SE Adjusted EBITDA Went Negative Again — This Time, It Wasn't an Accounting Illusion

The Illusion Is Gone. The Number Is Still Negative.

Q2 2021 broke Sea's four-quarter positive Total Adjusted EBITDA^» streak, but that swing traced almost entirely to a shrinking Garena deferred-revenue comp against an unusually easy Q2 2020 — the operating loss actually narrowed that quarter. The third quarter breaks the streak again, more deeply: Total Adjusted EBITDA fell to $(165.5) million from +$120.4 million a year earlier. This time there's no comp trick to explain it away. Operating loss widened 50.1% year-over-year, to $(458.6) million from $(305.5) million — the underlying business lost more money, not less, and the entire framing from last quarter's post doesn't apply here.

The driver is straightforward once the segment reconciliation table is opened: E-commerce's Adjusted EBITDA loss more than doubled, to $(683.8) million from $(301.6) million, as Shopee scaled marketing spend alongside order growth. Digital Entertainment's Adjusted EBITDA still grew a healthy 22.3% to $715.1 million, but its own growth engine is decelerating hard — bookings growth fell to 29.2% year-over-year from 64.8% in Q2, and quarterly active user growth fell to 27.4% from 45.1%. A full year past the initial pandemic surge, Garena's lockdown-era growth rates are normalizing, even as one of Sea's own markets (Vietnam, per management's own framing on the call) was in its strictest lockdown of the entire pandemic this quarter — the COVID tailwind that inflated 2020's comps is fading for Garena specifically, not because engagement is weakening but because the base it's growing against is no longer an easy one.

Total GAAP revenue reached $2.7 billion, up 121.8% year-over-year, and gross profit grew 147.5% to $1.0 billion — both still comfortably triple-digit, continuing the growth this series has tracked every quarter since Q3 2019. Sea raised its full-year 2021 e-commerce GAAP revenue guidance for the second time this year, to $5.0-5.2 billion (135.3% growth at the midpoint, up from the previous $4.7-4.9 billion) — a guidance raise announced in the same release as the segment's Adjusted EBITDA loss more than doubling, which is the tension this quarter's numbers actually sit in: growth accelerating on the top line, unit economics deteriorating on the bottom line, in the same business.

The Prescription

Sea should stop treating "prudent and sustainable growth" as a phrase that survives a quarter where the metric it's meant to describe — E-commerce's Adjusted EBITDA loss — more than doubled year-over-year while the per-order loss figure stayed merely "in line" rather than improving as it had in Q2 2021 (down 19.6% year-over-year that quarter). If the strategy really is disciplined investment rather than a scramble to defend market share against Amazon's and Lazada's regional pushes, the next earnings call should show per-order economics resuming their improvement, not just holding flat while the absolute dollar loss balloons with volume. Growing losses in line with order growth isn't the same as growing efficiently — the distinction matters, and this quarter's numbers don't yet show which one Sea is actually doing.

What Sea should stop doing: describing SeaMoney's losses purely as the cost of "continuing efforts to drive mobile wallet adoption," the same phrase used every quarter since Q3 2020 regardless of whether the loss narrows or widens (see SeaMoney's stalled inflection below) — three quarters after the one quarter it actually improved, the language hasn't updated to reflect that the improvement didn't stick.

Key Financial Metrics

Q3 2021 vs. Q3 2020, all figures in USD as reported (Sea reports in US dollars)

Metric	Q3 2021	Q3 2020	YoY
Total GAAP revenue	$2,688.9M	$1,212.2M	✅ +121.8%
Total gross profit	$1,008.7M	$407.6M	✅ +147.5%
Operating loss	-$458.6M	-$305.5M	🔴 Loss ~50.1% wider
Total Adjusted EBITDA»	-$165.5M	+$120.4M	🔴 Swung sharply negative
Net loss	-$571.0M	-$425.3M	🔴 Loss ~34.3% wider
Net loss excl. share-based compensation	-$448.0M	-$346.0M	⚠️ Loss ~29.5% wider
Basic and diluted loss per share (excl. SBC)	-$0.84	-$0.69	⚠️ Loss ~21.7% wider per share
Total Sea Limited shareholders' equity	$7,787.3M	n/a (Dec 2020: $3,382.9M)	✅ +130.2% vs Dec 2020
Cash and cash equivalents	$11,126.2M	n/a (Dec 2020: $6,166.9M)	✅ +80.4% vs Dec 2020
Net cash from operating activities (9 months)	+$513.5M	+$370.4M	✅ +38.6%

Free cash flow isn't included above, for the same reason as every prior quarter in this series: this quarter's disclosure doesn't break out a standalone capex figure, only a combined $(1,991.8) million nine-month net cash used in investing activities that mixes capital expenditure with short-term investment purchases and SeaMoney's growing loan book. The nine-month operating cash flow figure is worth reading quarter by quarter rather than as one number: Q1 2021 alone was +$318.3 million, Q2 2021 alone was roughly +$132.4 million, and this quarter's implied standalone figure — the nine-month total of $513.5 million minus the six-month total already disclosed — is roughly +$62.8 million. That's a third straight quarter of a shrinking operating cash contribution, even as the headline cash balance more than doubled on a huge financing inflow (see Beyond the Usual below). Both equity and cash swelling by more than 80% in nine months is a balance-sheet story, not an operating one — worth keeping separate from how much cash the actual business is throwing off, which is decelerating.

Three Segments, One Real Reversal and One Real Deceleration

Sea again reports three segments plus Other Services and unallocated corporate expenses, using the bookings-for-Digital-Entertainment / GAAP-revenue-for-everything-else metric set established in Q3 2020.

Digital Entertainment (Garena)

Bookings reached $1.2 billion, up 29.2% year-over-year — a sharp deceleration from Q2 2021's 64.8% growth — and Adjusted EBITDA grew 22.3% to $715.1 million, representing 58.6% of bookings, down from 61.9% a year earlier. Quarterly active users reached 729.0 million (+27.4% YoY, down from +45.1% in Q2) and quarterly paying users hit 93.2 million (+42.7% YoY, down from +84.8% in Q2), pushing the paying-user ratio to 12.8% from 11.4%. Every growth-rate comparison in this segment decelerated this quarter — bookings, users, and paying users alike — a full year past the initial pandemic surge that inflated 2020's comps, and consistent with a genuine normalization of gaming demand rather than anything company-specific. Free Fire launched a standalone "Free Fire MAX" version globally in late September, celebrated its fourth anniversary, and remained the top-grossing mobile game in Southeast Asia and Latin America for a ninth consecutive quarter and in India for a fourth. Operating income for the segment more than doubled to $611.7 million from $278.6 million, even as the deferred-revenue add-back that inflated last year's comp fell to $93.7 million from $299.2 million — Digital Entertainment's underlying profitability is genuinely growing, just off a slower top-line growth rate than earlier in the pandemic.

E-commerce (Shopee)

Gross orders reached 1.7 billion, up 123.2% year-over-year — a ninth consecutive quarter of triple-digit order growth — and GMV hit $16.8 billion, up 80.6%, both decelerating only modestly from Q2 2021's 127.4%/87.5% growth. GAAP revenue grew 134.4% to $1,450.4 million, split between roughly $1.2 billion of marketplace revenue (+151.4% YoY) and $279.6 million of product revenue (+82.2% YoY) — Sea's release states the marketplace figure only to the nearest $0.1 billion, not to the dollar — revenue as a share of GMV improved to 8.6% from 6.7% a year earlier, continuing the deepening-monetization trend. The unit-economics story stalled, though: Adjusted EBITDA loss per order was $0.41, described by management on the call as merely "in line with" Q3 2020 — a break from Q2 2021's 19.6% year-over-year improvement in the same metric. With flat per-order economics against 123.2% more orders, the total Adjusted EBITDA loss more than doubled to $(683.8) million from $(301.6) million, a 126.7% widening. Management noted per-order losses improved both year-over-year and quarter-over-quarter in Southeast Asia/Taiwan and in Shopee's other markets separately — a detail that doesn't reconcile cleanly with the flat combined figure and points to mix shift (newer, less efficient markets like the just-launched Poland, France, Spain, and India growing faster) rather than a company-wide efficiency stall.

Digital Financial Services (SeaMoney)

GAAP revenue reached $132.2 million, up 818% from $14.4 million a year earlier — again the fastest-growing segment by percentage, off the smallest base, and now large enough in absolute dollars to be meaningfully aided by SeaMoney's growing on-balance-sheet loan book (see Beyond the Usual). Mobile wallet total payment volume^» reached $4.6 billion, up 111% year-over-year, and quarterly paying users for the mobile wallet grew to 39.3 million (+120% YoY). Adjusted EBITDA loss was $(159.0) million, up 6.5% year-over-year from $(149.3) million — and, three quarters after Q1 2021's first-ever sequential narrowing, the loss keeps drifting slightly wider each quarter rather than resuming that trend: $(153.1) million in Q1, $(155.0) million in Q2, now $(159.0) million in Q3. Management again attributed the loss simply to "continued efforts to drive mobile wallet adoption," unchanged language since Q3 2020. Beyond the wallet, management flagged early-stage initiatives in buy-now-pay-later, digital banking, and insurtech — all described as still in early development, with no metrics disclosed.

Segment Comparison

Segment	Q3 2021 Revenue/Bookings	Q3 2020	YoY	Q3 2021 Adj. EBITDA	Q2 2021 Adj. EBITDA (QoQ ref)	Key Operating Metric
Digital Entertainment (Garena)	$1.2B bookings	$0.9B (implied by +29.2% YoY)	✅ +29.2% (down from +64.8% in Q2)	✅ $715.1M (58.6% margin)	$740.9M	QAU 729.0M (+27.4% YoY), QPU 93.2M (+42.7% YoY)
E-commerce (Shopee)	$1,450.4M GAAP revenue	$618.7M	✅ +134.4%	🔴 -$683.8M (loss ~126.7% wider YoY)	-$579.8M	GMV $16.8B (+80.6% YoY), 1.7B gross orders (+123.2% YoY)
Digital Financial Services (SeaMoney)	$132.2M GAAP revenue	$14.4M	✅ +818%	⚠️ -$159.0M (loss ~6.5% wider YoY, ~2.6% wider QoQ)	-$155.0M	TPV $4.6B (+111% YoY), QPU 39.3M for mobile wallet
Other Services	$6.9M GAAP revenue	$10.1M	🔴 -31.4%	🔴 -$31.9M (loss ~250.2% wider YoY)	-$23.3M	—
Unallocated expenses	—	—	—	-$5.8M	-$7.0M	—
Total	$2,688.9M GAAP revenue	$1,212.2M	✅ +121.8%	🔴 -$165.5M (from +$120.4M)	-$24.1M	—

Garena's $715.1 million of Adjusted EBITDA no longer comes close to covering everyone else's combined losses ($(880.6) million across E-commerce, SeaMoney, Other Services, and unallocated expenses) — Q2 2021 already showed the gap turning slightly negative at the group level; this quarter the gap between Garena's profit and everyone else's losses widened to $(165.5) million, roughly seven times larger than Q2's shortfall. Every non-Garena segment's loss widened both year-over-year and quarter-over-quarter this quarter — there's no offsetting improvement anywhere else in the business to lean on.

Trailing Quarters: Total GAAP Revenue and Total Adjusted EBITDA

Quarter	Total GAAP Revenue	Total Adjusted EBITDA
Q4 2019	$777.2M	-$104.9M
Q1 2020	$714.9M	-$69.9M
Q2 2020	$882.0M	+$7.7M
Q3 2020	$1,212.2M	+$120.4M
Q4 2020	$1,566.6M	+$48.7M
Q1 2021	$1,763.6M	+$88.1M
Q2 2021	$2,280.5M	-$24.1M
Q3 2021	$2,688.9M	-$165.5M

Revenue has now grown for ten straight quarters without interruption, and $2,688.9 million is again the largest quarterly figure in the series. Total Adjusted EBITDA tells a different story: after four consecutive positive quarters through Q1 2021, this is now the second straight negative quarter, and the second one is far deeper than the first. Summing the trailing four quarters (Q4 2020 through Q3 2021) puts trailing-twelve-month Total Adjusted EBITDA at roughly $(52.7) million — negative for the first time since the positive streak began in Q3 2020, when the trailing-twelve-month figure last turned positive. There's no structural Q3 seasonality specific to Sea's business the way there is around Q4 shopping festivals (Shopee's 11.11 sale falls in Q4, not Q3), so this isn't a seasonal read — it's a genuine two-quarter shift in the trend this table has tracked since the series began.

What Management Chose to Emphasize on the Call

Forrest Li's prepared remarks opened with "another quarter of strong results across all our businesses" and spent a substantial share of airtime on community and SME-support initiatives — Shopee Export MSME Campuses in Indonesia, rural-seller digitization with Malaysia's government, a young-entrepreneur program with Thailand's Digital Economy Promotion Agency, and passing one million registered local sellers in Brazil — continuing the pattern from Q2 2021's call. He also announced Chris Feng's appointment as Group President effective January 1, 2022, continuing to run Shopee and SeaMoney while taking on cross-business strategic coordination — a leadership change worth tracking in future quarters given it consolidates operating authority over the two segments now driving the group's losses. On e-commerce specifically, Forrest's framing was that Shopee's investment is happening "in a prudent and sustainable manner," and CFO Tony Hou echoed that the E-commerce Adjusted EBITDA loss reflected the company's continuing "investments to fully capture the opportunities in our markets" — neither executive acknowledged that the loss had more than doubled year-over-year, or that the per-order economics that had been the evidence for "prudent" spending in earlier quarters had stopped improving (see The Prescription above). As in Q2 2021, Tony Hou walked through every segment's Adjusted EBITDA individually by name — Garena's $715 million, E-commerce's $684 million loss, SeaMoney's $159 million loss — but at no point in either executive's prepared remarks was the consolidated Total Adjusted EBITDA figure mentioned, the second straight quarter this specific number has gone unmentioned while every component of it gets named. The filed transcript again contains only prepared remarks and ends before the Q&A that followed.

Beyond the Usual

Management skipped the one number that reversed, for the second straight quarter

Q2 2021 flagged that CFO Tony Hou named every segment's Adjusted EBITDA individually on the call while never mentioning the consolidated Total Adjusted EBITDA figure that had just swung negative. The same pattern repeats this quarter: Garena's $715 million, E-commerce's $684 million loss, and SeaMoney's $159 million loss are each stated by name, but the group figure — $(165.5) million, now roughly seven times deeper than Q2's swing and part of the first negative trailing-twelve-month reading since the positive streak began — is never spoken in either executive's prepared remarks. Last quarter this series noted the swing was largely a comp-driven accounting effect management could have explained but didn't; this quarter the swing is a real operating deterioration (operating loss up 50.1% year-over-year), which makes the second consecutive omission a harder pattern to read as coincidental.

SeaMoney's loss-narrowing inflection is now three quarters gone

Q1 2021 flagged the first sequential narrowing in SeaMoney's Adjusted EBITDA loss after eight straight quarters of widening, to $(153.1) million from Q4 2020's $(171.3) million. Q2 2021 found that improvement had stalled at essentially flat, $(155.0) million. This quarter the loss widened again to $(159.0) million — a small but real move in the wrong direction for a third straight quarter since the one-quarter inflection. Three quarters is long enough to say the Q1 improvement looks like a one-off rather than the start of a trend, without yet calling the trajectory a clear deterioration either.

Goodwill's balance essentially stopped changing this quarter — $476.662 million at September 30, 2021 versus $476.3 million at June 30, 2021, an increase of roughly $0.4 million, effectively flat. This closes out the pattern this series has tracked since Q1 2020's first unexplained jump: a 116% jump in Q1 2021, a sharp deceleration to 1.8% in Q2 2021, and now no meaningful movement at all. No acquisition has ever been named for any of these increases in a quarterly source document (the FY2020 20-F later named three unnamed acquisitions behind the original 2020 jump), but the balance has now been flat for a full quarter, so there's nothing further to track here going forward absent a new jump.

Total Sea Limited shareholders' equity more than doubled over nine months, to $7,787.3 million at September 30, 2021 from $3,382.9 million at December 31, 2020 (+130.2%), and cash and cash equivalents rose similarly, to $11,126.2 million from $6,166.9 million (+80.4%) — both driven predominantly by financing activities, which contributed $7,015.6 million of cash over the nine-month period versus $784.0 million in the same period a year earlier. Additional paid-in capital grew by $5,874.4 million over the same nine months, from $8,526.6 million to $14,401.0 million. The filed statements disclose only $196.0 million of convertible-note conversions between July 31 and October 31, 2021 — nowhere near large enough to explain a jump this size — so the bulk of the paid-in-capital increase reflects capital-raising activity earlier in the nine-month window that isn't itemized in this quarter's own release. Loans receivable (current) more than doubled again, to $1,049.7 million from $285.9 million at year-end, as SeaMoney's on-balance-sheet lending book — first disclosed at FY2020 year-end — keeps scaling faster than almost any other balance-sheet line. As in every prior quarter, related-party balances (amounts due from related parties: $10.3 million, down from $19.4 million at year-end; amounts due to related parties: $66.5 million, up from $42.6 million) are disclosed only as two aggregate lines with no breakdown of counterparties or transactions.

No 10-Q-equivalent quarterly filing with footnotes was located for this quarter — as a foreign private issuer, Sea files detailed footnoted financials only annually, in its Form 20-F, so the deeper disclosures (off-balance-sheet arrangements, lease schedules, related-party detail, purchase commitments) that would normally round out this section aren't available until that annual filing.

Target Valuation Range

No numeric fair-value range is computable this quarter: trailing-twelve-month Total Adjusted EBITDA turned negative, at $(52.7) million, so the EV/EBITDA anchor this series has used every quarter since Q3 2020 is gone, against an implied market cap of roughly $176.8 billion. Price-to-sales (≈21.3x) and price-to-book (≈22.7x) are still computable and both eased from Q2's stretched levels, but the P/B improvement is balance-sheet noise from a huge financing inflow, not a real reassessment, and neither is a substitute for the EV/EBITDA read that's no longer available.

Sea's ADS closed the quarter at $318.73 on September 30, 2021, up 16.1% from $274.60 at the end of Q2 2021, continuing the same steady climb this series has tracked since Q3 2020. No stock split has occurred at any point in Sea's history through this quarter, so this and all prior prices in this series remain directly comparable on a nominal basis.

Market cap → enterprise value	Q3 2021
Share price (period-end)	$318.73
Shares outstanding (per October 31, 2021 disclosure)	554,625,132
Market capitalization	~$176.8 billion
Plus: non-current convertible notes (debt)	$3,503.5 million
Less: cash and equivalents	$11,126.2 million
Enterprise value	~$169.2 billion

Peer-multiple sanity check	Q2 2021 (TTM)	Q3 2021 (TTM)
TTM GAAP revenue	$6,822.9M	$8,299.6M
Price-to-sales	21.6x	21.3x
Price-to-book	41.5x	22.7x
TTM Total Adjusted EBITDA	$233.1M	$(52.7)M
EV/EBITDA	~619x	uncomputable (negative EBITDA)

The P/B improvement is a balance-sheet-driven denominator change — equity nearly doubled this quarter (see Beyond the Usual above) while the share price rose only 16.1% — not the market reassessing the business more favorably. Trailing-twelve-month Total Adjusted EBITDA turned negative (Q4 2020's +$48.7M, Q1 2021's +$88.1M, Q2 2021's -$24.1M, and this quarter's -$165.5M), so an EV/EBITDA multiple can no longer be computed at all — the metric this series has used every quarter since Q3 2020 to sanity-check the valuation has broken. A real DCF still isn't attempted here for the same reason as every prior quarter: Digital Entertainment's growth is now genuinely decelerating rather than merely guided to decelerate, and Shopee and SeaMoney's unit economics, which had been improving on a per-unit basis, stalled or reversed this quarter — not yet a multi-year record a defensible intrinsic-value range can be built on. With the EV/EBITDA anchor gone and P/S and P/B both reflecting balance-sheet noise from this quarter's huge financing inflow rather than operating performance, this quarter's data doesn't support even the peer-multiple sanity check as cleanly as prior quarters did — a genuine read on this valuation will need at least one more quarter of the EBITDA trend to actually resolve one way or the other.

---