NU Nu Hit $1 Billion in Annual Profit. Its Founder Called It Time to Plant, Not Harvest.

Brazil Is Already a 40% ROE Business. The Group Reports 26%.

Nu Holdings' fourth-quarter and full-year 2023 results (period ended December 31, 2023, its second annual report as a public company after nu/2021-12 and nu/2022-03) are the first Nu filing where the growth story and the profit story actually agree with each other. Full-year revenue reached $8.03 billion, up 67.5% nominally from $4.79 billion in FY2022. Full-year net income was $1.03 billion, a swing of nearly $1.4 billion from FY2022's $364.6 million loss. None of that profit is a non-IFRS construction this time - Adjusted Net Income (which strips out share-based compensation and its tax effects) was $1.2 billion, but the IFRS net income is also genuinely positive, for the first time in a full fiscal year since Nu's 2013 founding.

The number worth sitting with, though, is one management volunteered rather than one buried in a footnote: "if one were to look at our operations in Brazil alone, our return on equity continued to increase and remains above 40%," CEO David Vélez told analysts on the call - against a consolidated annualized ROE of 26%. The 14-point gap is Mexico and Colombia, which in 2023 brought in just 6% of consolidated revenue while absorbing roughly 21% of Nu's total headcount (both figures per Vélez's own recap of the year). Brazil is now a mature, extremely profitable bank-like business; Mexico and Colombia are pre-profit investments being funded out of Brazil's earnings. That's not new information - Nu has said as much since the IPO - but FY2023 is the first year the math behind it is large and clean enough to actually see.

The Prescription

Vélez closed his prepared remarks with a line that doubles as Nu's actual strategy for 2024: "we believe it is time for planting, not for harvesting." Concretely, that means Nu should keep pushing Brazil's already-proven playbook (upmarket share of wallet, secured and unsecured lending expansion, the NuCel/insurance/investments cross-sell) precisely because it's now throwing off enough excess capital and cash flow to fund Mexico and Colombia's build-out without touching the balance sheet's $2.4 billion of holding-company excess capital or raising new equity. A digital bank with a 40%+ ROE core market and two genuinely early, still-small international markets is a rare enough setup that under-investing in the international leg now - to protect the consolidated margin optics for another year or two - would be the more costly mistake.

What Nu should stop doing: presenting its own YoY profitability comparisons using a base period that isn't the period it actually reported. Q4'23's $360.9 million net income is compared throughout the earnings release and call against "a $58 million profit in Q4'22" - but Q4'22's actual, filed result was a $297.6 million loss; the $58 million figure only exists after backing out a one-time $355.6 million Contingent Share Award termination charge that Nu itself incurred and reported (see Beyond the Usual). Nu already has Adjusted Net Income as its headline non-IFRS profitability metric - re-basing the comparison quarter on top of that, in the body of the earnings release rather than clearly flagged in the headline, is one adjustment layer more than a genuinely profitable, standalone-strong quarter needs to look good.

Key Financial Metrics

FY2023 vs. FY2022 and Q4'23 vs. Q4'22, consolidated, reported in USD - Nu reports natively in US dollars, so no FX conversion is needed

Metric	FY2023	FY2022	YoY
Total revenue	$8,029.0M	$4,792.2M	✅ +67.5% nominal
Gross profit	$3,491.0M	$1,663.0M	✅ +109.9% (margin 43.5% vs. 35.0%)
Total operating expenses	$1,951.9M	$1,971.9M	✅ -1.0% (24.3% of revenue vs. 41.1% - see note below)
Profit (loss) before income taxes	$1,539.1M	$(308.9)M	✅ Swung positive
Income tax expense	$508.5M	$55.7M	⚠️ Effective rate 33.0% vs. (18.0)% - see Beyond the Usual
Net income (loss) for the year	$1,030.6M	$(364.6)M	✅ Swung positive, first full profitable fiscal year
Adjusted Net Income, non-IFRS	$1,196.5M	$204.1M	✅ +486.2%
Cash and cash equivalents (year-end)	$5,923.4M	n/a in this filing	—

Metric	Q4'23	Q4'22 (as reported)	YoY
Total revenue	~$2.4B	~$1.5B (57% FXN growth implied)	✅ Record high
Gross profit	~$1.1B	n/a in this filing	✅ +87% FXN, margin 48%
Net income (loss), as actually reported	$360.9M	$(297.6)M	✅ Swung positive - see note on the Q4'22 comparison below
Adjusted Net Income	$395.8M	$113.8M	✅ +247.9%

Balance sheet metric	Dec 2023	Dec 2022 (per FY2023 disclosures)	Change
Total assets	$43,345.2M	—	—
Total liabilities	$36,938.8M	—	—
Total equity	$6,406.4M	—	—
Deposits	$23,691.1M	$15,847.0M (implied, per 38% FXN growth on $23.7B)	✅ +38% FXN
Interest-earning portfolio (credit cards + loans)	$8.2B	$4.0B	✅ +91% FXN

Nu's own press release states Q4'23 net income of "$360.9 million... compared to a $58 million profit in Q4'22" - but that $58 million isn't what Nu actually reported for Q4'22. Q4'22's real, filed net result was a $297.6 million loss, driven by a one-time $355.6 million Contingent Share Award (CSA) termination charge; the $58 million figure only appears after backing that charge out. Full-year, the same pattern repeats: FY2022's "$9.1 million net loss" figure used for comparison is also CSA-adjusted, against the actual filed FY2022 loss of $364.6 million (see Beyond the Usual for why this is flagged as a finding rather than a footnote).

Operating expenses actually fell 1.0% year-over-year in dollar terms while revenue grew 67.5% - genuine operating leverage, not just a favorable mix shift, and it drove the efficiency ratio (opex plus transactional expenses, divided by net interest income and fee income) to 36.0% for FY2023, a 19-percentage-point improvement over FY2022. Some of that improvement is mechanical: FY2022's opex included the one-off $355.6 million CSA charge, so FY2023's opex is being compared to an artificially high base. Even backing that out, FY2022 core G&A would have been roughly $977.7 million against FY2023's $1,042.3 million - a real 6.6% increase, still far below revenue growth, which is the more honest read of the operating-leverage story.

Key Operational Metrics

• Customers: 93.9 million globally (Dec 2023) vs. 74.6 million (Dec 2022) - ✅ +19.3 million net adds, +26% YoY. Brazil alone: 87.8 million, or 53% of the country's adult population - Nu is now the fourth-largest financial institution in Brazil by customer count, per Central Bank data cited in the release.
• Monthly ARPAC^»: $10.6 in Q4'23, up 23% FXN YoY; more mature cohorts already monetizing at $27/month, versus a blended base still climbing from a much lower start - the same cohort-maturation dynamic flagged in nu/2022-03.
• Monthly cost to serve per active customer: $0.9, essentially flat YoY but still sub-$1 - the other half of the operating-leverage story, holding steady even as ARPAC climbs.
• Activity rate: a new all-time high of 83.1%, with Nu becoming the primary banking relationship for over 61% of monthly active customers.
• Purchase volume: $111.1 billion for FY2023, +37.2% YoY.
• Credit quality: Brazil's 15-90 day NPL^» ratio eased slightly to 4.1%; the 90+ ratio held stable at 6.1%, described by management as normal delinquency-bucket stacking from prior periods rather than a fresh deterioration.
• Geographic mix: Mexico crossed 5.2 million customers (with over $1 billion in Cuenta Nu deposits by year-end); Colombia crossed 800,000. Combined, Mexico and Colombia generated only 6% of FY2023 consolidated revenue while absorbing about 21% of total headcount, per management's own 2023 recap - the clearest quantification yet of how much of the group's cost base is subsidizing pre-profit international expansion. Nu still reports a single consolidated operating segment; no segment-level P&L breakout is available in the interim disclosures, only in this annual filing's country-level color above.
• Capital: management describes $2.4 billion of "excess cash held by Nu Holdings" at the holding-company level, on top of capital adequacy ratios roughly 2x the regulatory minimum in each operating country - both management-characterized, non-reconciled figures rather than a single filed balance-sheet line.

Beyond the Usual

A second, further-adjusted number is used to make Q4'22 look like a profit it wasn't

Nu's FY2023 earnings release states Q4'23 net income of $360.9 million "compared to a $58 million profit in Q4'22," and frames full-year net income of $1.03 billion against "a $9.1 million Net Loss in FY'22." Neither $58 million nor $9.1 million is what Nu actually reported for those periods - the real, filed results were a $297.6 million loss for Q4'22 and a $364.6 million loss for FY2022. Both comparison figures only exist after excluding a one-time $355.6 million Contingent Share Award termination charge that Nu itself recognized and disclosed as a real GAAP expense in 2022. Nu already has a standing non-IFRS metric (Adjusted Net Income, which strips out share-based compensation) for exactly this kind of "underlying performance" framing; re-basing the prior-year comparison figure itself on a second, undisclosed-in-the-headline adjustment - rather than simply showing the real $297.6 million and $364.6 million losses alongside the explanation - makes an already-genuinely-strong swing to profitability look even larger than the real numbers support.

The effective tax rate swung from negative to a third of pre-tax profit

FY2022's effective tax rate was negative 18.0% - Nu paid $55.7 million of tax despite a pre-tax loss, largely a function of taxable income in Brazil sitting alongside a pre-tax loss driven by the CSA charge and Cayman-domiciled holding-company items that don't generate a tax benefit. FY2023's effective rate normalized to a more conventional 33.0%. That's a large enough swing in either direction that a reader modeling Nu's future earnings power should treat the group effective tax rate as structurally volatile - driven by the cross-border mix between a lightly-taxed Cayman Islands holding company and a heavily-taxed Brazilian operating base - rather than assume 33% simply continues.

The credit loss allowance ratio climbed faster than the loan book that's supposed to justify it

Total expected-credit-loss allowance for credit cards and loans combined was 14.2% of the total receivable balance at year-end 2023, up from 11.9% a year earlier. Credit card receivables classified as high-risk (level 3) more than doubled the pace of overall book growth - up 84.8% YoY (from $598.8 million to $1,106.4 million) while total credit card receivables grew 56.4%. Nu frames this as proportional to a fast-growing, still-maturing book; the level-3 concentration growing faster than the book itself is the detail worth actually watching next quarter, not the headline NPL ratios, which held roughly flat.

A legacy pre-acquisition tax dispute at Nu Invest (the brokerage Nu acquired in 2022) was resolved against the company during 2023: total judicial deposits fell from $18.9 million to $3.5 million, "due to the withdrawal of a tax dispute that resulted in the transfer of the deposited amount to the Government" - plain language for losing (or settling on unfavorable terms) a withholding-tax proceeding predating the acquisition, with the escrowed funds paid out rather than returned to Nu.

Two nearly identical lawsuits from Brazilian payment processors - Getnet (filed October 2021) and Banco Safra (filed January 2023) - both allege that Nu Pagamentos and Mastercard Brasil charge excessive interchange fees on transactions processed through their point-of-sale devices using Nu prepaid cards, seeking a cap of 0.5-0.8%. Getnet's claim (R$64 million in alleged damages) was dismissed at trial and again on appeal in December 2023, with Getnet still pursuing a clarification motion; Safra's identical claim was dismissed at trial in October 2023 and is now on appeal. Nu estimates the likelihood of loss on the Getnet matter as remote and doesn't disclose a specific loss estimate for Safra.

A newer class action, filed September 2023 by Instituto de Defesa Coletiva against Nu Pagamentos and four other financial institutions, seeks R$50 million in damages tied to the 2023 collapse of 123 Milhas, a Brazilian travel-package retailer - the claim alleges the banks failed to let affected customers dispute or reverse credit card charges for undelivered travel packages once 123 Milhas suspended operations, a novel theory of card-issuer liability for a merchant's own failure that's worth watching if other Brazilian card issuers face similar claims.

Nu Holdings remains guarantor on a growing stack of subsidiary borrowings in Mexico and Colombia that don't appear as debt on the parent's own balance sheet, only as contingent guarantees: a $70 million JPMorgan México term facility, a $10 million second JPMorgan México facility (also guaranteed by Nu Pagamentos), a roughly $460 million combined draw on the April 2022 $650 million Citigroup/Morgan Stanley/Goldman Sachs/HSBC syndicated facility (split between Nu Mexico and Nu Colombia), and $265.1 million drawn under an IFC A/B loan facility in Colombia - all guaranteed by Nu Holdings, none of it new in kind from the $135 million Nu Servicios guarantee flagged in nu/2022-03, but meaningfully larger in scale as the international expansion accelerates.

Management's Key Message from the Call

Vélez's framing of 2023 was explicitly a scorecard against three stated priorities from the start of the year - scaling Brazilian lending (secured and unsecured), growing share of wallet in the upmarket segment, and ramping local-currency deposits in Mexico and Colombia - and he claimed a clean sweep on all three, before pivoting to the "time for planting, not harvesting" framing for 2024 discussed in The Prescription above. The specific numbers he volunteered to make that case - Brazil-alone ROE above 40%, and Mexico/Colombia's 6%-of-revenue-but-21%-of-headcount split - are unusually precise disclosure for what's still, formally, a single reportable segment with no required segment-level financial statements; management chose transparency here that the filing itself doesn't require.

What got much less airtime on the call than in the release: the CSA-termination-adjusted comparison figures flagged in Beyond the Usual. The call's prepared remarks used the same "$58 million profit in Q4'22" framing as the press release, with no verbal caveat distinguishing it from the real $297.6 million reported loss - a detail available only in the release's own footnotes and the non-IFRS reconciliation appendix, not flagged verbally to analysts on the call itself.

Target Valuation Range

DCF-implied equity value: roughly $24.8 billion (bear) to $47.8 billion (bull), base case $31.4 billion - against today's $39.7 billion market cap, only the bull case clears the current price. Priced for continued hyper-growth rather than for the ~$1 billion-a-year, mid-30s-P/E bank it has just become on paper - the base case suggests the market is paying up for execution risk in Mexico and Colombia that hasn't been proven yet.

Nu closed FY2023 at $8.33 (December 29, 2023), up 104.7% from $4.07 at the end of FY2022 - one of 2023's stronger large-cap recoveries, tracking the broader growth-stock re-rating as rate-hike fears eased, on top of Nu's own swing to real profitability. With 4,765,937,154 total shares outstanding (3,682,625,012 Class A, 1,083,312,142 Class B, both classes unaffected by any stock split - Nu has never split its shares since its December 2021 IPO):

Market cap → enterprise value	FY2023
Share price (period-end)	$8.33
Shares outstanding	4,765,937,154
Market capitalization	~$39.7B
Total liabilities	n/a (net cash basis used)
Less: cash and equivalents	$5.92B (vs. ~$1.17B combined borrowings/financing/leases)
Enterprise value	~$35.0B

Peer-multiple sanity check	Q3 2023	FY2023	Change
Revenue basis	TTM	Full year 2023	-
Enterprise value	~$32.4B	~$35.0B	⚠️ up
EV/Revenue	~4.6x	~4.4x	- roughly flat, revenue base grew into the multiple
P/E	~26.8x (annualized)	~38.5x (trailing IFRS) / ~33.2x (Adjusted Net Income)	⚠️ up, genuinely comparable to a growth bank for the first time
Peer P/E (qualitative)	n/a	Nu's ~33-38x vs. single-digit-to-low-teens P/E typical of large incumbent Brazilian banks (Itaú, Bradesco) - not independently verified against their own filings, so qualitative only	-

DCF (base case, using FY2023 net income of $1,030.6 million as the starting base, projected forward on a declining growth glide path, discounted at a 15-16% cost of equity reflecting Nu's Brazil/Mexico/Colombia country risk and beta - Nu carries negligible net debt, so cost of equity and WACC are effectively the same here):

Scenario	Growth glide path (Y1-Y5)	Discount rate	Terminal growth	Implied equity value
Current (FY2023 close)	— actual market price, for reference	—	—	$39.7B
Bear	45% → 18%	16%	5%	~$24.8B
Base	50% → 20%	15%	5%	~$31.4B
Bull	55% → 25%	13%	5%	~$47.8B

Against the actual $39.7 billion market capitalization, the bear and base cases both imply Nu is trading rich (21-38% above intrinsic value on those assumptions); only the bull case - which assumes both faster sustained growth and a lower discount rate than Nu's emerging-market risk profile arguably justifies - gets to a fair value above today's price. Reverse-DCF read: the current price requires something close to the bull-case growth path to be right, which in practice means Mexico and Colombia need to scale toward Brazil-like profitability within the next few years, not merely keep growing revenue - exactly the bet Vélez described in Management's Key Message above. This is a real, arguable bet, not an obviously wrong one, but it is a bet: the base case built on Nu's own trailing numbers alone doesn't yet justify the current price.

---