- Fintech Brainfood
- Posts
- Fintech π§ Food - Worldcoin and The Future of Identity
Fintech π§ Food - Worldcoin and The Future of Identity
Plus; Robinhood has its first profit (aww), Venmo is up 8% & HSBC profit doubles following SVB UK acquisition
Hey everyone π, 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 32,450 others by clicking below, and to the regular readers, thank you. π
Hey Fintech Nerds π
Q2 earnings season is in full swing ποΈββοΈ
HSBC turned Β£1 into Β£1.5bn of profit by acquiring SVB UK.
That might go down as the greatest acquisition of all time. (Covered in π Things to Know)
But it's not just the banks enjoying high-interest rates.
The counter-intuitive insight is Fintech companies are benefitting against the assumption of many.
Like mammals when the dinosaurs died, Fintech companies adapted faster.
Meanwhile, bank relevance continues to erode to younger, default global consumers. (Covered in π Things to Know)
Digital wallets are becoming the default instead of cards for a generation of consumers.
Venmo transactions are up 8% YoY; I'll bet all wallets see this. Banks and Fintech companies that aren't building a wallet or playing with one will erode in relevance to consumers. I unpack why by using PayPal q2 Earnings. (Covered in π Things to Know)
Open AI's Sam Altman's other project Worldcoin officially launched this week.
Since then, Kenya has tried to ban it, and Europe, the home of GDPR, has questions.
But identity is broken.
We need something like Worldcoin to replace it. I dive into why in the π£ Weekly Rant.
That and much more in this weekend's Fintech Brainfood π§ π
PS. Yes, the Worldcoin Orb gives me Half-Life Portal 2 vibes.
Here's this week's Brainfood in summary
π£ Rant: Worldcoin and the Future of Identity
πΈ 4 Fintech Companies:
π Things to Know:
HSBC Profits double. The SVB acquisition might be the GOAT.
PayPal reports Q2 results with revenues up 7% YoY. eCommerce is back.
Robinhood reports its first profit since going public. Sweet net interest income π€€
π Good Read:
Weekly Rant π£
Worldcoin and the Future of Identity
This Rant was triggered by the launch of Worldcoin by OpenAI founder Sam Altman.
Worldcoin is a radical new vision for digital identity and universal basic income (UBI). It attempts to answer the question of what will humans do when AI takes all the jobs?
Worldcoin, if you haven't seen it, uses:
Iris scanning via a device called "the Orb."
A digital wallet for uses to manage identity and finance
A digital token that is distributed to users as a sort of Universal Basic Income (UBI)
Zero Knowledge Proofs (ZKPs) to store any information about users in a decentralized way (built on Ethereum and Optimism)
On the surface, this project wins the entire "Tech Utopian BS" game on the Bingo Card.
There is a market for Worldcoin credentials in China, with identities from emerging economies like Kenya and Cambodia. Kenya has now moved to ban Worldcoin.
France, Germany, Spain, and the UK are investigating it for possible privacy breaches.
The sci-fi fan in me can't help but think of this scene from demolition man whenever I hear eyeball scanning.
Yet.
Worldcoin or something like it will change the world.
Eventually.
The core concept of "Proof of personhood" is an alternative to using Government issued identity as a primary key for who you are and how you interact in the economy.
Worldcoin is just one way to approach that broader idea.
The idea has massive merit, not as we do identity traditionally "OR" we do proof of personhood.
But as an "AND."
Because the Government Identity as the primary key is broken.
How companies and governments manage citizen identity creates more crime and cyber risk than it solves. It is exclusionary, creates privacy risks, and is incompatible with an increasingly global and digital world. Even the best examples, like India or Scandinavia, are not fully ready for the global world we're heading for.
The most counter-intuitive insight into digital identity, financial crime, and security is that more user privacy will create less crime.
How is that possible? For that, we must understand how identity works today, its problems, and the possible solutions.
Every time a Government tries to prevent end-to-end encryption or create backdoors, they take a step backward. They have the right intentions but the wrong solutions.
πββοΈ
What is regular identity?
Identity is an intuitive but complex concept
"Legal person" identity is a core primitive of our economy
Flavor 1: Government-issued identity (analog)
Flavor 2: Commercial-issued identity
Flavor 3: Government-issued identity (digital)
Regular identity is broken
Current identity models create privacy issues and more cyber risks than they solve
Digitized identity creates yet more privacy risk
Current identity models don't prevent the harms they're intended to
Current identity models are exclusionary
Current identity models are not ready for a global world
The premise of identity itself is flawed
An Alternative: Proof of Personhood
The goal is to prove a human exists and is unique and continues to be unique
Similar but different to Verifiable Credentials (VCs) or NFT-based authentication
Proof of personhood better than traditional identity because it has no central or single point of failure
However, it's unproven and currently incompatible with legal identity
What is Worldcoin?
It's made of 3 things, a wallet, a token, and an identity record
Users get an account through an iris scanner called the Orb
We need it to align AI and give humans income when AI rules the economy
The Case for Worldcoin π
It is at scale and is built with a team who can scale it further
It is making a genuine attempt to be privacy-preserving
If anything can become compatible with a legal identity, it might be Worldcoin
The case against Worldcoin π«
The human is the weak link. Social engineering will be a problem
European Governments are objecting and Kenya has tried to ban it on privacy grounds
Meaning it could end up centralized and have no privacy at all
Where I stand on Worldcoin π€
We need something better than AML/KYC predicated on identity
If anyone can get it done, it's this team
Worldcoin could be the best or worst thing to happen to humanity
Some thoughts that didn't fit the narrative π
Identity is evolving outside of Worldcoin, Apple, CLEAR and European eIDAS are part of the landscape
We need to move from identity being binary to confidence based
Reality is messy; we need moonshots like Worldcoin
1. What is regular identity? π
a) Identity is an intuitive but complex concept. Identity is a loaded word. In daily life, it can describe the 1000s of attributes that make a human individual distinct from another. It can describe hobbies, appearance, character traits, or where you grew up. We instinctively understand that there's a human and observe that human has attributes. Some are static, and some evolve over time.
Put another way. Humans are unique, have attributes, and some of their unique attributes persist over time.
That's not how the administrative states think about it.
b) "Legal person" identity is the core primitive of our economy. Every transaction, asset, legal action, or commercial arrangement between Governments, businesses, and individuals involves a "legal person." A legal person is a "thing" that can enter contracts, own property, sue, and be sued. The term "thing" could mean a human, business, or non-commercial entity (like a charity or Government).
Not only does a legal person get protected by law. You have to be a legal person to do anything. You can commit unlimited crimes if you successfully forge an identity or evade these controls. It is the most critical risk management component we have built our legal system around.
c) Government-issued identity (analog). Community registers date back 1000s of years, with records from China and Europe working with Churches to register land and family ownership. This model evolved as societies became more complex in the 20th century to include Government-issued identity documents. Examples include Passports, driver's licenses, and ID cards.
d) Commercial-issued identity. In addition to Government-issued identity, banks, utility companies, or administrators like local municipalities can help to confirm an identity. This is especially helpful in providing some additional confidence that the legal person is who they claim to be. This might take the form of a bank statement or utility bill.
e) Government-issued identity (digital). Some countries like India, China, and several European nations have issued fully digital identities. Aadhaar in India is the canonical example of a single Government database of more than 1.3bn people who receive a unique 12-digit number. While it doesn't replace the passport, it is roughly equivalent to a more secure Social Security Number. However, users can also use that number to quickly sign up for commercial products like banks, insurance, or healthcare (although this later became subject to e-KYC).
2. Identity is broken. π
a) Current identity models create privacy issues and cyber risks. If a Government or company that holds your identity is hacked, the "legal person" secure identity is compromised. Equifax, Sony, and Target leaked 100s millions of legal person identities. Identity fraud counts for 70% of all fraud reported in the UK. Creating a central database makes a single point of failure. Consumer sentiment is that governments and corporations collect too much data, and 81% feel they have no control over that data.
GDPR was an attempt to manage this concern, but it created a cookie consent form on every web page. The legal system and administrative state have not solved privacy but continue to tie themselves in knots trying to.
b) Digitized identity creates more privacy risk. We won't solve this problem with ID cards or legislation. Countries like India have created a more efficient administrative state but have not solved the single point of failure or privacy risk. China has used its central digital identity system to create a "social credit score." One way to read this is to build reputation and credibility for business transactions; another is for a state to coerce and control the population. The idea of reputation is useful, but should one (hackable) actor control that database?
c) Current identity models don't prevent the harms they're intended to. Identity is how we prevent terrorist financing, child sex exploitation (CSAM), and human trafficking. Yet our analog controls have an abysmal track record. I cite this research paper more than almost any other. It concluded that criminal enterprises retain 99.95% of the proceeds of crime. Estimates put the total amount lost to money laundering and economic crime between $800bn and $2trillion annually. On the low end, that's the GDP of Indonesia; on the high end, it's nearly the GDP of France.
d) Current identity models are exclusionary. Migrants, rural workers, and low-income populations disproportionately lack a "legal person" identity. The World Bank estimates 850m people lack an identity globally. Some people were never given documents at birth or lived too far from an administrative building.
e) Current identity models are not ready for a global world. Ask anyone who's tried to work across borders or immigrated to another country. The digital world is default-global. Identity is not. Immigration, currency, and trade are political issues, but at the personal and corporate levels, they're barriers to getting on with life.
f) The premise of identity itself is flawed. The assumption is that a government can issue documents or credentials granting you legal personhood. But this will never capture the broad spectrum of opportunity, risk, or data needed for a more efficient and fair global economy.
We need to shift the center of gravity in identity.
From central actor points at a human and declaring a legal person (pictured below)
To human (or thing) can manage its identity and attributes, with central actors helping validate that (pictured below).
There may be a better way.
3. An Alternative: What is proof of personhood? π€
Vitalik Buterin describes Proof of Personhood as
"a limited form of real-world identity that asserts that a given registered account is controlled by a real person (and a different real person from every other registered account), ideally without revealing which real person it is."
a) The goal is to prove a human exists and is unique and continues to be unique. No central owner controls the system. Typically projects distribute a token of nominal value to users (sometimes called a UBI token). The goal is to create incentives that are open and democratic, avoiding too much power concentrated on the wealthy or centralized actors. The two primary approaches are social-graph-based and biometric. To quote Vitalik:
Social-graph-based proof of personhood relies on some form of vouching: if Alice, Bob, Charlie, and David are all verified humans, and they all say that Emily is a verified human, then Emily is probably also a verified human. [..] Biometric proof of personhood involves verifying some physical or behavioral trait of Emily that distinguishes humans from bots (and individual humans from each other). Most projects use a combination of the two techniques.
b) Similar but different to Verifiable Credentials (VCs) or NFT-based authentication. VCs are an open standard for digital credentials; they represent information from a legal person's credentials (such as passport number or bank account ownership). They do not require (but might be supported by) a Blockchain. Many new projects also issue users with an NFT after completing checks (like KYC) to prevent having to complete those checks again elsewhere.
The takeaway is that Worldcoin is a biometric-based proof of personhood project.
Often these concepts are complementary rather than binary. A world where legal identity uses verifiable credentials and proof of personhood can make similar claims with VCs, NFTs, etc.
c) Why is proof of personhood better than traditional identity. No central or single point of failure. The incentive mechanisms, technology platforms, and cryptography are designed to avoid holding data or being subject to control. The systems can still theoretically compromise, but the likelihood is lower. It is global by default and less exclusionary because the single requirement is being a human that is provably distinct from others via the social graph or biometrics.
d) It's unproven and currently incompatible with legal identity. Despite projects operating for several years, they haven't operated at scale. Proof of Humanity noted they have 15,000 users as of last year. I can't find examples of governments or corporations actively collaborating with Worldcoin.
There are tradeoffs with every type of identity, and I could spend 1000s of words on those, but that's not the point. The point is Worldcoin.
Worldcoin says it has 2.1m users; it is backed by Y-Combinator and OpenAI founder Sam Altman so it is by far the most likely to meaningfully change the identity landscape.
4. WTF is Worldcoin? π
a) It's made of 3 things, a wallet, a token, and an identity record. To quote Worldcoin
Worldcoin is intended to be the world's largest, most inclusive identity and financial public utility, owned by everyone. The Worldcoin protocol currently consists of:
b) How to get an account. For users, the process is as follows (Quoting Vitalik)
Each Worldcoin user installs an app on their phone, which generates a private and public key, much like an Ethereum wallet. They then go in person to visit an "Orb." The user stares into the Orb's camera and, at the same time, shows the Orb a QR code generated by their Worldcoin app, which contains their public key. The Orb scans the user's eyes and uses complicated hardware scanning and machine-learned classifiers to verify that:
The iris is considered to offer strong fraud resistance and data richness. The Orb is only required to receive the tokens (WRLD). The iris doesn't verify who a person is, only that they're unique. The data is never stored in the device; the only data that ever leaves is a validation that the human is unique.
The Orb device creates a "hash" (long string of numbers) representing the user's iris. This hash gets uploaded to a centralized database but is intended to be decentralized "once the team provides the iris scanning works at scale." This gives users a WorldID, which lives in their World App (wallet) and is periodically issued the WRLD token.
c) The thesis for why we need it. The thesis goes: People won't have jobs when we have AGI, so we need a universal basic income. Interestingly on the blog, the founders also describe Worldcoin as an attempt at "Global scale alignment." They don't go into detail, but I'd speculate this refers to the alignment problem with AI. This makes sense if you consider Sam Altman's other project Open AI.
TL;DR, if you care about the future of AI, you should pay attention to Worldcoin.
5. The Case for Worldcoin π
a) It is at scale and is built with a team who can scale it further. 2m users are far from an irrelevant number, and they're scaling this up to 35 cities in 20 countries. Worldcoin is available on Okta's OAuth marketplace (making it potentially compatible with social networks). Sam Altman has an unrivaled network of the Tech elite and is positioned to help bake Worlldcoin into generative AI projects. If Tech and Generative AI are in the future
b) It is genuinely attempting to be privacy-preserving. Co-founder Alex Blania has spent substantial time and energy appearing on Tech, mainstream, and Crypto podcasts and media outlets. The Crypto podcasts are his toughest critics diving deep into the technical nuances and risk of hacks or compromise. The use of Ethereum and Optimism L2 is slowly winning credibility. The depth of content published on their engineering blog about the device and Cryptography is far more transparent than it could have been. But this community is burned by media-savvy founders like former FTX CEO Sam Blankman-fried.
c) If anything can become compatible with a legal identity, it might be Worldcoin. Decentralized ID was a bust until Worldcoin. In a rush to regulate Generative AI, Sam Altman has proven he can engage with Governments and regulators. He might do the same with Worldcoin.
6. The case against Worldcoin π«
a) Will it create a market for eyeballs? My biggest worry is the social engineering aspect and getting 1000s of villagers to scan their eyeballs on the threat of violence to collect WRLD for themselves. The problem with identity is not just the human; it's the coercion and scamming of the human. This isn't just a worry. There is a market for Worldcoin credentials in China, with identities from emerging economies like Kenya and Cambodia.
There's also the creepiness factor. There's a scene in the classic early 90s movie demolition man where the bad guy steals an eyeball to escape prison. Worldcoin hasn't tested for the "stolen eyeballs edge case." Although they have noted on podcasts, the camera looks for signs of liveness in the eyeball (like blood pumping through it).
b) Will it remain private? Germany, Spain, France, and the UK have all raised concerns. The continent that brought you GDPR is actively investigating Worldcoin. The Worldcoin subsidiary is registered in Germany (of all places) and so falls under the Bavarian authority on privacy, which might be the world's most stringent. It's unclear if users having their eyeballs scanned understand what they consent to or who controls that data once it's scanned.
I get it. Worldcoin is decentralized, so these questions are hard to answer. But when it comes to consumer identity and their fucking eyeball, they should be in control. What happens if someone goes wrong? What happens if Tools for Humanity (the for-profit registered in Germany) is hacked? What is Worldcoin Foundation (the non-profit registered in the Cayman Islands) responsible for? Despite the effort to make the tech private, the issue might not be the tech.
c) Will it get centralized or be coerced? As of now, the entire biometrics database is centrally run and managed. If a Government arrests the founders and developers, it could compromise how the system works, no matter how well designed.
7. Where I stand on Worldcoin π€
a) We need something better than AML/KYC predicated on identity. It is ineffective. I've written about this numerous times before, but this piece captures it beautifully. We need something better than how we do identity today.
"AML/KYC laws costs about $300 billion a year and recover perhaps $3 billion a year in illicit funds⦠Indeed, AML/KYC laws have probably increased crime because they require so many companies to store personal information."
b) If anyone can get it done, it's this team. Worldcoin learns fast, engages transparently, and understands the tradeoffs between pragmatic building, privacy, and decentralization. I hope their fast learning includes the sheer number of edge cases social engineering creates.
c) Worldcoin could be the best or worst thing to happen to humanity. Like splitting the atom or AGI, the most ambitious breakthrough technologies are dual use. Worldcoin + AGI could be how we find ourselves unable to escape our AI overlords who can identify each of us with the tech we gave them. Or they could create unlimited wealth freeing humans up to discover a new sense of purpose in the stars as explorers.
8. Some thoughts that didn't fit the narrative π
a) Identity is evolving outside of Worldcoin. In the United States, companies like Apple and CLEAR are making a play to push identity closer to the device with more classic Web2 and mobile technologies. I suspect these forms of validation will or should merge (using something like Verifiable Credentials).
b) We need to move from identity being binary to confidence based. I dislike binary outcomes, and identity is predicated on one process of fully identifying a human. The reality is we change over time. Our preferences, behavior, and even some biometrics change. Every organization or human interacting with another needs a consistent confidence score of the risk and reputation of a counterparty. Are they who they say they are? Are they creditworthy? Has that changed over time? (This is a broader theme that needs its own Rant).
9. Most things are AND not OR
Reality is messy.
It's rarely doomsday or utopia; it's a muddle forward toward incremental improvements for the species.
Identity is the Higgs Boson of the digital economy.
Solve that, and everything else fits into place.
Better identity is a tipping point for our entire digital economy. But when it's such a crucial point of control, Governments will want to maintain it. Geopolitics will get involved (what happens if China compromised Worldcoin? Or the US did?)
We need weird projects like Worldcoin to happen and for them to hit enough scale to work out the kinks.
Identity is the hardest problem in the digital economy.
Worldcoin is just one project in a patchwork of exciting developments, and I hope they and others succeed.
Because we need a much better identity.
Identity that users own.
Identity that is effective, private, and secure.
ST.
4 Fintech Companies πΈ
1. Coris - LLM powered Merchant risk management
Coris allows payment facilitators or Vertical SaaS companies that offer payments to screen for risk at merchant onboarding or during operations. The platform uses a custom GPT-4 model to identify a merchant category (MCC) and NAIC code using the business name and address. It features real-time alerts (such as a website shutdown or a license suspension). Coris aims to save time on manual reviews with "Sitereview," which will check the website's professionalism and ensure that shipping policies are well documented.
π€ The superpower of LLMs is saving manual work. That schlep work that is contextual that humans are really good at suits LLMs. How often do you look at a website and make a snap judgment on its professionalism? That "feel," and speed is something payments companies pay humans to do; for humans, it's boring and time-consuming. The challenge of LLMs for a company like Coris is LLMs are so easy to use most companies of scale will build this themselves, or it will be a feature of a platform they're already using.
2. Fuse - Plaid for Account Aggregators
Fuse allows a single API to connect Neobanks, lenders, and expense management platforms to multiple account aggregation services (like Plaid and Finicity). This increases conversion, decreases disconnections, and allows companies to support multiple geographies.
π€ The worst thing about account aggregation is how bad the experience is when it doesn't work. When signing up for a new app or pulling data to make a payment, some banks won't connect to some aggregation services. The way around this is connecting to multiple aggregators (like MX and Finicity) and building a waterfall which can get expensive. Perhaps even more interesting is that as Fintech becomes default global, account aggregation also needs to. Fuse charges a fee on top of the aggregator fee, and I can't help but think this should be a feature of a BaaS platform or live somewhere else.
3. Goodwith - Where money meets mental health
Goodwith combines cognitive behavioral therapy (CBT) with AI and a chat interface to help users get better financial outcomes. Users track moods vs spending habits, and the app surfaces "education" or data about how to get better at managing finance based on where the user is in their journey.
π€ 8 out of 10 people's finances suffer when they have ill mental health. Often healthcare focuses on fixing the clinical but not the financial issue. 81% of young adults regularly feel anxious about money. The two are intrinsically linked. I love that Goodwith is working hard to land the right advice to the right person in the right context. I worry it's just another app in a sea of too many apps. This is where the big Fintech and tech companies could make a difference if packaged well.
4. Pay.so - Send Crypto, and they receive Fiat
Pay.so is a Crypto "off-ramp" as an API that connects to 40 currencies in 170 countries. Pay.so also offers a Crypto checkout with no-code, low-code, and developer-facing documentation.
π€ What's the moat here? I like the cleanness of the product, but it looks like a repackaged Nium or Routefusion for this use case. I believe Crypto as a rail for payouts is a huge opportunity. Getting volume and scale will require trust, which will require being great on the risk and regs side. They're regulated out of Lithuania as an e-money issuer, so they're almost certainly partnering for the Fiat off-ramping.
Things to know π
The London HQ'd bank posted a pre-tax profit of $21.7bn in H1, compared to $9.2bn in H1 2022. The CEO attributed much of this growth to the rise in interest rates driving up net interest income but noted: "tougher times are ahead." A provisional gain of $1.5bn was attributed to its acquisition of SVB UK.
π€ SVB UK was and is a great business. HSBC acquired a company for Β£1 ($1.28) and is making a $1.5bn profit from that line of business. That has to be one of the best deals of all time.
π€ Banking growth companies is an excellent spot in the market now with a wide open field of play. The spend management companies are aiming at it, but will more banks make a serious play here? What, if anything, JP Morgan might do with First Republic remains to be seen, and if Citizens can steward what remains of the SVB franchise. I speculate these will work out to be GOAT acquisitions looking back in a decade.
π€ The banks know the good times won't last. The warnings from CEOs are that the pain of interest rate rises is not yet showing in the data, especially in the UK, where inflation is still stubbornly higher than in the US, and rates may stay higher for longer. Banks have fewer deposits and a market where fewer can afford lending.
PayPal's Total Processed Volume is up 11% on mid-single-digit eCommerce growth. BNPL saw a 25% increase in first-time users, but with delinquency levels normalizing to "pre-COVID" levels. Venmo transactions also grew by 8%.
π€ eCommerce is stabilizing after the pandemic boom and bust cycle. The growth of eCommerce remains a secular trend, as does the default global nature of eCommerce. Perhaps the tech / VC crowd is getting a little over-exuberant with the AI-fuelled hype and growth, but as always, zoom out. Fintech is also a secular trend that PayPal benefits from.
π€ Is BNPL becoming a crutch for consumers facing higher costs? Kinda but not really. The economic growth in the US is partially attributed to a consumer who refuses to stop spending and an incredibly tight labor market. But are they spending on credit? Credit card debt is averaging at pre-pandemic levels, suggesting we're reverting to the mean. BNPL is growing, but so is eCommerce. Again, reverting to the mean. The bigger concern is what happens in 12 to 18 months? Do balances continue to grow, and can consumers service those debts now their deposits are all gone?
π€ Wallets like Venmo are where the growth in consumer payments will come from. While for many mainstream 40-something consumers, the card is the default, for 20-somethings, especially those outside the West, their default is a digital wallet. Any Fintech player not considering how they interact with wallets has a gaping hole in their strategy.
Robinhood reported revenue of $486m and a net income of $25m. Monthly actives are down to 10.8m from 14m a year ago. Transaction revenues were down at $194m, but net interest revenues were up 3x to $234m.
π€ Fintech companies are winning in a high-interest rate environment. Banks are having a fantastic earnings season, but the surprise is how well the Fintech companies are doing from net interest income.
π€ Banks are eroding in relevance to consumers. Bank revenues are driven by deposits and demand for lending, both declining in the high-interest rate environment.
π€ Fintech's ability to react faster isn't factored into most big bank strategies. It makes me think of the dinosaurs and the mammals. The dinosaurs didn't disappear overnight, but as the world changed, the mammals adapted faster.
π€ What goes up must come down. Robinhood and now likely many other Fintech companies' entire business is predicated on high-interest rates. Those rates won't stay this high forever. What happens, then?
π€ Does Robinhood have to Pivot its core brand identity? The trading side is not a draw for new or active users. There is nothing to sniff at 10.8m monthly active users and half a billion in revenue. But it can make more money as it diversifies and might acquire more users that way too. A repositioning would help do that but would be challenging from where they are.
Good Reads π
Adverse selection happens when a buyer or seller has more information than the other party (AKA information asymmetry). In lending, the borrower knows what they intend to use the money for and how serious they are about repaying; the lender doesn't. So lenders look at past performance (like your credit file) and try to figure out during the application if you're doing anything dodgy (like using a stolen identity). Trying to get the facts is expensive. While underwriting isn't perfect, it becomes not worthwhile if it gets too expensive. Credit scores are far from perfect, but they're efficient.
π€ More data about consumers can lead to better lending. Alex gives an example of a bank that does auto loans and spends much more time and energy underwriting to lend profitably. The missing killer blow is that that's exactly what you could do in embedded finance. A BNPL provider has all kinds of data about where you shop, how you shop, and what else you buy.
π€ Most consumer regulation is trying to fix information asymmetry. Information asymmetry works both ways. Professional investors might know more about a company in financial markets than consumers. That's why SEC regulations focus on "disclosures." The disclosures a lender has to make about how they lend and fee structure are also designed to avoid hiding things in the small print.
π€ All risk problems are data science problems. Great data science needs more data, cleaner data, and well-managed data. In turn, that creates a privacy issue. We would have far less lending or fraud risk if consumers had zero privacy. We can progress on this tradeoff with better cryptography and data science (see below).
Tweets of the week π
That's all, folks. π
Remember, if you're enjoying this content, please do tell all your fintech friends to check it out and hit the subscribe button :)
Disclosures: (1) All content and views expressed here are the authors' personal opinions and do not reflect the views of any of their employers or employees. (2) All companies or assets mentioned by the author in which the author has a personal and/or financial interest are denoted with a * (3) Any companies mentioned in Rants are top of mind and used for illustrative purposes only. (4) I'm not an expert at everything you read here. Some of it is me thinking out loud and learning as I go; please don't take it as gospelβstrong opinions, weakly held.