CPNG The First Positive Adjusted EBITDA Quarter Has an Asterisk

A Fire-Sized Asterisk on the Best Quarter Yet

By the headline numbers, this is Coupang's best quarter since its IPO. Consolidated Adjusted EBITDA^» was positive $66.2 million - the first positive quarterly Adjusted EBITDA in the company's history as a public company, following five straight negative quarters going back to Q1 2021. Net loss narrowed 85% year-over-year to $75.5 million. Operating loss narrowed 87%. And Product Commerce swung to Segment Adjusted EBITDA of $97.8 million, from a $47.9 million loss a year earlier.

Before taking any of that at face value, it's worth remembering what was sitting inside last year's comparison quarter. This site flagged in the Q2 2021 post that a June 2021 fire at Coupang's Deokpyeong fulfillment center added $296 million to that quarter's net loss - $158 million of inventory losses in cost of sales, $127 million of property losses in operating expenses, and $11 million of other directly related costs. None of that $285 million operating-level hit was a recurring cost, and none of it repeats this year. Every year-over-year comparison in this post's headline numbers is measuring Q2 2022 against a quarter that was still absorbing a one-off warehouse fire.

Stripping the fire back out of Q2 2021 (as that quarter's post already did) puts ex-fire operating loss at roughly $(230) million and ex-fire net loss at roughly $(222.6) million. Against those normalized figures, this quarter's actual operating loss of $(67.1) million and net loss of $(75.5) million are still real, substantial improvements - operating loss down about 71% and net loss down about 66% even on a fire-adjusted basis. The improvement is genuine; it's just smaller than the raw headline suggests, and the segment-level picture (see Segment Performance below) complicates the "Product Commerce turned the corner" reading further.

The Prescription

Coupang should start disclosing a fire-adjusted comparison itself, at least until the June 2021 event rolls fully out of its trailing-twelve-month and year-over-year comparisons in Q2 2023. A company that already discloses the fire's exact dollar impact in a footnote (as this 10-Q still does, more than a year later) has effectively conceded the comparison is distorted - the only thing missing is doing the subtraction for the reader instead of leaving it to whoever bothers to dig through Note 3. Given how much of this quarter's positive-EBITDA headline rides on that easy comp, silence here reads as convenient rather than neutral.

What it should stop doing: letting the Product Commerce narrative run ahead of what the segment note can actually support. Last quarter's post called Product Commerce's swing to near-breakeven a real answer to whether the core retail business is profitable in isolation - and it still is, directionally. But this quarter's Segment Adjusted EBITDA table doesn't disclose how the fire's $285 million hit was allocated between segments, and the Deokpyeong facility supported owned-inventory retail fulfillment, which sits inside Product Commerce. If most of that hit landed there (a reasonable inference, not a disclosed fact - see Beyond the Usual below), Product Commerce's underlying ex-fire profitability a year ago was already comparable to or larger than this quarter's $97.8 million. Coupang doesn't need to change anything about the business to fix this - it needs to stop letting an undisclosed allocation choice make its healthiest segment's story sound more dramatic than the numbers alone support.

Key Financial Metrics

Q2 2022 vs. Q2 2021, consolidated - reported in USD

Coupang reports natively in US dollars, so no FX conversion is needed here. Figures below are drawn directly from this 10-Q's condensed consolidated financial statements and MD&A tables, covering the quarter ended June 30, 2022. Recall that Q2 2021's figures include the $296 million FC Fire hit discussed above - not adjusted for it in the table itself, since the filing doesn't restate it that way, but essential context for reading the "YoY" column.

Metric	Q2 2022	Q2 2021	YoY
Total net revenues	$5,037.8M	$4,478.1M	✅ +12%
Gross profit	$1,153.8M	$658.5M	✅ +75%, margin 22.9% vs 14.7% (includes $158M of fire-related inventory losses in the 2021 base)
Operating loss	$(67.1)M	$(514.9)M	✅ loss narrowed 87% (2021 base includes $285M of fire-related costs)
Adjusted EBITDA»	$66.2M	$(122.1)M	✅ first positive quarter; margin 1.3% vs -2.7%
Net loss	$(75.5)M	$(518.6)M	✅ loss narrowed 85%, margin -1.5% vs -11.6%
Net cash (used in) provided by operating activities	$(18.3)M	$30.9M	⚠️ swung negative
Free cash flow»	$(195.5)M	$(137.7)M	⚠️ outflow widened 42%
Cash & cash equivalents (period end)	$3,107.5M	n/a¹	⚠️ down from $3,487.7M at Dec 2021

¹ Q2 2021's period-end cash figure was already covered in that quarter's post and isn't restated here; the sequential comparison against Dec 2021 is the more useful read for cash movement over the first half of this year.

Balance sheet metric	Jun 2022	Dec 2021	Change
Total assets	$8,425.2M	$8,641.8M	⚠️ -2.5%
Total liabilities	$6,357.1M	$6,465.9M	✅ -1.7%
Total stockholders' equity	$2,068.1M	$2,176.0M	⚠️ -5.0%

The two metrics moving the wrong way this quarter - operating cash flow and free cash flow - aren't about the fire comparison at all. Operating cash flow swung from $30.9 million provided a year ago to $18.3 million used this quarter, and free cash flow's outflow widened 42% to $195.5 million. Working-capital timing explains most of it: accounts payable fell to $3,294.5 million from $3,442.7 million at year-end, a real cash use that has nothing to do with the fire and everything to do with normal payment-cycle movement. This is the metric worth watching next quarter, not the Adjusted EBITDA headline - a company can post its first positive Adjusted EBITDA quarter while still burning more cash than the year before, and this quarter is exactly that.

Six-month (H1 2022) totals: net revenue $10,154.5 million (+17% YoY), net loss $(284.8) million (down 65% from $(813.6) million), Adjusted EBITDA $(24.7) million (down 90% in loss terms from $(255.1) million). The first half is still Adjusted EBITDA-negative overall - Q1 2022's $(90.9) million loss outweighs Q2's $66.2 million gain - so "first positive quarter" describes Q2 specifically, not a first-half trend yet.

Segment Performance: Product Commerce vs. Developing Offerings

Coupang reports two segments this quarter, the same structure introduced in Q1 2022: Product Commerce (core retail, marketplace, Rocket Fresh, and related advertising) and Developing Offerings (Coupang Eats, Coupang Play, international expansion, and fintech).

Metric	Product Commerce Q2'22	Product Commerce Q2'21	Developing Offerings Q2'22	Developing Offerings Q2'21
Net revenues	$4,877.5M	$4,332.7M	$160.3M	$145.4M
Revenue growth	✅ +13% (27% constant currency)	—	✅ +10% (24% constant currency)	—
Segment Adjusted EBITDA	✅ $97.8M	$(47.9)M	✅ $(31.7)M	$(74.3)M

Product Commerce is 96.8% of consolidated revenue and, on its face, the segment driving this quarter's Adjusted EBITDA turn. But see the opening section above: the Deokpyeong fulfillment center that burned in June 2021 supported owned-inventory retail operations, which sit inside Product Commerce. The segment note doesn't disclose how the fire's $285 million operating-level cost was split between the two segments, but if most of it landed in Product Commerce - the reasonable inference given what the facility did - then this segment's underlying, fire-adjusted profitability a year ago was likely already in the neighborhood of this quarter's $97.8 million, or larger. That would mean Product Commerce's segment margin isn't newly positive this year so much as it's back to roughly where it would have been without the fire - a much less dramatic story than the raw comparison implies.

Developing Offerings is the segment with a cleaner read, because nothing about the fire touches it. Its Segment Adjusted EBITDA loss narrowed 57% year-over-year, to $(31.7) million, while revenue still grew 10% (24% constant currency) to $160.3 million. That's the opposite of last quarter's trend, where Developing Offerings' loss widened 47% faster than its 65% revenue growth. This quarter, the segment got more efficient per dollar of revenue even as revenue kept growing - loss narrowing against growing revenue is the base case this site laid out last quarter for how Developing Offerings' trajectory should look if it's maturing rather than just burning more to grow faster, and this quarter is the first one that actually looks like that. Notably, Developing Offerings' revenue also fell sequentially, from $180.6 million in Q1 2022 to $160.3 million this quarter, while its loss shrank at the same time - consistent with a deliberate pullback in spending on the segment's lower-margin activity rather than simply riding a bigger revenue base to better leverage.

Key Operational Metrics

Metric	Q2 2022	Q2 2021	YoY
Active Customers	17.885M	17.022M	✅ +5.1%
Total net revenue per Active Customer	$282	$263	✅ +7.2%

Quarter	Active Customers	Revenue per Active Customer
Q3 2021	16.823M	$276
Q4 2021	17.936M	$283
Q1 2022	18.112M	$283
Q2 2022	17.885M	$282

Active Customers fell sequentially this quarter, from 18.112 million to 17.885 million - only the second sequential decline since the IPO, after Q3 2021's first-ever drop. Revenue per Active Customer also slipped slightly, from $283 to $282, meaning both halves of the growth equation - customer count and per-customer spend - moved backward at the same time this quarter, something that hasn't happened together in a single quarter before on this site's tracking. Year-over-year the picture still looks fine (+5.1% customers, +7.2% revenue per customer), and Q2 is not a seasonally strong quarter for Korean e-commerce the way Q4's holiday period is, so a modest post-Q1 pullback isn't itself alarming - but two metrics moving the same direction at once, however small, is worth watching into Q3.

Net retail sales and net other revenue (third-party commissions, advertising, and other income) weren't broken out with the same granularity as prior quarters in this filing's MD&A; the segment table above is the more reliable read on where growth is actually coming from this quarter. "Not available" again this quarter: orders per customer, Rocket WOW membership subscriber count, and third-party seller/merchant count - none disclosed in this filing, consistent with every prior quarter.

There still isn't a same-quarter e-commerce peer covered on this site with a directly comparable fiscal calendar, so this remains a company-only trend for now.

Beyond the Usual

An undisclosed segment allocation is doing a lot of work in this quarter's headline number

Product Commerce's Segment Adjusted EBITDA swung from $(47.9) million to $97.8 million year-over-year - the single number most responsible for this quarter's "Coupang turned profitable" headline. But the segment note doesn't disclose how the $285 million of fire-related costs from June 2021 (see the opening section above) was allocated across Product Commerce and Developing Offerings. The Deokpyeong facility supported owned-inventory retail fulfillment, which points toward Product Commerce as the likely home for most of that cost - meaning a meaningful share of this quarter's dramatic segment swing may be an artifact of an easy year-over-year comparison rather than a genuine step-change in the segment's economics. This isn't a claim that the filing is misleading; the fire's total dollar impact is disclosed, just not split by segment. But a reader relying on the segment table alone, without also reading Note 3, would come away with a more dramatic story than the underlying numbers can actually support.

The Korean severance liability fell for the first time since the IPO

For the first time since this site started tracking it, Coupang's defined severance benefits obligation - the mandatory Korean labor-law liability first flagged in the Q4 2021 post and tracked at $253.4 million as of March 2022 - fell this quarter, to $229.5 million at June 30, 2022. The decline comes from a $34.2 million actuarial gain recognized in other comprehensive income this quarter (against a $9.9 million actuarial loss in the same quarter last year), consistent with the kind of swing that typically comes from a higher discount rate being applied to the liability rather than from headcount actually shrinking - South Korea's own policy rate rose meaningfully over this period. It's not evidence Coupang is cutting its Korean workforce; it's a reminder that this liability moves with interest-rate assumptions as much as with headcount, and this quarter that movement happened to go in Coupang's favor.

Total contractual commitments declined for the first time this site has tracked them

Coupang's disclosed minimum contractual commitments - unconditional purchase obligations, long-term debt including interest, and operating lease payments, all on an undiscounted basis - totaled $3.319 billion as of June 30, 2022, down from the $3.462 billion flagged last quarter and the $3.119 billion the quarter before that. This is the first sequential decrease in this figure across the four quarters this site has tracked it, a break from the steady fulfillment-center-buildout growth pattern seen since Q4 2021. One quarter of decline isn't enough to call a trend change in Coupang's capex/commitment appetite, but it's worth checking again next quarter against whether it's a genuine pullback or just contract-timing noise.

The revolving credit facility is still entirely undrawn

Coupang's $1.0 billion revolving credit facility remained fully undrawn as of June 30, 2022, same as every prior quarter this site has tracked it. Combined with $3.108 billion of cash and cash equivalents ($3.272 billion including restricted cash), the facility continues to sit as unused liquidity rather than a funding source actually being tapped, even in a quarter where operating cash flow and free cash flow both moved in the wrong direction (see Key Financial Metrics above).

Target Valuation Range

EV $19.9 billion, ~0.99x EV/Revenue. Bottom line: the stock fell a further 28% over the quarter to close at $12.75, on top of Q1's roughly 40% decline - even as the company posted its first positive Adjusted EBITDA quarter, the market kept repricing Coupang lower, which is the clearest evidence yet that this stock's 2022 decline has been driven by macro multiple compression across unprofitable growth names, not by anything specific to Coupang's own quarterly execution.

Coupang closed Q2 2022 (June 30, 2022) at $12.75, down from $17.68 at the end of Q1 2022 - a further 28% decline, on top of Q1's roughly 40% drop, for a combined decline of about 57% from where the stock started the year. With 1,590,313,751 Class A shares and 174,802,990 Class B shares outstanding (1,765,116,741 total) at quarter-end, that implies a market capitalization of approximately $22.5 billion, down from $31.1 billion three months earlier and $51.5 billion at the end of 2021.

Using cash and cash equivalents of $3.108 billion plus restricted cash of $163.5 million against total debt (short-term borrowings plus current and long-term debt) of approximately $649.2 million, net cash comes to roughly $2.62 billion, putting enterprise value at approximately $19.9 billion.

Market cap → enterprise value	Q2 2022
Share price (period-end)	$12.75
Shares outstanding (Class A + B)	1,765,116,741
Market capitalization	$22.5 billion
Net cash (cash & equivalents less total debt)	$2.62 billion
Enterprise value	$19.9 billion

EV/Revenue sanity check	Q1 2022	Q2 2022	Change
Enterprise value	$28.2 billion	$19.9 billion	down
EV/Revenue	1.38x	0.99x	down

• EV/Revenue^» (annualizing Q2 2022's $5.038 billion revenue to a $20.15 billion run-rate): ~0.99x, down from Q1 2022's ~1.38x and continuing the same direction of travel - the multiple has now roughly halved in two quarters even as the underlying business posted its best quarter yet on an Adjusted EBITDA basis. The market's repricing has clearly diverged from the operating trend this year.
• Using trailing-twelve-month revenue (Q3 2021 through Q2 2022) of approximately $19.9 billion instead of the annualized quarterly run-rate, EV/Revenue works out to essentially the same ~1.0x - the two methods converge this quarter because growth has slowed enough that the annualized and trailing bases are close together.

DCF and reverse DCF (illustrative only): six quarters of public-company data now include one quarter distorted by a warehouse fire and one quarter (this one) benefiting from that fire's comparison rolling off, which still makes a clean multi-year cash flow projection thin. Directionally, checking this quarter against last quarter's base case:

Scenario	Assumption / outcome
Current (period-end close)	Actual EV/Revenue this quarter, for reference: 0.99x
Bear case	Bear case (from last quarter) assumed Product Commerce's breakeven might be a one-quarter blip and Developing Offerings' loss growth kept widening. Neither happened - Product Commerce stayed at Segment Adjusted EBITDA-positive (even accounting for the fire-comp caveat above) and Developing Offerings' loss narrowed rather than widened further. The bear case's operating assumptions did not play out this quarter.
Base case	Base case (from last quarter) predicted Developing Offerings' loss growth would decelerate from Q1's 47% widening pace and that the multiple would stabilize near 1.4-1.5x. Developing Offerings did better than decelerate - its loss reversed to a 57% narrowing - but the multiple did not stabilize; it kept compressing, to roughly 0.99x. The operating half of the base case held or beat expectations; the valuation half did not.
Bull case	Bull case (from last quarter) called for Product Commerce margin expansion, Developing Offerings' loss flattening, and multiples recovering toward 2.0-2.5x. The operating half is arguably ahead of this case - both segments improved - but the multiple moved in the opposite direction from what the bull case required.
Revised base case for next quarter	Revised base case for next quarter: if Developing Offerings' improvement and Product Commerce's segment profitability both hold without the fire-comp tailwind available for comparison, and the multiple stabilizes near this quarter's ~1.0x rather than continuing to compress, that would be the first quarter since the IPO where operating results and market pricing move in the same direction rather than opposite ones.

---