Trade Alert 1/8/24
If you’ve been following along for the last several months, you may not be surprised by this trade alert.
Simply put, we think the valuation is better for the Nu company (terrible pun) than the ones we’re trimming. It’s certainly not an apples-to-apples comparison since you can’t really compare software and banks but just projecting out 5 years, there seems to be more upside to Nu than the other companies.
Below is a lot more detail on the company to get you up to speed, starting with the background of the founder, then diving into the quality of the business, and finishing up with the valuation. Enjoy!
Born in Colombia to an entrepreneurial family (all 17 of his aunts/uncles run businesses), David Velez saw firsthand some of the drawbacks of the Latin American banking system. This was especially relevant in 2008 when Velez was asked to open a regional office in Brazil for the private equity firm he worked at. To open a bank account, he had to go through a security screening, wait in line for hours, and even after that, it took four more months to finally get the account opened.
After that experience, while doing his MBA at Stanford, a friend introduced him to Doug Leone of Sequoia Capital. After graduating, Leone asked Velez to head up their search for Brazilian entrepreneurs as Sequoia thought that would be the next big innovation hub. However, after realizing that the entire country of Brazil graduated just 42 computer science majors, Sequoia shuttered its Brazil plans in October of 2012. Rather than accepting a job back at Sequoia, Velez decided to start his own business – a Brazilian neobank.
The Brazilian banking space was ripe for disruption. The top five Brazilian banks owned 90% of the market and at an average of 25% return on equity – these banks were among some of the most profitable in the world. A myriad of late fees and onerous interest rates were just part of the daily experience for a Brazilian. Velez raised some money from Leone and recruited two co-founders, Edward Wible as CTO, and Cristina Junqueira as the CEO of Brazil.
The team started out by offering a Mastercard credit card with no annual fee. Visa wasn’t willing to work with start-ups at that point and the Brazilian regulations were changing such that if Nubank wasn’t approved by April 2013, they’d be forced to obtain a banking license which would delay the launch date by three years. In fact, Velez once flew to the Netherlands to pick up a contract from a Mastercard office just because they couldn’t wait two days for FedEx shipping. Supposedly, Nubank’s initial product is the fastest that Mastercard has ever given approval.
In the early days, Nubank made money through the merchant discount rate. When customers would use the credit card, Nubank would get between 3-5% from the merchants for taking on the risk of the customer. This is how most neobanks start out as they can institute low credit maximums and quickly stem losses. Further, neobanks typically don’t have large deposit bases from which they can do all sorts of lending.
Importantly, Nubank is very good at focusing before moving on to new products. After launching in 2013, it took them four years until their second core product, a bank account. Then it took another year to roll out debit cards and personal loans. I’ve heard from employees that you wouldn’t believe how focused the company can be and how important the customer experience is. I think the natural inclination for some fintechs is to grow as fast as possible but without a strong foundation, too much growth can become quite risky.
However, Nubank has quite a robust suite of products now. What started out with a simple credit card, has morphed into bank accounts, secured and unsecured personal loans, an investing platform, small business accounts, and money transfers. The company also partners for other types of loans like mortgages or auto.
The thesis here is pretty simple – Nubank has the best customer experience in a vastly underbanked population. The company’s growth has been nothing short of astounding. Since the beginning of 2017, total customers have grown from 1.6 million to over 90 million. That is a CAGR of 99%! Over 90% of those customers come from Brazil – the rest are from Mexico and Colombia. These ~84 million Brazilian customers make up 1/3rd of the population over 14 years old. It’s safe to say that the company has some pretty strong brand recognition by now. Further, NPS scores in Brazil are 90, which are some of the highest I’ve ever seen. Not to be outdone though, the most recent scores out of Mexico are 95. This means that 95% of customers would strongly recommend using the product.
The story these days isn’t so much about Brazilian customer acquisition – though it has grown over 40% YoY, adding about 5 million customers every quarter – it’s about providing more services to those same customers. The monthly average revenue per customer (ARPAC) is $10. This includes the interchange fees on purchase volume, credit card and personal loan interest, and then lending revenue from customer deposits. At the end of 2019, ARPAC was under $3, which is a 35% CAGR. While rising interest rates make up about half of that growth, the company is still doing a great job at becoming more embedded in customers’ lives.
In the latest quarter, the company announced that it’s already the largest credit card issuer in Colombia and Mexico. Mr. Velez even commented that he’s surprised the progress in these two countries has been better than the early days of Brazil. The fact we’re seeing evidence of a repeatable playbook is quite exciting for the future growth of the company. But growth isn’t everything, the company also has a strong and growing moat.
The moat here is that Nubank is the lowest-cost producer. By utilizing technology, the company has the lowest customer acquisition cost and the lowest cost to serve. Without the need for physical branches and large marketing campaigns, the company can focus its efforts on improving the customer experience rather than acquiring customers. While 2020 was certainly a unique year, the company spent $19 million in marketing and for 2021, brought in 20 million new customers. In the US, I’ve seen estimates that the average customer acquisition cost can hover around $200. The thing that originally caught my eye when looking at Nubank was that up to 90% of new customers come organically, without any paid acquisition. That’s simply because of how good the product is. Low fees, ease of use, and no waiting four months for approval!
Another piece of evidence for being the low-cost provider is that Nubank has one employee for every 12,000 customers whereas incumbents need 12x the number of employees. Extremely low-cost acquisition coupled with a low cost-to-serve means that the long-term economics are quite interesting. One data point here is that at the end of 2020, the monthly cost to serve a customer was about $1.2 and the average monthly revenue per active customer (ARPAC) was $3.3. So that’s about a 64% gross margin. Well, in the past two years, cost to serve has dropped to $0.80 and ARPAC has increased to $10, or more like a 92% gross margin. The main takeaway here is that costs don’t increase linearly with revenue. The platform that Nubank has built is incredibly scalable and customer satisfaction isn’t being deprioritized, judging by the NPS scores.
And bank accounts are notoriously sticky, especially the account that is tied to where customers receive their salaries. Payroll loans are quite popular in Latin American countries and Nubank has yet to really focus on that yet. The main takeaway here is that Nubank’s retention rates will only increase as customers use more of its products. In short, the switching costs will only get stronger.
Nubank was trading for about $40 billion. The company did $7 billion in TTM sales and roughly $1 billion on the bottom line. Currently, the company trades for roughly 4-5x book value which is quite high for a bank. In contrast, Bank of America is around 1.2x and Chase is at 1.5x. But this also makes sense because these banks aren’t growing assets by over 70%. In fact, Chase’s assets will be about flat this year; it has taken Chase 9 years to grow 70%. Let’s put it this way – if Nubank’s book value per share can grow at 30% over the next five years, and the stock compounds at 15%, it will trade at 2x book, which I think would be a fair value. So you can sort of think of our expected return as the delta between the multiple shrinking to 2x and whatever the company ends up growing book value per share at. So if Nubank can grow at 40% over the next five years, then our return would be more like 25%.
That’s one way of thinking about the value but below is a slightly more nuanced way of looking at the scenarios.
Without including any more countries, the company has 130 million more Brazilian customers to go after, 90 million Mexican customers, and 40 million Colombian customers (these are the total populations over 15 years old subtracting current customers).
Even if Nubank acquires 50% of these customers, with current ARPAC rates, that gives the company $16 billion in eventual revenue. However, current Brazilian banks have monthly ARPAC’s more like $30-35. I don’t think Nubank will get close to that because it wants to keep fees low but I do think that it will continue bundling more and more products, deepening its customer relationships. I don’t think $15 of monthly ARPAC is out of the equation, undercutting competitors by 50%. With a 50% penetration rate (the current Brazilian penetration is 33%), assuming no more geographic expansion, and a little less than a doubling of revenue per customer, that gets us to $36 billion in revenue. The biggest bank in Brazil, Itau, does about $25 billion in revenue so this isn’t very far-fetched.
I don’t see why the company shouldn’t be able to trade at 3x sales, considering that 30% margins should be very doable. Right now, Itau’s margins are 25% and it has a much higher cost to serve, accounting for its 4,000 bank branches. If we assume that Itau can maintain higher margins because of extra fees but then cancel that o
ut because of Nubank’s efficiency, maybe a reasonable assumption is a similar margin profile.
At a 15x PE, that’s 4.5x sales. On our $36 billion revenue estimate, that’s a $162 billion company. Now, the question is just how long will that take. Five years seems pretty quick as the revenue CAGR would be around 60%. In ten years, the revenue CAGR would be more like 26%, yielding a 15% overall stock return.
Now, of course, lots of things could change, especially in the geopolitical landscape in Latin America. However, Nubank has accomplished this growth despite its core market basically being in a recession the entire time. Since the company’s founding, GDP/capita in Brazil has fallen over 30% on an absolute basis. Any semblance of an economic tailwind over the next decade should be good news for the company.
And now here’s what the portfolio looks like: