Latest / AI Investment Edge: Stocks, Markets & Money / EP 41 | Nu Holdings Stock Analysis: Will Nubank Leave Big Banks Behind?
Transcript
- 0:00SPEAKER_01: NewBank just crossed a billion dollars in quarterly profit.
- 0:03SPEAKER_00: Yeah, it's huge.
- 0:03SPEAKER_01: It is, right?
- 0:04SPEAKER_01: I mean, it's a staggering milestone for a financial disruptor, but it sets up this,
- 0:09SPEAKER_01: well, this massive tension for anyone watching the stock.
- 0:12SPEAKER_00: Exactly.
- 0:13SPEAKER_01: Because, you know, we're looking at this and wondering,
- 0:15SPEAKER_01: are we looking at a purely digital architecture that is permanently dismantling
- 0:18SPEAKER_01: the economics of traditional banking?
- 0:21SPEAKER_01: Or is the market just like being fooled?
- 0:23SPEAKER_00: Mistaking a really good temporary credit cycle for a permanent technology advantage.
- 0:29SPEAKER_01: Exactly. Welcome to AI Investment Edge.
- 0:31SPEAKER_01: Today, we are doing a deep dive into new holdings.
- 0:35SPEAKER_01: That's the listed parent company of Nubank.
- 0:37SPEAKER_00: Right.
- 0:38SPEAKER_01: And the core takeaway right off the top is pretty striking.
- 0:40SPEAKER_01: The current price adjusted weighted iPulse AI rating checked on November 8th using
- 0:45SPEAKER_01: the November 6th close is a verified strong buy classification at both the shorter
- 0:51SPEAKER_01: and longer forecast horizons.
- 0:53SPEAKER_00: Which is definitely a strong signal.
- 0:56SPEAKER_01: Yeah. So we have this immediate action label of strong buy and keep.
- 1:00SPEAKER_01: Okay, let's unpack this because seeing that label on a high growth financial firm,
- 1:04SPEAKER_01: I mean, that requires a serious reality check on the underlying business model,
- 1:08SPEAKER_01: right? Can these research classifications actually hold up when we dig in?
- 1:12SPEAKER_00: It absolutely does require a reality check, especially when, you know,
- 1:16SPEAKER_00: we are dealing with the structural complexity of a bank. Right.
- 1:19SPEAKER_00: Let's lay out the Plain Business Foundation before we get into the valuation.
- 1:23SPEAKER_00: New is a digital financial group serving consumers and small businesses,
- 1:27SPEAKER_00: primarily across Brazil, Mexico, and Colombia.
- 1:30SPEAKER_00: But from an ownership perspective, the corporate structure matters immensely here.
- 1:34SPEAKER_00: They have a Cayman Islands parent company.
- 1:37SPEAKER_00: And what trades on the market is a New York Stock Exchange listed share,
- 1:42SPEAKER_00: and that's quoted in US dollars.
- 1:44SPEAKER_01: Even though the actual operating business uses Latin American currencies.
- 1:47SPEAKER_00: Exactly.
- 1:48SPEAKER_00: Those regulated operating subsidiaries, they are distinct legal entities from the Cayman parent.
- 1:54SPEAKER_00: And they are running this at massive scale.
- 1:56SPEAKER_00: I mean, well over 100 million total customers.
- 1:59SPEAKER_01: Though we should note for the listener that total customers are not all monthly active users.
- 2:04SPEAKER_00: Right. Good point.
- 2:06SPEAKER_01: And when you look at how they service those active users, there's a metric that stands out.
- 2:10SPEAKER_01: But I mean, it can be incredibly misleading.
- 2:13SPEAKER_00: The $17 metric.
- 2:14SPEAKER_01: Yes.
- 2:15SPEAKER_01: You see roughly $17 of monthly revenue per active customer measured against about $1
- 2:21SPEAKER_01: of monthly servicing cost.
- 2:22SPEAKER_01: Here's where it gets really interesting.
- 2:24SPEAKER_01: So are they just making $16 in pure profit per person?
- 2:28SPEAKER_00: No, definitely not.
- 2:29SPEAKER_00: That is a huge misconception.
- 2:31SPEAKER_01: Right, because that sounds too good to be true.
- 2:34SPEAKER_00: Because it is.
- 2:35SPEAKER_00: That is just a narrow servicing metric.
- 2:37SPEAKER_00: It's not a reflection of near zero all-in processing costs.
- 2:41SPEAKER_01: Okay, so what's missing there?
- 2:43SPEAKER_00: Well, a cheap app interface doesn't negate the fundamental cost of banking.
- 2:46SPEAKER_00: They still have to pay funding costs, overhead, growth investments,
- 2:50SPEAKER_00: and most importantly, credit losses.
- 2:52SPEAKER_01: The actual defaults.
- 2:54SPEAKER_00: Exactly.
- 2:54SPEAKER_00: They are actively expanding into higher risk lending, which inherently changes expected losses.
- 3:00SPEAKER_00: And we have to remember, expected losses are not guaranteed defaults.
- 3:04SPEAKER_00: But taking on that risk costs money.
- 3:06SPEAKER_01: That dynamic instantly reminds me of our analysis of Adyen back in episode 31.
- 3:11SPEAKER_00: Oh yeah, the pure payments platform.
- 3:13SPEAKER_01: Right.
- 3:13SPEAKER_01: We looked closely at how scale and low processing costs drive everything for them.
- 3:18SPEAKER_01: But new is completely different because they actually keep credit risk on their balance sheet.
- 3:23SPEAKER_00: Exactly.
- 3:23SPEAKER_01: If a borrower defaults, new takes the hit directly, whereas ad yen, you know,
- 3:27SPEAKER_01: they don't carry that specific exposure.
- 3:29SPEAKER_00: Right. And holding that risk bridges us perfectly
- 3:32SPEAKER_00: into the four quantified financial findings we need to evaluate today.
- 3:35SPEAKER_01: Okay, let's start with earnings.
- 3:37SPEAKER_00: So their Q2 IFRS revenue was over $5 billion.
- 3:41SPEAKER_01: Wow.
- 3:41SPEAKER_00: Yeah.
- 3:42SPEAKER_01: Over $5 billion.
- 3:43SPEAKER_00: Yeah. And their quarterly accounting profit was just over a billion dollars.
- 3:47SPEAKER_01: Just over a billion.
- 3:48SPEAKER_01: That's a huge milestone.
- 3:49SPEAKER_00: It is. It's a profound demonstration of real accrual earnings progress.
- 3:53SPEAKER_00: But, and I cannot emphasize this enough, this is accrual accounting earnings.
- 3:59SPEAKER_01: Not cash in the bank.
- 4:01SPEAKER_00: Right. It is not proof of cash collected, and it's certainly not free cash flow.
- 4:05SPEAKER_00: When analyzing a bank, you can never say that a profit milestone proves cash conversion.
- 4:10SPEAKER_01: Right, because accrual accounting recognizes revenue when it's earned on paper.
- 4:13SPEAKER_00: Exactly. Not necessarily when the cash hits the vault.
- 4:17SPEAKER_01: Okay, so that's the first finding.
- 4:18SPEAKER_01: The second one is profitability and credit, which feels like the real stress test here.
- 4:23SPEAKER_00: It is. We saw there after credit cost interest margin hit roughly 12%.
- 4:27SPEAKER_01: Which improved from Q1, right?
- 4:29SPEAKER_00: It did, but management explicitly notes that 12% is not a floor.
- 4:34SPEAKER_00: But the credit quality data...
- 4:37SPEAKER_00: Uh, that is where the friction really shows up.
- 4:39SPEAKER_01: Yeah, I was looking at this.
- 4:40SPEAKER_01: Their early arrears, the loans that are just starting to miss payments,
- 4:43SPEAKER_01: those actually improved seasonally.
- 4:45SPEAKER_00: Right.
- 4:45SPEAKER_01: But meanwhile, their late arrears, the loans more than 90 days overdue, rose to almost 7%.
- 4:51SPEAKER_01: I mean, on the surface, early delinquencies getting better while severe defaults get worse.
- 4:55SPEAKER_01: That sounds totally contradictory.
- 4:57SPEAKER_00: It sounds contradictory until you look at the mechanics of it.
- 5:01SPEAKER_00: It's a denominator effect driven by rapid loan growth.
- 5:05SPEAKER_01: Okay, walk me through that.
- 5:07SPEAKER_00: Think of it like pouring a turn of fresh water into a murky pond.
- 5:10SPEAKER_00: When a bank originates a fresh flood of new loans,
- 5:13SPEAKER_00: that immediately increases the total size of the loan portfolio.
- 5:17SPEAKER_01: And those new loans haven't missed a payment yet.
- 5:19SPEAKER_00: Exactly.
- 5:20SPEAKER_00: Because they haven't had time to go bad, your early arrears ratio artificially drops.
- 5:25SPEAKER_00: It dilutes the bad loans.
- 5:27SPEAKER_01: Ah, so it makes the immediate picture look cleaner.
- 5:29SPEAKER_00: Right.
- 5:30SPEAKER_00: But those new loans simply aren't old enough to reveal their inevitable losses.
- 5:35SPEAKER_00: So that late arrears number rising to almost 7%,
- 5:38SPEAKER_00: that reflects the older loan cohorts finally seasoning.
- 5:42SPEAKER_00: The profitable growth is real, but the credit seasoning is catching up.
- 5:46SPEAKER_01: And that seasoning directly eats into their capital.
- 5:49SPEAKER_01: Which brings us to the third finding, the balance sheet.
- 5:52SPEAKER_00: Yes, the funding side.
- 5:53SPEAKER_01: They have customer deposits of about $45 billion.
- 5:57SPEAKER_01: But we have to treat those as customer liabilities, right?
- 6:00SPEAKER_01: Not spare shareholder cash.
- 6:02SPEAKER_00: Exactly. That is money owed to depositors.
- 6:04SPEAKER_01: And at the same time, their Brazil Prudential Core Capital dropped from 13% to roughly 12%.
- 6:10SPEAKER_01: This connects directly to the framework we used for BNP Paribus in episode 33.
- 6:15SPEAKER_00: Oh, absolutely.
- 6:16SPEAKER_01: Where we talked about how a bank's profit, its liquidity,
- 6:19SPEAKER_01: and its distributable capital are completely different questions.
- 6:23SPEAKER_00: Right, because national regulators heavily ring-fence that capital.
- 6:27SPEAKER_00: Dropping to roughly 12% core capital shows that even a highly profitable bank burns
- 6:32SPEAKER_00: through its regulatory headroom when it expands as fast.
- 6:35SPEAKER_01: Which leads to our fourth finding, investment.
- 6:38SPEAKER_01: Growing this fast isn't cheap.
- 6:40SPEAKER_00: No, it's not.
- 6:41SPEAKER_01: In the first half, their physical and intangible investment used about $100 million.
- 6:46SPEAKER_00: Okay, about $100 million.
- 6:47SPEAKER_01: But separately, their infrastructure and data operating expense reached about $340 million.
- 6:54SPEAKER_00: Wow.
- 6:55SPEAKER_01: Yeah, that is up roughly half from the previous year.
- 6:58SPEAKER_01: So this whole branch light model still requires massive economic scale and technology.
- 7:03SPEAKER_00: Exactly.
- 7:04SPEAKER_01: Okay. Let's look at how all this impacts their liquidity.
- 7:07SPEAKER_01: Walk me through that H1 cash bridge.
- 7:09SPEAKER_00: Sure.
- 7:10SPEAKER_00: So they had an opening balance of roughly $15 billion.
- 7:14SPEAKER_00: Then operations used about $1,200 million.
- 7:18SPEAKER_01: That's a huge outflow.
- 7:19SPEAKER_00: It is. And then investing used $100 million.
- 7:22SPEAKER_00: Financing used $400 million.
- 7:25SPEAKER_01: And then the currency translation, right?
- 7:27SPEAKER_00: Yeah.
- 7:27SPEAKER_00: Currency added $300 million, which brings the ending balance to almost $14 billion.
- 7:32SPEAKER_01: Now, that negative operating cash flow,
- 7:34SPEAKER_01: the fact that operations used about $1,200 million, that always trips people up.
- 7:39SPEAKER_01: But as we discussed with PayPal back in episode 11,
- 7:42SPEAKER_01: we have to distinguish customer money from corporate cash.
- 7:45SPEAKER_00: Precisely. That outflow is driven by loan and deposit flows.
- 7:49SPEAKER_00: This is bank cash, not industrial free cash flow.
- 7:52SPEAKER_01: Okay. So let's tie all these financial realities back to the valuation.
- 7:57SPEAKER_01: We're looking at a current price of about $12 dated November 6th close.
- 8:01SPEAKER_00: Right.
- 8:01SPEAKER_01: And the original September forecast endpoints imply roughly 39% cumulative price
- 8:06SPEAKER_01: appreciation by original Q3 2027, which is, you know, less than a year away.
- 8:12SPEAKER_00: Yeah, very close.
- 8:13SPEAKER_01: And further out, it implies roughly 170% cumulative price appreciation by original Q3 2031.
- 8:19SPEAKER_01: So less than five years away.
- 8:21SPEAKER_00: Assuming zero future dividend cash or no uplift because they retain everything for growth.
- 8:26SPEAKER_01: Right.
- 8:27SPEAKER_01: So outline the base case for us.
- 8:29SPEAKER_01: What has to happen to hit those numbers?
- 8:31SPEAKER_00: The weighted base case basically requires deeper primary account relationships
- 8:35SPEAKER_00: and more useful products per customer.
- 8:38SPEAKER_00: Top line revenue has to grow faster than the cost of credit, funding and operations.
- 8:42SPEAKER_01: And what about Mexico?
- 8:44SPEAKER_00: Well,
- 8:44SPEAKER_00: Mexico's move into actual bank operations definitely widens their capabilities,
- 8:48SPEAKER_00: but regulatory approval alone doesn't guarantee profitability.
- 8:52SPEAKER_01: Got it. And the bull case.
- 8:54SPEAKER_00: The bull case is
- 8:55SPEAKER_00: that they maintain disciplined underwriting plus successful country replication,
- 9:00SPEAKER_00: which leads to durable margins supporting a richer valuation.
- 9:04SPEAKER_01: Okay, so the bear case is the opposite.
- 9:06SPEAKER_01: Mature credit losses pile up, deposits get too expensive,
- 9:09SPEAKER_01: and those overseas investments absorb all the operating leverage.
- 9:13SPEAKER_00: Exactly.
- 9:14SPEAKER_01: So the bear case is that investors wake up and value them like
- 9:17SPEAKER_01: a cyclical traditional lender instead of pricing them
- 9:20SPEAKER_01: at a premium software valuation.
- 9:22SPEAKER_00: I agree completely. That's the core risk.
- 9:25SPEAKER_00: And remember, these are qualitative, stressed cases, not invented numerical probabilities.
- 9:30SPEAKER_01: Right. So if we look at the four quadrants of their business model,
- 9:33SPEAKER_01: what are the primary drivers and frictions?
- 9:36SPEAKER_00: The drivers are clear.
- 9:37SPEAKER_00: Their cheap service architecture, their deposit funding, and their massive country scale.
- 9:42SPEAKER_01: And the frictions pushing back?
- 9:45SPEAKER_00: Seasoned unsecured credit, deposit yield competition, strict local capital rules,
- 9:49SPEAKER_00: and currency translation.
- 9:51SPEAKER_01: Which opens up both tails of the distribution.
- 9:54SPEAKER_01: On the upside tail, you've got them building a successful U.S.
- 9:58SPEAKER_01: banking franchise or maybe accelerating carefully underwritten payroll lending.
- 10:02SPEAKER_00: Right. The payroll loans.
- 10:03SPEAKER_01: But, I mean, payroll deductions don't magically eliminate defaults, right?
- 10:07SPEAKER_00: No, they do not.
- 10:08SPEAKER_01: Which?
- 10:09SPEAKER_00: If the customer loses their job, that payroll deduction is meaningless.
- 10:12SPEAKER_01: Which ties perfectly to the downside tail.
- 10:14SPEAKER_00: Exactly. Right.
- 10:15SPEAKER_00: The downside tail is in employment
- 10:17SPEAKER_00: and credit shock hitting right alongside weaker local currencies.
- 10:20SPEAKER_00: That would damage both their local profits and the U.S. dollar translation of those profits.
- 10:25SPEAKER_01: Okay, so what does this all mean for the alpha gap?
- 10:28SPEAKER_01: Where is the market missing the plot?
- 10:30SPEAKER_00: The alpha gap is basically the space between the market seeing
- 10:33SPEAKER_00: a familiar Latin American consumer lender and the research seeing
- 10:38SPEAKER_00: an exceptionally cheap distribution and data engine.
- 10:41SPEAKER_01: But the hard question is, how much of that structural advantage actually survives
- 10:46SPEAKER_01: a full credit cycle?
- 10:47SPEAKER_00: Yes.
- 10:48SPEAKER_00: After funding, after defaults, and after required capital, AI can improve decisions,
- 10:54SPEAKER_00: sure, but management's claims aren't independent proof
- 10:57SPEAKER_00: that underwriting risk has vanished.
- 10:59SPEAKER_01: Right.
- 11:00SPEAKER_01: Profitable, seasoned customers resolve the gap, not just adding more customers.
- 11:04SPEAKER_00: Exactly.
- 11:05SPEAKER_01: Which probably explains the substantive advisor disagreement we're seeing.
- 11:10SPEAKER_00: I mean, we know a weighted consensus does exist that strong buy
- 11:14SPEAKER_00: at both horizons we talked about. Right.
- 11:16SPEAKER_00: But the dispersed individual evidence weights are not votes and they aren't proof
- 11:21SPEAKER_00: of an evenly split panel.
- 11:23SPEAKER_01: So what's the core debate there?
- 11:25SPEAKER_00: One side accepts the cost moat but expects credit seasoning to moderate.
- 11:29SPEAKER_00: So they apply restrained bank multiples and worry about capital restrictions.
- 11:33SPEAKER_01: And the other side?
- 11:34SPEAKER_00: The optimistic views model very optimistic country replication and expect
- 11:38SPEAKER_00: a persistent technology premium.
- 11:41SPEAKER_01: But wait, does good business growth automatically guarantee stock gains?
- 11:45SPEAKER_00: No, absolutely not.
- 11:46SPEAKER_00: A cautious value view can see profits growing
- 11:49SPEAKER_00: while the shares valuation multiple actually contracts.
- 11:52SPEAKER_00: The real disagreement is the durability of credit adjusted earnings versus
- 11:56SPEAKER_00: the price you pay for them today.
- 11:58SPEAKER_01: Okay. So to track that durability, we have to look at the quarterly catalysts.
- 12:03SPEAKER_01: Q3 results on November 12th are upcoming, and then there's the investor day on December 8th.
- 12:08SPEAKER_00: Right. And since we don't have inventor'd Q3 actuals yet,
- 12:11SPEAKER_00: we have to be very clear about the essential path forward.
- 12:14SPEAKER_01: Okay. Lay out that path.
- 12:15SPEAKER_00: After the 2027 credit seasoning test, the model's middle years, 2028 and 2029,
- 12:21SPEAKER_00: rely on deeper customer use and better overseas economics.
- 12:24SPEAKER_01: Okay, so 28 and 29 are about usage and overseas.
- 12:27SPEAKER_00: Right.
- 12:28SPEAKER_00: Then by 2030 and 2031, durable profits and retained regulatory capital must support
- 12:34SPEAKER_00: the terminal valuation.
- 12:35SPEAKER_01: But those are just models, right?
- 12:37SPEAKER_00: Emphatically, yes.
- 12:38SPEAKER_00: Those are model assumptions, not promised company profitability dates.
- 12:42SPEAKER_00: We have to keep these causal windows clear.
- 12:44SPEAKER_00: Long-term return figures alone do not replace them.
- 12:47SPEAKER_01: Got it.
- 12:48SPEAKER_01: Okay, keeping those causal windows in mind, let's talk about the genuine business deltas.
- 12:52SPEAKER_01: What actually changed between July and September?
- 12:55SPEAKER_00: Well, the June buyback authorization was already known in July.
- 12:59SPEAKER_01: Right, so that's not a new delta.
- 13:00SPEAKER_00: Exactly.
- 13:02SPEAKER_01: The genuine business deltas for September were the later Mexican authorization
- 13:06SPEAKER_01: and the actual August bank opening.
- 13:08SPEAKER_00: Plus the Q2 monetization data, right?
- 13:11SPEAKER_01: Yes, and the worst later years and expense efficiency drag we discussed earlier.
- 13:15SPEAKER_01: We also saw that U.S. lead bank partner launch.
- 13:18SPEAKER_00: Which differs from new operating under their own conditional charter.
- 13:22SPEAKER_01: Right. Data learning isn't the same as legal weighting.
- 13:25SPEAKER_01: It's a commercial experiment with Gates.
- 13:26SPEAKER_00: Exactly.
- 13:27SPEAKER_01: That commercial experiment approach reminds me of RxCite in episode 40.
- 13:32SPEAKER_00: Ah, yes.
- 13:33SPEAKER_01: Where they had to carefully test clinical adoption before scaling.
- 13:36SPEAKER_00: It's a good analogy for testing the waters, but we have to contrast them carefully.
- 13:40SPEAKER_00: RxSite tested clinical adoption and procedure economics.
- 13:44SPEAKER_00: New must turn app engagement into risk-adjusted banking profit.
- 13:49SPEAKER_01: Right. Completely different beasts.
- 13:50SPEAKER_00: You cannot compare a medical device company's cash reserve
- 13:54SPEAKER_00: with a deposit-funded bank's liquidity.
- 13:56SPEAKER_01: Fair point.
- 13:57SPEAKER_01: Okay, so with all of this in play, what are the concrete tests to break this thesis?
- 14:01SPEAKER_01: What specific red flags would actually invalidate the research?
- 14:05SPEAKER_00: What's fascinating here is that the mechanisms for invalidation are highly specific.
- 14:10SPEAKER_00: The thesis breaks if older loan cohorts deteriorate faster than revenue can cover the losses.
- 14:15SPEAKER_00: It breaks if the after credit margin falls while late arrears rise.
- 14:20SPEAKER_01: Which we're already watching closely.
- 14:22SPEAKER_00: Right.
- 14:23SPEAKER_00: It also breaks if deposit prices consume their cheap distribution advantage
- 14:27SPEAKER_00: or if weak currencies turn local growth into disappointing dollar results.
- 14:32SPEAKER_01: Okay. Synthesizing all of this.
- 14:34SPEAKER_01: Nubank has demonstrated real scale in earnings, obviously,
- 14:38SPEAKER_01: but truly leaving established banks behind requires a cost advantage
- 14:41SPEAKER_01: that survives underwriting, regulation, and valuation.
- 14:45SPEAKER_01: The weighted strong buy classifications are conditional research conclusions, not guarantees.
- 14:51SPEAKER_01: It will be really interesting for you, the listener,
- 14:54SPEAKER_01: to test which financial interpretation wins as the upcoming Q3 results arrive.
- 14:59SPEAKER_01: For the bigger picture, visit iPulse AI to explore our complete rankings
- 15:03SPEAKER_01: of covered assets, discover the strongest buys,
- 15:06SPEAKER_01: and see the research behind the ratings all in one place.
- 15:09SPEAKER_00: Primary research conducted through the iPulse AI platform by Future Edge Group.
- 15:14SPEAKER_01: If you enjoyed this deep dive, please like, comment, and subscribe.
- 15:17SPEAKER_01: We are publishing a new podcast every day of the week, except Thursday.
- 15:21SPEAKER_01: Stay tuned for the next podcast on Eden Red.
- 15:24SPEAKER_01: We'll be looking into how employee benefits and payments turn everyday spending
- 15:28SPEAKER_01: into a fascinating business model.
- 15:30SPEAKER_01: Thanks for listening.