlord is essentially taking on a new part-time job, and there are very few tools that help to do it well.  I like how broad Knox is thinking. (Although their website is ugly compared to, say Hammock in the UK).

2. Boom Pay - Build credit as you pay your rent

  • BoomPay allows users to pay rent however they choose (directly from their bank or debit card), even if their landlord prefers a check. BoomPay also enables users to change their payment date (to align with when they get paid, for example).  Lastly, all payments are reported to the credit bureaus to help users build credit from paying rent on time.

  • Lower-income consumers can be stuck in a cycle of renting and find it hard to build their credit score without any credit history.  By taking an activity, they already had to do and reporting it, you could see how BoomPay could eventually help its users become homeowners.

3. Finalis - The platform for M&A / Private Placement dealmaking

  • Finalis allows bankers or deal makers to build a deal from compliance-ready templates, manage the deal with collaboration tools and manage the various files and data room challenges that come with making a deal.  Finalis is FINRA affiliated, has state registration, and an in-house compliance team that can quickly review any deal for users of its platform.

  • Putting together a merger, acquisition, or alternative investment can be incredibly time-consuming, manual, and exhausting.  Much of the work is managing the emails, PDFs, and spreadsheets flying back and forth between lawyers, compliance teams, and counterparties.  

4. Qapita - Carta for APAC

  • Qapita allows entrepreneurs to manage their cap table, issue stock and options to employees, and model investment scenarios.  Qapita also offers services to help entrepreneurs structure their deals with more hands-on advice.

  • The old adage is that raising for your startup is a full-time job. Tools like Carta have gained tremendous popularity with investors and entrepreneurs in Europe and the US.  Qapita is bringing those tools to Singapore (a hub of startup activity for the region).

Things to know 👀

  • Using Apple Card, families can share the Apple Card and "build credit together." Users can track, share and manage their spending. Two adults can use the card and extended to include any children (including spending limits and transaction blocking features). The card can help families (and children) build a credit score together.

  • 🤔 My Analysis: There's a lot to like in the Apple Card family.  It is an example of beautifully executed multiplayer fintech, solving family use cases with Apple quality design. Generally, I'm also a fan of products that create on-ramps to being credit-worthy.

  • 🤔 My Analysis: The unintended consequences here could be massive.  If a parent doesn't set their teenagers spending limits right, a teenage child could ruin the credit score of the whole family.  I imagine (and hope) someone at Apple is thinking about this, but damn. Remember, underneath this is a bank (Goldman) managing the risk appetite and reporting to the credit bureaux.

  • 🤔 My Analysis: This is a real shot at Greenlight and other teenager cards that monetize via a monthly fee. Apple has a real opportunity to displace that type of card, especially in the affluent segment.

  • Stripe issuing allows users to create, manage and distribute virtual and physical payment cards.  Stripe gives Zipcar an example that offers its drivers cards that drivers can only use to buy petrol as a key use case.

  • 🤔 My AnalysisBanking as a Service has been very regional for a long time; by expanding internationally, finally this is starting to change.  

  • 🤔 My Analysis: Stripe in Europe was nowhere near the same behemoth in the US, with Adyen and Checkout.com both enjoying significant market share. What Stripe has going for it is now a breadth of offering and increasingly global availability.

  • 🤔 My Analysis: Another player to watch for geographic spread is Railsbank.  Railsbank gets a tiny fraction of the press Stripe does, and is frankly, much earlier but has significant geographic coverage and is generally loved by developers. BaaS is going to get interesting.

Quick hits: 

  • UK Government announces access to the central bank balance sheet for fintech. Access to the payment systems and balance sheets puts many fintech companies on the same level as banks. This could be massive in time.  The treasury is also announcing a CBDC task force. What do you do when you don't know what to do? Announce a task force. China is live with a pilot; we need a little less conversation and a little more action in CBDC.

  • Razor pay raises $160m- Razorpay feels a bit like India's Stripe connect + stripe capital + modern treasury. Embedding payment operations with embedded lending & payment acceptance is a combo I'm not sure I have seen before, and I like it.  Razorpay feels like an API first upgrade to Global Transaction Banking.

Good Reads 📚

  • Wherever the creator goes, the audience and attention follow. Platforms are bootstrapping by attracting those creators.  The big platforms (Spotify, Tiktok, etc.) offer significant money to attract the top creators, but this only works for the very top creators.

  • Mirror is an example of a creator-focused protocol (not platform), where committees make product decisions, and ownership is shared between creators and developers.  

  • This sounds like one of those crypto utopia ideas that would never work in practice, but it is.  From Uniswap, Aave, and Mirror itself, creators can make a significant income from this model. The protocols are incentivized to build tools that best serve their community to keep the community.

  • 🤔 My Analysis: Mirror is an example of a service that uses multiple crypto protocols.   Mirror uses Arweave to store essays and blog post content, ENS to create a decentralized domain name linked to a CashApp handle.   Not everything will be built on Eth, different protocols will play different roles, and the best services will combine those protocols for different use cases.

  • 🤔 My Analysis: Read this. The sheer volume of ideas and creativity in a single essay is staggering.

  • Wells Fargo published better than expected earning results as the economy begins to right itself.  For the first time, we have experienced a recession without a credit cycle (restriction of access to credit for the economy, AKA, banks are lending a lot less).

  • Deposits are way up, making bank balance sheets strong, possibly too strong. Consumer deposits are up 62%, making banks less profitable as loan demand is unusually low. Given banks make money from the spread between loans and deposits; this is a bad thing. The measure of profitability (Net Interest Margin - AKA "NIM") for JP Morgan is down from 2.37% a year ago to 1.69% in Q1 2021.

  • The problem for banks is made worse because they have to lodge significant capital against deposits after the financial crisis. Except for Wells Fargo, which has been forced not to grow its balance sheet or attract deposits by regulators after its miss-selling scandal.  In a twist of fate, the ruling designed to harm Wells Fargo helped them.

  • 🤔 My Analysis: When you look at this, you can see why the very largest banks moan about the regulations they're subject to post financial crisis. In this weird market, the biggest risk isn't that borrowers will default. The most significant threat facing banks is their business model doesn't work in today's economy.  Maybe it's time for regulation to catch up with the market. Perhaps it's also time for banks to find new ways to make money.

Tweets of the week 🕊

That's all, folks. 👋