October 2024 Fintech Newsletter: Fintech isn’t dead. AI is driving a new beginning 

Fintech isn’t dead. AI is driving a new beginning  ?

Angela Strange, James da Costa

Over the past 20 years, almost all innovation in the fintech sector has occurred in line with wider product cycles. 

• In the early internet and PC eras, fintech companies created new experiences by taking existing financial products and “putting them on a website.” For example, LendingClub started to make loans on the internet instead of in bank branches; PayPal enabled users to send money over email instead of writing a check. 

• During the cloud era, every layer of the fintech stack became available “as a service.” This led to our thesis that “Every company will become a fintech company,” as fintech became a key source of monetization for every software company. Companies like Moov offer payments, Sardine and Sentilink offer fraud and compliance solutions, and Spade offers real-time merchant intelligence. 
• In the mobile era, winning fintechs took advantage of being in their customers’ pockets. Chime’s app, for example, brought banking to users’ fingertips with a full stack bank and attracted customers with a “get your paycheck early” wedge. Robinhood built a commission-free mobile trading app that enabled users to trade fractional shares. 

In this new AI product cycle, labor is becoming software. For the world of fintech, that means thousands of white-collar jobs at financial institutions are in line to be augmented by AI copilots and agents. Take compliance officers, who represent the fifth-fastest growing role in the U.S. over the past 20 years. At one of the large banks, 30,000 of 210,000 employees work solely in compliance. In response to an increased volume of compliance requests and more complexity, banks and financial services organizations threw more labor at the problem. With AI, companies can think software first.  

Banks and insurance companies have an opportunity to take a hard look at their legacy software stacks and challenge the notion of “XYZ is too difficult to rip and replace.” New AI competitors might be 10x better and make the revenue upside and cost savings opportunity more than worth it: just take a look at what Vesta is doing for mortgage loan origination, Valon for mortgage servicing, and hyperexponential for insurance pricing and underwriting, to name a few.      

Why You Shouldn't Fear "Trading Like A Bank"

Santiago Rodriguez, Alex Immerman

Last month we wrote about when pursuing a bank license makes sense and how to think through the various licensing options. 

This month, we want to address the myth that if you become a bank, you will “trade like a bank” and be valued at lower multiples, which is one of the most common concerns we hear from fintech companies when deciding to launch new products that may make them look more like a bank.

When a fintech gets a bank license, there isn’t a market switch that flips and suddenly starts valuing them as banks. Valuations are nuanced and depend on the nature of the fintech company’s business model. The more a company’s economic model looks like a traditional bank’s at maturity — meaning it generates the majority of its income via interest and requires equity investments to fund growth — the more likely it is to be valued as a bank (on the basis of its expected ROE). 

But having a bank license does not mean a lower price-to-book multiple, as illustrated by a number of public companies today. For example, Block, American Express, Discover, and Nubank hold bank charters but trade at higher P/B multiples than non-banks like Upstart or Affirm. Simply put, they trade at higher book value multiples because their economic model is very different from a bank (like Block) or because investors expect them to earn a higher ROE at maturity (like Nubank).

Why do banks trade like banks?

Investors tend to value banks on a price-to-book (“P/B”) multiple or equity valuation divided by book value of equity. This multiple is shorthand for residual income models that forecast the “excess” return that a bank earns—that is, the difference between its return on equity (“ROE”) and its cost of equity. A multiple of 1x implies a business will earn a ROE equal to its cost of capital. For example, Bank of America valued at 1.1x implies the market expects it to generate an ROE just above its cost of capital. 

Large banks in the U.S. generate a ~8-12% ROE and trade for ~0.8-1.2x book value, meaning they are expected to return just about their cost of capital (~10%). Public valuation data demonstrate that the market makes assumptions about the ROE a bank will generate and uses shorthand P/B multiples to value it. Companies that are able to generate sustainably higher ROEs command higher valuations.

Why successful fintech models can trade better than traditional banks

Fintech disruptors have an economic model capable of earning higher ROEs than incumbent financial institutions for two main reasons: at maturity, we believe they will earn higher net income margins and will have higher asset turnover.

1. Higher Net Income Margins. While most fintech companies have not reached maturity, we believe they will earn materially higher margins than banks. First, higher margins will be driven by a lower cost to serve. Digital distribution enables a reduction in the cost to serve consumers by as much as 85% (as demonstrated by Nubank). Second, tech-driven underwriting capabilities can be significantly more effective in separating the risk of payment impairment. This means fintech companies can offer credit to the right high margin customers previously underserved by banking institutions. If a company can make a loan to the same customer with a lower cost to serve and a lower cost of risk than a traditional bank, it is poised to earn a higher margin.
2. Asset Turnover. Many fintech companies are able to generate more revenue dollars per unit of assets than traditional banks because their business models are typically driven by fees in addition to interest income. Fee revenue streams are unconstrained by the size of the asset base (unlike interest income) and can therefore drive higher asset turnover. There’s also examples of companies with and without bank licenses, like Block (with Cash App, Square and AfterPay), American Express, or Adyen, that generate the majority of their revenue from transaction fees and therefore do not trade like banks. 

There are two key examples that drive the point home. First, American Express is a bank that earns 75%+ of its revenues via fees, generates a 30%+ ROE, and is valued at >5x book value. Second, Nubank is in the early innings of proving out its better business model, but already achieved an annualized 40% Adj. ROE in its core geography of Brazil and therefore trades at a premium book multiple. Together, higher margin and asset turnover drives a higher ROE, which in turn leads to valuations un-anchored from bank comparables that trade for 1x P/B.

Successful fintech disruptors will do so with a differentiated model that unlocks a long growth runway. 

Differentiation may come from a number of sources, including:

• A unique product, like Greenlight’s financial toolkit for kids and teens or Monzo’s Personal Finance Management tools.
• Distinct distribution and network effects, like Revolut or Cash App’s peer-to-peer payment network.
• Product quality relative to incumbents, evidenced by Nubank’s NPS of ~2x other fintech companies and ~4x incumbent banks in Brazil. 

These competitive advantages will support high ROEs for many years to come, which brings us to our last point: growth. The next fintech champions will be those that earn high returns and have the opportunity to continue to invest in distinct distribution, customer acquisition, and aggressive product roadmaps that drive monetization. 

The financial services industry has unparalleled scale: $5T+ of gross profit and $30T+ of global public market cap. Companies that crack the code of building an efficient economic model with high growth potential have the opportunity to become the next big public success story. We remain excited about backing fintech disruptors that have a better economic model than incumbent banking providers.

More from the a16z Fintech Team

In this episode of In the Vault, a16z General Partner David Haber talks to John Stecher, Chief Technology Officer at Blackstone, about how Stecher covers how he decides whether to build tech in-house or partner with startups, what qualities he’s looking for in early stage companies, and how he sees AI impacting real estate, credit, energy, and ecommerce.

In this episode of the a16z Podcast, a16z General Partner Angela Strange and a16z investment partner Gabriel Vasquez chat with Dileep Thazhmon, cofounder and CEO of Jeeves, and Santiago Suarez, cofounder and CEO of Addi, about the future of fintech in Latin America and the unique approach required to succeed in this diverse market.

This Past Month in Fintech News

• Klarna shut down using Salesforce and Workday as service providers last month in favor of using custom AI-native software. They are aiming to employ as few as 2,000 employees (down from 5000 in 2022) in the coming years as they use AI in tasks such as customer service and marketing. 
• Apple now supports California IDs in the Apple Wallet, while Google is beta testing the ability to create a Digital ID from U.S. passports. Users will soon be able to tap their phone to go through airport security. 
• The U.S. Justice Department accused Visa of stifling competition in its debit card business. Details of the lawsuit are still emerging. 
• PayPal expanded into U.S. point-of-sale payments by integrating its debit card with Apple’s mobile wallet, driving competition for in-store payments.

Recent M&A Deals and Market Intel 

Workday announced its acquisition of Evisort, an AI-powered document intelligence platform, on September 17. Workday will make Evisort’s solutions available across its finance and HR suite with a range of use cases, including accounting, procurement, and employee knowledge base.

Mastercard announced its acquisition of Recorded Future for $2.65 billion on September 12. Recorded Future is the world’s largest threat-intelligence company and counts the governments of 45 countries and more than 50% of the Fortune 100 as clients. 

Hopscotch, the invoicing software provider for small businesses, announced its sale to Avalara on September 8. Hopscotch supports integrations to strengthen Avalara’s platform while continuing to operate as a standalone product.

Mercury announced its acquisition of Teal, a seed-stage startup that builds accounting products, on September 6. The acquisition is part of the company’s efforts to simplify complex financial workflows around the bank account.

Paylocity announced its acquisition of Airbase for $325 million on September 4. The acquisition expands Paylocity’s product suite and is expected to represent ~1% of FY25 revenue while diluting EBITDA margins by ~100 bps. An estimated revenue of $15 million for Airbase would represent an implied 22x multiple.

Visa is reportedly in advanced talks about a possible acquisition of fraud-prevention firm Featurespace. One source said the deal could be worth as much as $925 million, while another said the value could be markedly less.

Klarna is reportedly close to selecting Goldman Sachs to lead its U.S. IPO next year. The company is in talks with investors for a sale of existing shares that would come before the proposed IPO.  It had considered seeking a valuation of around $20 billion in the IPO.