• Fintech Brainfood
  • Posts
  • Fintech 🧠 Food - 7th Feb 2020 - PayPal results, Balance B2B, Valon & Divvy Homes 🔥 and the fintech stack

Fintech 🧠 Food - 7th Feb 2020 - PayPal results, Balance B2B, Valon & Divvy Homes 🔥 and the fintech stack

Hey everyone 👋, thanks so much for coming back for more brainfood. A space to learn in public and hopefully process everything happening in fintech.

Thanks to the 114 people who signed up in the past week for this free newsletter.

If you haven't joined yet, you can subscribe right here 👇

We also have a ton more content over at the 11:FS Youtube

Weekly Rant 📣

As I wrote up "4 fintech's" this week, it became clear the fintech stack is re-making itself. Charley Ma (I believe) coined the term "fintech stack" as the default tools used by builders and operators in fintech. I like this term because it communicates a powerful idea, that in B2B fintech, there is a provider power law and that the community quickly adopts the market leader. 

There's a B2B fintech API for almost every part of the banking value chain and stack in consumer and now increasingly SMB and B2B payments. Over the past year, we've also seen increasingly lending products (upstart, lendflow) and more specialist APIs emerge (alloy, hummingbird).

To fully understand the fintech stack, I think we have to look through a few lenses.

  1. The types of customer

  2. The products

  3. The providers

  4. The geography

The customer 🧑

My mental model for the market is that fintech is 1% finished. The full video explainer is here. 👇

But it looks something like the picture below…

  • Higher volume, low value per customer products are closer to retail and prime segments that banks traditionally served.

  • To the left of this (even higher volume, low value per customer) is the underbanked sector. This sector sees significant disruption as fintech companies like Square and Current create aspirational brands with lower-cost technology and operating models than incumbent banks.  

  • To the right are larger value per customer but lower volume—small businesses, high net worth, corporates, financial institutions, and finally sovereigns.

Each point on this scale has different products to deal with different challenges (paying, getting paid, managing cash flow). The problems for a high net worth individual are often quite different for the underbanked. As a result, the financial products they buy and solutions they require change.  

Historically banks' "digital transformation" journeys are backward-looking. People look at what everyone else is doing today and build from there. Foresight is about understanding the deeper customer problem and the job to be done and building from there.

To effectively service directly to a target segment, any builder or business needs the *foresight* to see what's coming around the corner and what customers will want in 3 years.  

The products 🎁

There are many core product types in finance (e.g., Store value, move value, lend value). There are many secondary activities (Know who's interacting, what their risk is if the transaction is legitimate, audit, reporting). Around that are tertiary activities (managing balance sheet, re-selling loans, regulatory reporting).

Historically banks built vertical solutions to manage all of these activities. This created a mental model where customers buy products, and the technology and operational stack to support that product was unique and custom-built.

Increasingly the areas of that value chain that are horizontal are being unbundled (e.g., KYC, Fraud, Payments) by B2B API providers (covered next), but the underlying product engines are still very silo'ed.

There are newer, modern, composable product engines or great BaaS APIs, but their mental model is often based on yesterday's products. Creating a "smart" or composable loan is still creating a loan, requiring some vertical integration.

The market is building *primitives* for many of the nondifferentiated areas of building fintech and embedded finance, with horizontal, B2B API providers. (Although this is yet to come to the construction and orchestration of financial products properly).

The providers 🔌

The fintech stack is preferred by anyone building new fintech or embedded finance solutions because the APIs are clean, modern and the business model (often transactional or pay as you use) is low friction for builders.

Most banks struggle to partner with these fintechs.  The reality for big banks is they can probably invest in these companies far faster than they could operationally integrate them. For example, the UK startup Flux removes paper receipts and stores them automatically in your banking app. Flux has been live with most UK challenger banks for years but just secured its first major UK banking partner.

SaaS companies and fintechs need to identify and integrate many B2B API providers.  Many SaaS companies embedding finance lack the depth of knowledge to differentiate between the providers easily.  Fintech's and SaaS companies each have to build a significant amount of nondifferentiated code to handle all of these integrations (even with clean APIs from all involved).

So the choice is often between working with an all-in-one BaaS player, with less flexibility, or renting and assembling your stack. It's like the difference between buying the lego death star or having the bricks yourself to make whatever you like.

Increasingly the ability to rapidly identify who the best player in the fintech stack is and integrate with them will become a competitive advantage.

To effectively create fintech or embedded finance solutions, builders need *insight* into who the B2B fintech APIs are and their capabilities.

The Geo's 🌍

Have you ever wondered why "Plaid for X Geo" is a thing? Or why Apple Card isn't available outside of the USA yet?

That's because the underlying financial infrastructure in each Geo is wildly different and complex. The best provider for the USA is often not the best provider for Europe, and you can forget about APAC. The simple solution for this could be to wait until Plaid and Stripe acquire everything. But that's unlikely in the medium term.

Builders in fintech and embedded finance don't have a global solution. Even the very best scale API providers tend to be quite USA centric. So if you're a fintech, bank, or global SaaS platform, it's hard to scale globally.  

Where does that leave us? 🤔

If you were going to build a new fintech or embedded finance product tomorrow, you'd be well served by BaaS providers and B2B API providers who can get you moving quickly.  

But there's a catch. Well, several actually;

  • The best API provider changes over time, so you need to avoid vendor lock-in

  • Some BaaS providers offer credit and not debit.  

  • Some API providers only support certain product types or Geos.

  • There are almost no fintech B2B providers that you can truly compare apples to apples.

So the fintech stack is a helpful hack. 

But the fintech stack varies by Geo and changes over time.  So that leaves any builder with some key challenges.

  • How do you re-bundle all of the fintech APIs?

  • How could you change the fintech API provider without customer interruption over time?

  • How do you create consistent experiences across Geos?

You'd need 

  • Foresight into your customers' future product needs,

  • Composable primitives for those products 

  • Insight into the fintech stack providers, and how they vary by Geo.

Where do you get all of that?

4 Fintech* 🤑

*especially for dhaber

1. Balance - The first complete B2B e-commerce checkout

  • B2B payments are massive. Credit Suisse estimates it's a $95trn market globally (nearly double all consumer payments). Retail e-commerce has been the story of 2020, with businesses that enable consumer - merchant payments becoming massive growth engines. Shopify counts for nearly 46% of total revenues from payments.

  • B2B payments are also utterly broken. The majority flow through ACH and wire and are subject to a complexity that consumers don't have. There are multiple ways to pay (with an invoice, on delivery, or even recurring payments). This complexity is usually handled by a finance team or even the entrepreneurs using their accounting software directly in an early stage.

  • It's not a surprise that companies like Brex who's card-based solutions see massive traction.  Modern Treasury's customers love how they've simplified payments.  

  • But where both Brex and MT focus on the buyer experience, Balance is focused on the checkout experience.  As a merchant, marketplace, or SaaS business, how do you create a simple checkout experience that bakes in all invoicing, payment type and delivery options? That's what Balance is aiming at.

  • Built on Stipe and the US payments layer cake, Balance creates beautiful UX flows for B2B payments.  Lightspeed, Stripe, and Max Levchin's Sci-fi are early investors, and everything about this makes sense.

2. Persona - Twilio like APIs for ID+V, KYC, and AML

  • Persona provides a complete ID+V, KYC, and AML compliance platform that's API first.  Developers can drop persona into their app with a few lines of code, and operations teams can manage the various types of compliance risks behind the scenes.

  • Persona is an example of the fintech stack evolving.  It launched on product hunt five days ago but already counted companies like Stripe, Brex, Postmates, and Branch as its customers. The head of product at Square Capital credits persona with helping distribute their PPP loans.

  • We've had simple Identity and Verification (ID&V) startups for some time (e.g., Onfido, Jumio, Socure), and now there are much more comprehensive AML and compliance platforms like Alloy and Hummingbird. Persona combines both of these types of solution into one.  

3. Valon Mortgage - The digital mortgage servicer

  • Possibly the only thing more broken than B2B payments is mortgages and the house buying process.  The mortgage process has evolved over decades from a paper-based system, where each transaction had enough margin not to give incumbents much incentive for change. Mortgages have three main workflows. Origination is buying a mortgage and is usually handled by a lender (and/or broker to lender). Servicing (collecting monthly payments, or dealing with collections, missed payments). Transfer (selling a mortgage to another bank).

  • In the past few years, fully digital lending platforms like blend.com have done well to help banks that relied on branches bring their origination and servicing in house.  Valon is playing in a different space. Valon is a platform for servicing that works with many vendors, potentially (if more lenders sign up) creating a consistent experience for the consumer.  

  • In the same week, DivvyHomes raised $110m in Series C to help renters buy their homes. Divvy is focussing on creating new borrowers by building their savings. Increasingly the before and after the origination space has room for disruption.  How do you help people save, find the right area, find the right broker, prepare for moving day, removals company, lawyer, switch utilities? There's a vast set of problems around the mortgage still not solved.

4. Gatsby - Options trading app with social & rewards

  • Robinhood had quite the week. Despite some gloomy mood music, they raised $2.4bn (yes billion) and added around 600k new users. Democratizing access to finance is wildly popular. But Robinhood has a chasing pack that is differentiating on being social first (e.g., commonstock and public.com) and not offering more risky products such as options trading.

  • Gatsby is taking another approach, competing with Robinhood directly. It offers zero commission options trading but includes social features and rewards.  It will be interesting to see if Robinhood adopts these features or if consumers look to move in this direction.

  • There has to be a reasonable middle ground between pushing risky options products at retail and democratizing finance.  Instead, how could we gradient up the risk for a consumer based on their increasing available capital and long term behavior (e.g., via connecting to Plaid and seeing their affordability and creditworthiness). Robinhood CEO will have to testify before Congress. It seems the West Coast VC community has taken a "regulator bad, innovator good" approach to this, but the reality is much more nuanced.

I often find there's too much news. Fortunately, Nik from this week in fintech has your back.

Check out Nik Milanovic's This Week in Fintech for the headlines of the past week:

  • Ant Group to restructure as a financial holding company.

  • Square and AllTap both launch new POS systems.

  • India bans bitcoin and looks to launch its digital currency.

  • JP Morgan will launch a UK neobank.

  • Royal Bank of Canada and Bread team up on BNPL.

  • And mega-rounds for Robinhood ($3.4 billion), Stash ($125 million), Divvy Homes ($110 million), and more!

Things to know 👀

  • 73m accounts added, 31% total payment volume growth (TPV).  Jason Mikula points out that PayPal is growing users but only seeing a 1% increase in payments per account. More users, but not more DAU (although that's changing with new products). They shipped BNPL and Crypto this year.  BNPL has a 40% repeat use rate, and users with crypto log in 2x as often.

  • 🤔 My Analysis: PayPal is rising with the e-commerce tide and shipping new features at an incredible rate, but I worry if its endless game of catch up with Square is winnable. PayPal seems to be consistently a step behind Square, but perhaps that's their role in the market. By the time it hits PayPal, it's becoming mainstream. When PayPal does it, it's crossing the chasm.

  • Companies like Klarna and Affirm sit in a regulatory grey area in the UK.  BNPL providers don't extend credit to consumers, but consumers who miss payments can end up in a collections process, negatively impacting their credit score.  While the lender's risk (e.g., Affirm) is against the merchant, the UK regulator believed the UK consumer was not adequately protected.

  • 🤔 My Analysis: My real hope here is the regulation is nuanced. GDPR was well-intended but ultimately just made the internet a bit more annoying. The UK regulator tends to approach lending regulations with broad strokes like "let's make everyone display the APR (APY)" to inform consumers. You could imagine they'd also require a full credit assessment of the consumer before the transaction completes. This could break the experience of BNPL and not materially improve outcomes for those the regulator wants to protect.

  • 🤔 My Analysis: With open banking, we can check creditworthiness and affordability far better than historic credit bureaux credit checks.  I'd love to see better outcomes for consumers, but let's not assume the regulatory minimum is the right approach.

  • 🤔 My Analysis: I wonder if as GDPR made the internet harder to use for most of the world, the UK looking into BNPL means it is a matter of time until the Biden administration does the same?

Good reads 📚

  • Personal income increased in 2020, but only when you include stimulus and unemployment benefits. The underlying real income decreases as assets become more expensive, inflation is hidden, and demographics are trending against younger people.  We have higher populations in the old generations, whose healthcare needs are getting more expensive.

  • Nikhil also mentions an idea being credited to Brad Gerstner, to give every child a "birth dividend" of $2000 that they can't withdraw for decades.  The compounding could potentially allow them to pay for school or get started in property. We need more long term and structural policy responses than just stimulus checks.

  • 🤔 My Analysis: When I have this conversation with property owners with high income, they'll often point to low inflation levels and continued growth in earnings. But this is hiding a few things. When Gen X and boomers left school, their school fees were tiny compared to today's cohort, and buying property as a % of income was much more achievable.  There is a real difference between having a high wage and being able to afford to join the ownership class.

  • 🤔 My Analysis: When you become an owner, you begin to build wealth. Building wealth takes decades of compounding, and this is happening later and later in life.  Meanwhile, the planet is dying, the average politician is in their 70s, and the younger voters are outnumbered by older voters who are more concerned about their own healthcare. The aging society hides the problem when you look at the whole economy; we have to look at the cohorts.

  • If you want to understand what happened to Robinhood, forget order flow for a second and take a look at the DTCC.  The DTCC is the US securities market engine, processing $2.15 quadrillion worth of securities in 2019.  What most people don't realize is the same underlying asset can be bought and sold multiple times. This creates a problem two problems. 1) It's hard to track and reconcile all of those trades, and 2) paying the full price for something you may or may not buy in the future gets expensive quickly.

  • 🤔 My Analysis:Netting is probably the misunderstood superpower of financial markets.  With netting, you "pay the difference" between everything you owe others and everything they owe you. Imagine I had 2 transactions to pay you $1 each (=$2), and you had two transactions to pay me $2 each (=$4). If we "net settle," then we work out the difference and pay that (so you would pay me $4 - $2 = $2).

  • 🤔 My Analysis: Many of the hot-takes about payment order flow, or "Robinhood good, wall st bad," have missed the nuance of why netting exists.  Robinhood's margin call may have been massive ($3bn), but it also should not have been a surprise.  Brokers are expected to hold reserves aside to manage margin calls or throttle their trading activity. Instead, Robinhood was building up a massive order book, which carried a ton of risk and happened quickly.  

  • 🤔 My Analysis: Two things look likely from here. 1) Robinhood has to better build their collateral and cash reserves for retail-led market volatility (which is expected of brokers). 2) Regulators step in to limit retail exposure to options and futures contracts. But it makes you wonder, what would a decentralized DTCC look like?

Tweets of the week 🕊

That's all, folks. 👋