Discover more from Fintech Brain Food 🧠
5th June 2022 - Can Neobanks ever be profitable? Layoffs at Bolt, Stripe Marketplace & The importance of time to money
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 15,104 others by clicking below, and to the regular readers, thank you. 🙏
Hey Fintech Nerds 👋
It's been a sunny week in the UK with two public holidays and the Queen celebrating 70 years on the throne. I've enjoyed extra family time and bumping into so many of you getting into London the week before Money 2020 Amsterdam. Can we make this a thing?
Maybe every year before Money 2020 in Amsterdam, there's a fringe event series of nerds in London (or across Europe?) 🤔
The Fintech community really is that, a community. We support each other in good times and bad. I’ve felt this personally lately from so many of you ❤. Things aren't always easy, but when you're in an industry with so many nerds who get excited and what we can build together, that's as good as it gets. It has been a blessing meeting so many of you; now we can travel again.
I guess that's my way of saying thank you. To all of you, building, thinking, writing, investing, ranting, and just Fintech-loving nerds.
Lithic just published their legal library and free templates for card programs. That's the kind of pay it forward, good vibes stuff we love (and I know others do great things, Modern Treasury has a great learning center, and there are so many good resources, I'll never remember them all). But the point is.
We're simplifying the complex.
To build better finance.
And if finance underpins everything in the global economy. Changing finance changes the world.
Thanks for being you, Fintech Nerds 🤓
This week I've been thinking about can consumer Neobanks make money if interest rates come back? Is one-click checkout a feature or a product? And I'm obsessed with Ayo's latest blog on time to money.
Sometimes email clients clip the bottom of the content. If you don’t want to miss anything, click here for the website version.
Weekly Rant 📣
Most Consumer Neobanks lose Money. Is it game over?
Despite consumer Neobanks having more than 100 million accounts globally, less than 5% make a profit. Great UX and hook products have driven adoption, but few show signs of long-term sustainability. Even worse, those that have millions of users in the US and UK have spent vast sums getting a charter/license, which only increased their cost burn.
Whatever happened to money-making machines?
Is that it?
Should we all go home?
(Cue some banker saying, I told you so, they'll never be profitable)
We need to unpack:
Why Neobanks are chasing charters
The cost structure of Digital Banks and Neobanks
The revenue model and dynamics of Incumbent Banks require scale and breadth
The revenue model of Digital Banks and Neobanks had been narrow but is evolving
And Neobanks and Digital Banks have cost levers they can pull
So does a charter make sense?
Why get a charter?
Jason Mikula breaks down Varo's rationale here. TL;DR,
To begin lending against deposits and
Improve the take rate on the swipe fee interchange.
Lending is challenging if you're not a bank. You have to have the capital to first lend to customers, which has to come from somewhere. Typically this comes from specialist lenders (like Credigy) or other banks (like Cross River). As an originator of lending (i.e., the person selling the lending), you have all of the KYC and underwriting costs
Others like Monzo and Starling, when gaining their UK bank license, noted that for consumers having their deposits insured with FSCS (the equivalent to FDIC) would help consumers build trust and help them grow market share.
Costs for Incumbents vs. Digital Banks and Neobanks
Before we get into revenue, we need to talk about cost. Digital bank's primary economic advantage over incumbents is a much lower cost to acquire and manage a customer.
In my experience,*, incumbents will pay anywhere between $150 to $350 to acquire a customer (cost of acquisition or CAC). The lower figure is purely operational costs, and at the higher end, it might include a switching incentive like $100 for opening an account. These figures have changed somewhat as incumbents increasingly adopt digital onboarding platforms via mobile. But they still have to cover the cost of opening accounts in a physical branch with humans.
* These are finger-in-the-air numbers based on what I've seen, and there are countless edge cases and exceptions that they don't include.
For digital banks, the acquisition cost ranges from $5 to $15, with the top end tending to bake in all activities like sending a physical card.
The incumbent cost to serve a checking or current account again ranges from $150 to $300 per year, in my experience. This will include a lot of shared / operational overhead from branches, payment systems, and other shared infrastructure costs.
Neobanks have published figures ranging from $60 per year to $10 per year on an account basis. Some include operational overhead, some aren't, and a lot of this depends on their provider stack.
The same is true for chartered digital banks like Monzo, who would quote closer to $20 to $30 to run an account for a year.
In my experience, by the time a bank is big enough to get a charter, they're meaningfully displacing some of its more high-level providers (e.g., Banking-as-a-Service providers). Some even replace their Issuer Processor with in-house tech (e.g., Monzo, Starling, Nubank, and Current). They do this to improve their cost structure.
Digital banks have a meaningfully lower acquisition cost and cost to serve than incumbents, but they also don't do all of the same stuff in most cases.
And that means they can't always generate as much revenue either.
Revenues for Incumbents
Regulated consumer banks have two primary sources of income "net interest margin" (NIM) and fees. (Which is why the incredible Marc Rubenstein named his blog Net Interest, which I'm an avid reader of and you should be too)
Net interest margin is the difference between the interest banks must pay on deposits (e.g., 1%) and the revenue they make from lending (e.g., 5%). So revenue of 5% minus 1% paid to depositors = 4%.
Fees are everywhere, overdraft fees, payment fees, card issuing fees, and foreign exchange transaction fees. The largest banks also generate significant fees from asset management and advisory businesses (wealth and capital markets).
This summary from MX shows that interest income is typically around half of the large bank income and fees the other half.
But the important takeaway is how much of that revenue at the biggest banks comes from commercial, advisory, and asset management businesses. The big banks charge fees and are profitable per business line (like in consumer), but they're also a portfolio of customers and geographies. This is not true of Neobanks and digital banks.
I've often said winning at the banking business model reminds me of the movie Waterworld, where the polar ice caps have melted, and humans search for little high ground and land remains. Incumbent banks have struggled to sell more products in developed markets while the cost of infrastructure, capital, and compliance have all skyrocketed in the past decade.
The world's largest banks span consumer, corporate, and capital markets in 100s of Countries and 1000s of products. The economies of scale allow them to be profitable despite the burden of being a bank. It's much harder to disrupt that with one or two products only available to consumers and only in one market.
This chart from Oliver Wymann sums up the problem incumbent banks face perfectly (especially in EMEA and North America).
Loans have grown, but the net interest margin has not. And both are growing just about in line with GDP.
The bank's core business has stood still for a decade on a relative basis.
That's why banks have chased revenue growth through fees, and that is where Neobanks and digital banks challenged them.
It's also worth adding that bank regulatory, technology, and infrastructure costs have exploded in the past decade. Trying to fight that with ever more scale has worked to a point. But now they're much more seriously considering using Fintech companies as providers to reduce cost. We could be in for an exciting few years for incumbents if they invest the right way.
Incumbent banks can copy Neo and Digital Bank revenue models.
(Especially subscriptions, as Cornerstone pointed out in their research report Banking in 2022 - rebounding from the revenue recession).
Revenue sources for Digital Banks and Neobanks
Neobanks' and digital banks' primary revenue source has been fees from card usage (interchange), but some have branched out in various ways.
For Neobanks, interchange revenue is income every time one of their customers uses a card to pay. They may split this with their providers (the partner bank, the issuer processor, and possibly a Banking-as-a-Service provider). When interchange is high (above 1%), this can be almost enough to sustain a viable business, but rarely with wildly positive unit economics.
Interchange varies by card type and geography. It's higher for credit than debit. Higher for corporate than consumers and higher for smaller banks in the US than big banks (thank you, Durbin Amendment). It's higher in APAC than the Americas and higher in the Americas than Europe (although in the UK, travel cards now have high interchange again, which is why Monzo is having a renaissance, or a Monznaissance, if you will).
It is fair to say that Neobanks and digital banks initially bootstrapped their revenue engines with interchange and low-cost structures. The 2010 passing of the Durbin Amendment made this particularly profitable in the US. The introduction of PSD2 in Europe in September 2019 had the opposite effect, dramatically reducing revenue for European Neo and digital banks.
But the times are changing.
And we have seen several alternatives to interchange emerge like:
Subscriptions: The Neobanks and digital banks often have many qualify-of-life features (like subscription management, commission-free trading, cashback, purchase protection, or travel insurance). It's tempting to see this as the old "bundled account" product, but bundled accounts often included features customers didn't want. The subscription is a) way better positioning and b) focuses on what Gen Y and Z actually care about. This is another $5 to $15 per month revenue line.
Embedded Finance: Starling Bank in the UK is leaning into Embedded Finance in a market where the incumbent banks have been de-risking the money service business (Fintech) sector. This requires a charter/license to offer but is a way to diversify revenue and is taking the opportunity the incumbents leave on the table.
Real-time payments (US): As Ayo covers in his excellent "time to money" blog post (in good reads), delays in getting paid disproportionately impact vulnerable consumers and merchants. Not only can this be a competitive advantage, but it's also one worth paying for.
Stock trading: Everything from Cash App to Revolut has baked in stock buy/sell/hold capabilities to diversify their offering and defend market share vs. specialists like Robinhood. While the stock market is down, the rise of Investment-as-a-Service APIs like Drivewealth, Alpaca, and Embed makes offering this service much easier than it would have been in previous decades.
Crypto: In the bull run, Fintech companies that added Crypto saw a massive spike in user growth and new revenues. Yes, Robinhood, Cash App, and Revolut are the obvious examples, but now even Neobanks like Step (the account for teens) are thoughtfully adding Crypto. Like stock trading, the rise of Digital-assets-as-a-Service offerings like Zerohash and Conduit** has made this more accessible for Neobanks, digital banks, and incumbents.
"Tips" (US): Some Neobanks allow their users to offer "tips" or "donations" in return for a short-term lending facility. In theory, these are optional, but most apps are designed to make opting out of the tip complicated in practice. (I hope we see an end to this practice, it feels wrong somehow, I'm all for flat fee lending, but call it a fee).
Pay later: Several digital banks (and incumbents) offer users the ability to "pay later" or pay in installments for a given transaction. This is a popular form of managed lending, especially as it helps consumers manage cash flow.
Lending: Some digital banks have become effective at lending. Nubank is the poster child, deploying $3.1bn of loans in Q1 22' (up from $600m the year earlier).With more than $12bn in deposits, delivering $877m in revenue at a 34% gross margin. But digital banks like Tinkoff and Starling often don't get the plaudits they deserve for building effective and diversified lending across consumers and SMBs. It can be done even in lower interest rate markets like the UK.
**(Discl': I'm an investor)
Some thoughts on these business models 🤔
The frame on Neobanks and Digital Banks above has mainly been consumer, but the SMB sector is performing better. SMBs and larger corporations have higher average deposits and transaction volume. Corporate card interchange is also higher.
Since they're building a balance sheet, Starling getting into Embedded Finance is interesting to me. While the US has countless partner banks, they're typically small, and in Europe, we see specialist headless banks like Solaris and Griffin. I don't know if I've seen a mainstream consumer bank push embedded finance quite as strong as Starling is.
There are lessons here for incumbents and large Neobanks chasing a charter (or who just got one). Offering embedded finance requires some specialism, but once it works, you're effectively re-selling your balance sheet with a much lower distribution cost. Have someone else cover the CAC and take a part of the revenue.
Embedded Finance doesn't have enough firepower, especially for Crypto. There are so many companies looking for deposits or lending in Crypto, but the banks out there just don’t have the liquidity (balance sheet) to get it done.
The first bigger banks in each market to figure this out have a massive opportunity to attract deposits and deploy lending. (Europe is nowhere on this stuff, at least the US has Silvergate, Cross River, etc. LHV bank in Estonia is tiny).
Levers on cost, growth, and unit economics
As Starling and Nubank build their lending books, their operational and marginal costs are dramatically lower than incumbents. For example, Nubank quotes $0.73 per active customer per month (or $8.76). How on earth could an incumbent touch that? Lending is complex, and the multiples for those that succeed are often financial services multiples, but the revenue is real.
Where incumbents carry the branch infrastructure, they also have countless offices, data centers, old IT infrastructure, and stuff that's very hard to turn off. All of this tech and process debt makes taking cost out extremely challenging. The fact that the bigger banks are often roll-ups of 5, 10, or even 15 different banks historically mean system duplication. Conversely, countless business units might use a payments engine for different customer bases in different markets.
Incumbents live in a world of edge cases.
Incumbent's biggest cost lever remains branches and staff because the cost take out can be realized quickly. But trying to meaningfully reduce IT cost is a gordian knot while not having the estate fall over. The systems were never designed to be modular and built before modern software engineering practices evolved.
But perhaps their best hope is the Fintech provider ecosystem. Fintech providers grew massively, serving the consumer and SMB Fintech boom, but are now looking to expand. Incumbents need to get better at partnering and figure out where a Fintech company fits in its operations. There are easy wins (e.g., compliance process for partner banks with platforms like Themis), slightly more complicated things like SDKs inside the mobile app that reduce fraud, and then the tough like core banking replacement.
Neobanks and Digital Banks have the advantage of much more modern tech and process infrastructure, but they don't have the direct cost take-out as quickly either.
User growth at all costs is no longer an option, and unit economics matter.
Having focussed entirely on growth, getting compliance, fraud, and efficiency in the provider stack becomes critical. Digital Banks can't lose that growth and have to find ways to improve customer conversion while reducing costs.
Again the Fintech provider stack can be beneficial here, as specialists can essentially become your entire "department for x." As PackyM wrote in APIs all the way down, imagine if a single API allowed you to hire the Collisons to run the payments team. The same is true for Fraud, compliance, Crypto, Stock Trading, real-time payments, etc.
There may be other areas where the provider has to go. They worked to get to scale, but now it's time to go further down the stack or build internally to get those economics. This comes down to the Fintech companies' DNA and strengths. Are you great at customer? Great at engineering? Where do you win?
And where do you hire the world's best department of x via an API?
Do you really want a charter Neobanks?
Getting a charter is hard; keeping it is harder.
Getting scale and the flywheel is still harder unless you have amazing economics like Nubank. Nubank's home market is a cheat code, but so is its execution.
Getting a charter in the EMEA or North American markets makes it harder to achieve product velocity.
There's also seemingly a horrible rite of passage between being a Neobank and operating a charter at scale. Survive that. Then thrive.
If you get a charter, get great at lending, and consider embedded finance.
But maybe that's not the right outcome for every Neobank or consumer Fintech company. Some might be better off not getting a charter to focus on aggregating every new Fintech feature that appears.
Especially those who had focussed on low-income or niche consumer segments. A lower-income consumer is a lower deposit base, and a lower deposit base makes lending harder. Lending to low-income consumers is also fraught with risk to people who can't afford the high-cost debt.
I suspect we'll see all of the above continue to play out and evolve.
And now the incumbents are here to play, with more firepower.
I hope they start deploying that balance sheet as they are the CAPEX for "digital transformation."
And I hope Fintech companies and talent get through this rocky patch stronger than before. No doubt there will be casualties; finance touches everything.
And if we change finance, we change everything.
4 Fintech Companies 💸
1. Uprise - Free Wall St Grade Financial Plans
Uprise helps users optimize every financial decision and claims to help find $1.5M in additional net worth for users over their lifetime. It works by understanding user goals, optimizing investing, and tax planning. The app helps users figure out allocations of a pension vs. investment vs. managing debt, getting the most from employer benefits, and finding the credit card with rewards that will matter to you.
🤔 If Credit Karma and Charles Schwaab had a baby, this might be it. If it works, this is the app we've all been waiting for, the private banker in your pocket quietly optimizing shit while you catch up on Stranger Things. Uprise is monetizing referrals, which can be a good business at "Credit Karma scale," but like CK, I wonder if they'll be tempted to launch their own card in time? Also, every consumer megabank product division is trying to build this product but not shipping it. There's a limited window for them to acquire or partner in this private banker in your pocket app category. These apps can aggregate consumer demand and make money-saving apps (PFM) look like child play.
2. Bloom - Investing for Teens
Bloom is a stock-buying app aimed at high school-age consumers to buy US stocks and fractions of US stocks. It features a "learn to earn" set of quizzes where users are rewarded with shares (or fractions) for completion. Bloom works with non-profit financial literacy groups, is an SEC-registered advisor, and uses Alpaca as its technology broker.
🤔 Money is culture, and some of the financial advice on TikTok is better than some private bankers I've met. The one universal bit of financial advice that applies throughout history is to start saving early in life. Warren Buffet started saving at age 11 and is still doing so in his 90s; he has been much more able to compound with the stock market than most (who start at 30 and finish at 60). Bloom enables not just saving earlier but learning as you do. I'm seeing Alpaca quietly pop up in more places where I used to see Drivewealth.
3. Jupiter Exchange - Fractional ownership of real-world assets via NFT
Jupiter allows users to buy "ownership tokens" representing fractions of a curated, real-world asset collection (like sneakers, cards, art, or memorabilia). The founders say the use of NFTs makes the exchange element publicly available in a way that wasn't previously possible for consumers.
🤔 Countless apps already allow consumers to buy fractions of real-world assets. The value of an NFT and an exchange is that these apps could interoperate, or real-world asset tokens could start appearing in other apps (like your favorite Neobank). Now that everything bubble is popping, it will be interesting to see how the demand for alternative assets evolves. Pokemon cards and baseball cards have exploded in value in the past 10 years, but where is the longer-term equilibrium?
4. Cardda - Expense Management Card for LATAM
Cardda provides virtual cards for companies in LATAM and includes features like spending limits, single-use, and category restrictions. Cardda does not charge its users for its standard package and has no fees or minimums. The team has focussed the value proposition on international payments in particular (e.g., to suppliers like Facebook or Alibaba), where LATAM merchants historically have had issues completing payments with incumbents' offerings.
🤔 The virtual card for SMBs and expense management category continues to expand, as does the LATAM Fintech ecosystem. Based in Chile, I'm curious to see which markets this card is available for and what will make Cardda stand out over time. I also wonder if the likes of Brex (who have roots in Brazil) could be well placed for much more regional expansion.
Things to know 👀
Bolt will cut an estimated 250 staff, a source told Bloomberg. The payments startup noted this was a part of a "broader restructuring." In a company blog, the CEO stated the goal to secure their financial position so they can "secure their own destiny."
🤔 Can one-click checkout be a company, or is it a product? Famously Fast.com closed its doors with $600k ARR, but it had mainly focussed on smaller businesses. This bottoms-up, growth customer-led model worked well for Stripe historically, but Stripe also took good economics from every transaction. Competitors like Adyen did well in enterprise and came top-down into the market. Then, folks like Shopify have their Shop Pay product, which is aggressively capable. Why is this a business, not a feature?
🤔 Is Bolt another Fast? We haven't seen underlying revenues for Bolt yet, but both companies have very loud founders on Twitter. Ryan Breslow of Bolt, in particular, has claimed conspiracy theories about Stripe, Y Combinator, and Sequoia. But the data we do have shows there's some potential here. Bolt's last $11bn round was led by Blackrock and claims they have 860+ merchants (nearly 2x YoY), including companies like Forever 21 and Benefit. There's certainly a lot of noise, but Bolt's model of a shopping app for consumers and enterprise merchants is also not without merit.
🤔 What about Wyre? Bolt acquired Crypto on-ramp and payments startup Wyre for $1.5bn in April in an all-stock (or mostly stock) deal. Wyre is a key player in Crypto payments alongside MoonPay and Transak; they're one of the most popular ways to convert fiat into Crypto. If Bolt is in trouble short term, does that mean Wyre revenue would have propped it up? Has Wyre's revenue imploded with the Crypto market crash? I sense for Bolt that diversifying their sources of revenue allows them to become a payments company, not a checkout company. I think they'll be fine in the long run, but I'd rather have Wyre stock than Bolt stock (FWIW, I don't have shares in either or any of their competitors, NFA).
🤔 Before this market correction, many folks thought BNPL would be in the most trouble (and yes, Klarna has announced layoffs). But there's a difference between slowing growth and being an entirely unviable business. My sense is in 10 years, BNPL providers will still be around, but one-click checkout companies will either diversify, get acquired, or die.
2. Stripe just launched payment orchestration (kinda)
Stripe Marketplace allows users to integrate tools like Mailchimp, Xero, Docusign, and Dropbox to create and manage everyday business activities. The focus is more on an app marketplace than a pure workflow builder, but customers can use Stripe Events to trigger actions based on webhooks' conditions.
🤔 Cross this off the "what if stripe builds it" bingo card. That's it, that's the thought.
🤔 Although there's still a lot of work for users to do here. The payment automation and orchestration companies specialize in several verticals. In particular, e-commerce payments have to be fraud screened and sent to the payment processor; the user needs an email receipt and then a confirmation. That might use a different provider for each step in different markets. The easy way to solve that is to just take the offering from Stripe for all of the above (except email), but reality doesn't work that way. I guess that's a long way of saying; I don't think payments orchestration companies have much to worry about here. Yet.
🥊 Quick hits
Apple's Prepaid Card is now "Apple Cash." Apple has repositioned its prepaid card as a "cash" offer allowing simple P2P transfers or payments with Apple Pay. 🤔 As with everything Apple does, you wonder, why didn't someone do this sooner? Other people have used the label cash before, but something about how Apple does it just makes sense. It's cash, send it to friends, pay with it, simples. Also, does the marketing page look like rent-a-gen-z-fintech-homepage to anyone else?
Good Reads 📚
Helping customers get money sooner is a competitive advantage that isn't always obvious. Before Stripe and Square, the industry average for getting paid in commerce was 7 to 10 days. Time to money is the gap between you creating value and receiving it. Longer gaps require working capital (or debt) to help manage those gaps. This becomes a tax on the vulnerable; rich people are okay waiting for a few days, but lower-income segments become insolvent quickly.
There are three ways to enable faster funds in the USA, "RTP" or real-time payment is available for 26 banks or around 70% of checking accounts. Several Fintech companies use "Good Funds" to advance users in scenarios where they are confident they will be made whole (e.g., Chime's get paid early, Instant Deposit on Stripe and Square). Float loans (e.g., lending to an employer for earned wage access). Ayo suggests this leaves two massive opportunities on the table: real-time tax withholding and real-time paychecks.
🤔 Real-time paychecks are easy to build in theory but carry a ton of hidden risk. Instant payments are a feature to manage the buyer's delivery risk as they're a bug for the seller. What happens if you pay an employee but don't get your work delivered or get defrauded? To do paychecks in real-time is easier when you have high confidence work is happening in real-time. If you're Uber or Lyft you have the data to see that, but what about knowledge or creative work? The answer is getting more data to understand the risk of a negative outcome. Those who best understand the real-time risk can have the highest confidence and success paying in real-time.
🤔 Why aren't taxes just auto-calculated and pre-deducted? That's one of those schleps in life that most of us would pay to remove. This should be a feature in every merchant checkout and acquiring stack (no surprise Stripe is already heading that way). But for consumers, just wow, yes, let's do this. Let me feed you all my income, investments, open banking data, and auto-file.
🤔 While "time to money" delays are a quirk of the US payment system, it's still a global problem for consumers and merchants. Having a real-time payment system doesn't mean you'll suddenly get paid in real-time. Most suppliers still get paid on 30+ day invoice schedules in the rest of the world. Put another way, RTP won't fix everything. It comes back to delivery risk and having better data. How can I be confident I can pay in full for a value that hasn't been delivered yet?
Extra credit 🤓
Crypto on and off ramps by Liam Roberts is a great explainer of the space.
Varo Unit economics breakdown by Jason Mikula, I linked to this in the Rant, but it really does get into the cold start problem Neobanks have in becoming banks.
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 :)