SE — Deep-Dive Research v2.0 | Liquid Wheel Research

Liquid Wheel Research · Deep-Dive Framework · v2.0 quarterly refresh

🌊

Sea Limited: the growth held, the discipline didn't

Sales still grow almost 50% a year. Profit grew about 11%. Management opened a new spending venture with no budget, insiders sold about $344M, and the price fell through v1's exit line. The stock is cheap now. Whether it is cheap and safe gets decided at the 2026-11-10 earnings report.

⚠ IMPAIRED · ON WATCH v1 EXIT RULE FIRED · HONORED CONVICTION-SPEC · HALF OF v1 SIZE NO STOCK BUYS BEFORE 2026-11-10

6.0

/ 10 overall

v1: 7.5

Data as of close 2026-10-26 ($74.01) · Filings through Q2 2026 (6-K, filed 2026-08-11) · Report 2026-10-27 · Prior version v1.0, 2026-07-12 ($111.14) · Next earnings 2026-11-10 (date from Yahoo Finance, not yet confirmed by Sea)

Data as of 2026-10-26 close.

In plain English

Sea's businesses still grow fast, but profit is not keeping up. Management has started spending on a new venture. Insiders sold stock. The price fell through the exit line we wrote in July, and we follow our own rule. The stock is now cheap. Whether it is cheap and safe gets decided at the 2026-11-10 earnings report, not before.

Price

$74.01

−33.4% since v1 · −43.7% from $131.51 (2026-08-11) · −64.1% from the $199.30 peak

3.5% above the 52-week low ($71.53)

Revenue growth

+48.1%

Q2 2026 $7,787.8M vs a year ago · no acquisitions

2026 Wall Street estimate +36.0%

Net income growth

+10.6%

$458.1M · net margin 5.9% [Q2 2025: 7.9%]

Profit grew at about a fifth of the sales rate

Price-to-sales

1.64x

v1: 2.70x · second-lowest in the peer group

You pay $1.64 for each $1 of yearly sales

Net cash

+$9.63B

$11.86B cash and investments vs $2.22B debt excluding leases

About $401M of the $1B buyback left

Checklist

9 / 3 / 4

Pass / Watch / Fail · v1: 11 / 2 / 3

Insider gate fell to FAIL; efficiency to WATCH

Read in 30 secondsThe five things that matter, in plain English

Growth is fine+48.1%Q2 2026 revenue $7,787.8M. Shopee's gross merchandise value (GMV), the total value of goods sold on Shopee, was $38.3B, +28.4%. Loans 90+ days late were 1.0% of Monee's book. The businesses customers use still grow faster than almost any large company.

Profit conversion is not+10.6%Net income and adjusted EBITDA (roughly cash operating profit) both grew +10.6%. The tax rate jumped to 35.0% [Q2 2025: 25.2%]. Loan-loss charges rose +71.5%. Sales and marketing rose +64.5%. Each new dollar of sales brings home less profit than last year.

The plan changed on 2026-10-20−9.1% that dayLi's 9th-anniversary email launched Sea Cloud, an AI cloud for outside customers built by about 500 staff over 2 years. It floated a Monee listing: the fintech arm "may well ring that bell." No budget, timeline or deal structure. Volume ran 4–4.4x normal.

Insiders sold≈$344M · 0 buys9 insiders since v1. Li sold 1,139,734 shares ($147.3M) from 2026-08-11 to 2026-08-28 at $120–132, under a pre-arranged plan adopted 2025-09-10. He has sold nothing since.

Cheap, and a plan1.64x sales · 14.5x14.5x the 2027 earnings estimate. Re-run targets: base $109 (+47%), bear $46 (−38%), bull $203 (+174%). No stock buys before 2026-11-10. Before the print: cash-secured puts only, at the $65 and $60 December strikes. A $70 put only if it expires before the print (Nov-06).

Item	Call
Score	6.0/10 [v1: 7.5]
Growth thesis	INTACT
"Self-funded, founder-aligned" capital-allocation thesis	BROKEN
v1 price exit rule	FIRED. Honored.
Status	IMPAIRED, ON WATCH. No stock purchases before the 2026-11-10 print.
Tier	Conviction-spec. Moat 7/10 [v1: 7.5–8], below the 9/10 bar for full-size (Tier-1) positions. The new position is half of v1's size.
Checklist	9 PASS / 3 WATCH / 4 FAIL [v1: 11 / 2 / 3]
Macro	YELLOW. Rates near 5% weigh on the price investors will pay. A headwind, not a veto.

The one decision: what does "weekly close below $75" mean?

v1 said two different things about the same line. The entry-ladder section said a weekly close below $75 meant "stop adding, reassess." The breakers section said "exit, don't average." The stock closed the week of 2026-10-23 at $74.63, so the line was crossed.

PM decision: the breakers list is binding. We honor the exit. A position built under the v1 plan is exited. It is not averaged down.

Why: the rule was written before money was at risk. That is the only time a rule is worth anything. Two v1 non-price breakers had also fired before the price broke: founder Forrest Li sold stock, and sales-and-marketing spending grew faster than revenue for two quarters while Shopee's growth slowed.
The honest cost: this exit sells into a very oversold reading (daily Relative Strength Index of 17.7), 10 trading days before a print that could jump either way. If Q3 is strong and the stock jumps, honoring the rule costs that bounce. That is the price of having rules.
How we get back in: the v2 plan on the Trade Plan tab is a new, smaller position with its own rules. Before the print it uses only cash-secured puts at strikes 12–25% below today's price. After the print it buys stock only if the print passes the bar on the Management tab.
v2 fixes the ambiguity. Every rule now names its action: "exit," or "stop adding." Price lines stop adding. Business breakers trigger exits.

Why the score moved from 7.5 to 6.0

Change	Points	Reason
Insider gate: PASS → FAIL	−0.50	About $344M sold from 2026-07-10 to 2026-10-23, including Li's $147.3M. Zero open-market buys.
Capital-allocation regime change	−0.50	The 2026-10-20 memo announced Sea Cloud (no budget) and a possible listing of Monee, the fintech arm.
Profit is not keeping pace with sales	−0.50	Net income +10.6% on revenue +48.1%. Net margin 5.9% is below the 6.5% v1 used in its bear case.
TikTok keeps taking share	−0.25	Cube data: TikTok Shop went from 27% to 39% of the big-three Southeast Asian platforms in one year.
Founder's stake was overstated	−0.25	Li's own economic stake is about $3.5B, not v1's $10.96B (Management tab).
Operating-efficiency gate: PASS → WATCH	−0.25	Operating margin 8.35% vs 9.27% a year earlier.
Cheaper valuation	+0.50	Price-to-sales 1.64x [v1: 2.70x]. Next-year price-to-earnings 14.5x [v1: 27.0x].
Credit is steady	+0.25	Late loans 1.0%. Loss reserves are growing in line with the loan book.
Net	−1.50	7.5 → 6.0

the business got slightly better on growth and credit, and the stock got much cheaper. But the reasons v1 trusted management with a premium (insiders acting like owners, every dollar reinvested in the core) got worse. That is what cost the 1.5 points.

Company 101: what Sea actually is

Sea is a Singapore company listed in New York. It started in games, used game profits to build an online mall, and used the mall's data to build a lender. In October it added a fourth idea: renting out its own computer infrastructure as an AI cloud. The 8-year-old version: a phone mall, delivery bikes, a piggy bank and a game. Now also a rented-out computer room, which makes the story less simple.

Four businesses (three in v1)

Shopee: online marketplace plus its own delivery network (SPX Express). Q2 revenue $5,587.6M including goods Shopee sells itself, +48.2%.
Monee: consumer loans, digital banks and wallets (formerly SeaMoney). Q2 revenue $1,402.8M, +58.9%. Loan book $11.1B.
Garena: game publisher. Flagship is Free Fire. Q2 bookings (money players spent) $763.5M, +15.5%.
Sea Cloud (NEW, 2026-10-20): the internal cloud, rebuilt over 2 years by about 500 staff, now offered "to the world." No revenue, no budget and no timeline disclosed. A Dow Jones Market Talk item (2026-10-22, headline only) says it could take 3–4 years to break even UNVERIFIED.

Market size (total addressable market)

Southeast Asian online shopping: in the first half of 2026, Cube puts Shopee at 55%, TikTok Shop at 39% and Lazada at 6% of the big three. Indonesia is about 40% of Shopee's Southeast Asian GMV ($21.3B in the first half).
Derived size: Shopee's Southeast Asian GMV is about $53B a half-year, so the three platforms together move roughly $95–97B a half-year, or about $190B a year. Momentum Works puts TikTok Shop at $34.7B in the first half of 2026 (+80%), consistent with that math UNVERIFIED: cross-source estimate.
Brazil: Shopee had about 8.8% share vs Mercado Libre's 12.5% (v1; not updated).
Consumer credit: Monee has more than 40M active credit users (+34%) and a $1B+ book in the Philippines alone.
Cloud: no sizing in the filings. Sea Cloud enters against Amazon Web Services and Alibaba Cloud.
Runway: online-shopping penetration figures are not sourced in this refresh. Carried as an open item.

the core market is big and still growing about 25–30% a year, but it is now clearly a two-horse race. Sea Cloud is a bet in a market where Sea has no customers yet and faces the world's largest competitors.

Timeline (carried from v1; new items marked NEW)

Date	Milestone
2009	Founded as Garena in Singapore by Forrest Li, Gang Ye and David Chen.
2014	SeaMoney begins. Chris Feng joins from Rocket Internet/Lazada.
2015	Feng launches Shopee: mobile-first, with local teams in each country.
2017	Renamed Sea Limited. New York listing in October. Free Fire is built in-house for cheap Android phones.
~2019	Shopee overtakes Alibaba-backed Lazada as Southeast Asia's #1.
Oct 2021	Stock peaks near $366 (all-time high $372.70 on 2021-10-19).
Jan 2022	Tencent cuts its stake. Li consolidates control.
Feb 2022	Free Fire banned in India. Still not relaunched.
Sept 2022	The memo. Leadership takes $0 cash pay "until self-sufficiency." About 7,000 layoffs. Exits from France, India, Spain and Argentina. Stock about $35, −90%.
Q4 2022	First profitable quarter. Self-funding since.
2023–24	TikTok Shop forces live-video shopping. Shopee builds its own. SPX passes 50% of Shopee's orders.
May 2025	SeaMoney renamed Monee.
Sept 2025	Stock peaks at $199.30 (2025-09-11). NEW: Li's pre-arranged selling plan adopted on 2025-09-10. v1 missed this.
Nov 2025	$1B buyback announced.
2026-03-03	Q4 2025 earnings day. −16.5% on the day, then a low of $77.05 on 34.8M shares (7.2x average volume). CORRECTED: v1 called this unattributed.
2026-03-18	First insider-trade report (Form 4) under the new US reporting regime for foreign issuers. Earlier "no insider sales" findings reflect a disclosure gap, not an absence of selling UNVERIFIED: start date inferred from filings.
2026-05-12	Q1 2026. Revenue +46.6%. First quarter with adjusted EBITDA above $1B. Stock +13.1%.
2026-07-10	v1 as-of date, $111.14.
July 2026 NEW	MariBank Philippines gets a digital-bank licence (BigGo/DBS; not filed by Sea).
2026-08-11 NEW	Q2 2026 print. Stock +14.6% to $131.51 (high $131.94). Li's sales begin the same day.
2026-08-25/26 NEW	Daily "golden cross," the 50-day average crossing above the 200-day CONFLICT: sources disagree by one day.
2026-09-15 NEW	$996.3M of 0.25% convertible notes mature. Repayment in cash is presumed UNVERIFIED until the Q3 filing.
2026-09-16 NEW	Annual meeting. No proposals.
2026-10-01/02 NEW	DBS downgrades Buy → Hold, target $148 → $105 CONFLICT on the date.
2026-10-16 NEW	MSCI warns it may drop Indonesia from emerging-market indexes. Jakarta stocks −11.6% in a day. SE −3.83%.
2026-10-20 NEW	Li's anniversary memo: Sea Cloud plus a possible Monee listing. SE −9.1%.
2026-10-23 NEW	Weekly close $74.63. The v1 exit rule fires.
2026-10-26 NEW	New 52-week low of $71.53 intraday. Close $74.01.

The Flywheel: 5 steps, updated

In plain English

A flywheel is a loop where each step makes the next one stronger. Sea's loop still turns. This quarter two leaks appeared: profits flowing out to Sea Cloud, and a possible split of the lender from the mall.

Garena prints cash and users

Q2: 666.3M quarterly active users (flat vs a year ago) and 68.1M paying users (+10.2%). Bookings $763.5M (+15.5%). Adjusted EBITDA $429.8M = 56.3% of bookings [v1, Q1: 61.6% of $931.4M]. Weakest link: Free Fire is a "significant majority" of Garena's revenue and profit (annual report, 20-F). v1's 65–70% share is an analyst estimate.

More buyers → more sellers → more GMV on Shopee

GMV $38.3B (+28.4%; Q1 +30.2%). Orders 4.2B (+27.5%).

Order density feeds SPX, the in-house delivery network

Instant-delivery orders +~80%. Cost per order −~20% (Q2 call, per the Motley Fool summary; the verbatim transcript was not available). v1's −6% Asia / −21% Brazil figures used a different basis and were not updated. TikTok Shop rents its delivery. This is still Shopee's structural cost edge.

Checkout data plus the wallet feed Monee's lending

Loans $11.1B (+62.5%): $10.0B on Sea's balance sheet, $1.1B placed with partners. More than 40M credit users. Loans 90+ days late: 1.0%. NEW RISK: a Monee listing could weaken the data-and-funding link between Shopee and Monee.

Profits recycle

Take rate (Shopee revenue ÷ GMV) up to 14.59% from 12.64%, about +195 basis points (1.95 percentage points). Ad take rate +90 basis points. Profits fund delivery density, loan growth and Brazil. NEW LEAK: they now also fund Sea Cloud, which does not feed steps 1–4.

Sea Cloud breaks the "every dollar makes the loop stronger" logic. Until management shows Sea Cloud's budget and returns, part of each profit dollar leaves the flywheel.

What Matters: the make-or-break questions

In plain English

Six questions decide this stock. Two got worse this quarter: competition and capital allocation. One got better: credit quality. One was partly answered: the margin question.

Can TikTok Shop do to Shopee what Shopee did to Lazada? WORSE Still the question that decides the tier. Cube: Shopee fell from 61% to 55% and TikTok rose from 27% to 39% between the first half of 2025 and the first half of 2026. Lazada fell from 12% to 6%. Thailand: TikTok holds 48% of combined Q2 GMV, so Thailand may be the first major market where Shopee loses its majority. Vietnam: Shopee 53% vs TikTok 44% (Q&Me, 12 months to April 2026). Counter-evidence: Shopee raised its take rate about 195 basis points while losing share.
Shopee is giving up share to TikTok, but customers still pay Shopee more per sale. That is what the profitable half of a duopoly looks like. It is not yet a collapse.
Does Monee's loan book survive a downturn? BETTER Better on the main signal, worse on mix. Loss reserves grew in line with the book: reserve growth ÷ on-book loan growth ≈ 0.98–1.0x [v1, Q1: 2.7x]. But the annualized loan-loss charge rose to about 23.6% of average on-book loans [Q1: ≈21.7%]. Charges are 39.5% of Monee revenue. The shift to Brazil and to credit used outside Shopee is deliberate, and management says it earns lower returns. New watch item: Indonesia plans a 40% debt-to-income cap on buy-now-pay-later loans from 2027 (DBS) UNVERIFIED regulation and start date.
credit is not cracking, but Monee is choosing riskier customers to keep growing.
Is the margin plateau reinvestment or erosion? LEANED EROSION v1 said Q2 would be the test. Net income +10.6% and adjusted EBITDA +10.6% on revenue +48.1%. Shopee adjusted EBITDA was 0.67% of GMV [Q2 2025: 0.76%]. Li's long-run target is 2–3%.
"we are investing for growth" is believable only if the profit-per-dollar line turns up. It has not yet.
How long does one game carry Garena? Bookings growth is slowing (+20.1% → +15.5%) and users are flat. Deferred revenue (cash collected from players but not yet counted as revenue) built only $16.9M in Q2 [Q1: $254.8M], so recent revenue partly comes from bookings made earlier. Two new titles are announced but not launched: Palworld Online and Monster Hunter Outlanders.
Does Indonesia hurt again? The risk shifted from rules to country risk. On 2026-10-16 MSCI warned it may cut Indonesia out of emerging-market indexes. The Jakarta index fell 11.6%, its worst day since 1998, and is down about 40% for the year. The rupiah was 18,070 per dollar vs a record low of 18,190. $4.9B of foreign money left Indonesian stocks this year. No new anti-Shopee rule was found. A competition-authority (KPPU) logistics complaint against Tokopedia is pending.
NEW: what will management do with the cash? Sea Cloud has no budget. Q2 spending on buildings and equipment (capex) jumped 2.6x vs Q1, to $356M, with no explanation. A Monee listing could sell part of the fastest-growing business at a low point for emerging-market fintech valuations: Nubank (NU) fell 16.95% on 2026-10-26, cause unverified. About $401M of buyback authority is left, roughly one quarter at the Q2 pace, and no new program has been announced.
this question now matters as much as TikTok. The Q3 call must answer it.

§2 · WHAT CHANGED SINCE v1 · data as of 2026-10-26

What Changed since v1: tracker grades first

In plain English

We graded every claim from July. The growth calls were right. The calls about management's discipline, the entry zones and the near-term price path were wrong. Several v1 "facts" turned out to be measurement errors, and they are corrected below. "[v1: X]" means the value in the July report.

Business claims

Mostly KEPT

Growth, take rate, credit, buyback

Price-path & entry claims

Mostly MISSED

All three buy zones filled within 23 trading days

Capital allocation & insiders

BROKEN

Li sold; Sea Cloud opened with no budget

Validators: signs the thesis is getting stronger

v1 claim	Grade	Evidence
Shopee gross merchandise value (GMV) growth ≥25% with take rate flat or up	KEPT	GMV +28.4%. Take rate (Shopee revenue ÷ GMV) 14.59% [Q2 2025: 12.64%].
Monee loans 90+ days late ≤1.5%	KEPT	1.0%, steady vs Q1.
Adjusted EBITDA above $1B per quarter, sustained	MISSED	Q2: $917.2M [Q1: $1,034.3M].
Shopee adjusted EBITDA ≥ the 2025 level	SUPERSEDED	Management replaced it with an explicit 2026 target of ≥$1B.
Buyback ≥$168M per quarter	KEPT	Q2: $416.8M (4.7M shares).
Q2 fund filings show stabilizing breadth	PARTIAL	418 funds added, 305 cut. The hedge-fund holder count is not available.
Garena bookings growing double digits	KEPT, slowing	+20.1% → +15.5%.
Brazil share climbing	PENDING	No share data. Brazil is 15–16% of Monee loans (DBS).

Breakers: signs the thesis is breaking

v1 claim	Grade	Evidence
Weekly close < $75	TRIGGERED	Week ending 2026-10-23 closed at $74.63.
200-week average lost on >2x average volume	TRIGGERED	$74.63 close vs 200-week average $93.87. Week volume 57.56M = 2.79x the 20-week average.
Daily "death cross" with price < $85	PENDING, imminent	About 5 sessions away (~2026-11-02).
Forrest Li selling stock	TRIGGERED 2026-08-11	1,139,734 shares, $147.3M.
Sales-and-marketing growth > revenue growth for 2+ quarters with GMV slowing	TRIGGERED (on the letter)	Q1: +52% vs +46.6%. Q2: +64.5% vs +48.1%. GMV growth slowed from 30.2% to 28.4%.
Loan-loss charges > Monee adjusted EBITDA	RULE WAS WRONG	Always true by construction (see the error list below). Replaced on the Trade Plan tab.
Loans 90+ days late above 2%	Not triggered	1.0%.
Shopee regional share < 50%, or a take-rate cut	Not triggered	Cube: 55%. The take rate rose.
Free Fire ban in a top-3 market	Not triggered	No evidence of a ban.
Jakarta rule hitting local platforms	Not triggered	A new Indonesia country-risk shock appeared instead (Red Team tab).

v1 price targets

v1 claim	Grade	Evidence
Bear case $52	PENDING	Price is $74, 42% above it.
Base case $124	MISSED (time path)	Its 9% net-margin input looks wrong vs 6.0% over the last four quarters.
Bull case $211	NO LONGER RELEVANT	Not reachable on these inputs.

v1 trade-plan calls

v1 claim	Grade	Evidence
No-chase line $116	CORRECT	The 2026-08-11 jump to $131.94 fully reversed.
Trim 1 at $123–127	CORRECT, HIT	2026-08-11 high $131.94, close $131.51. The best call in v1.
Buy zones $95–100 / $86–91 / $78–83	MISSED	All three filled within 23 trading days. A blended fill near $88.80 is −16.7% today.
Invalidation $75	FIRED	Weekly close 2026-10-23.
"Valuation compresses into the Jul 14 inflation report and the Jul 28–29 Fed meeting"	MISSED	The stock rallied to $131.51 by 2026-08-11.
"Q2 reinvestment print repeats the March selloff"	MISSED (timing)	Earnings day was +14.6%. The punishment came on 2026-10-20 from the memo instead.

Earnings-call promises

Promise	Grade	Evidence
2026 GMV ~+25%	ON TRACK	First half +29.3%.
Shopee 2026 adjusted EBITDA ≥$1B, second half > first half	PENDING	Needs ≥$521.4M in the second half.
Garena double-digit bookings growth	KEPT year to date	No full-year guide repeated on the Q2 call.
$1B buyback	ON PACE	About $401M left.
Free Fire India relaunch	BROKEN	Still not relaunched.
Q2 earnings "~mid-August"	KEPT	2026-08-11.

v1 understood the business but misjudged how much the market would pay for it, and how disciplined management would stay.

The Stage-12 two-question check

1 · Did the growth story or execution change?

Growth: no. How growth turns into profit, and how cash gets spent: yes.

Adjusted EBITDA and net income both grew +10.6% on revenue +48.1%.
The tax rate jumped to 35%.
Loan-loss charges rose +71.5% as Monee deliberately moved into riskier loans: SPayLater credit used outside Shopee, and Brazil. Management itself says this mix earns a lower return on assets.
Sea Cloud is a new place to spend cash.
A Monee listing is on the table, with no date. DBS sees nothing before 2027.

2 · Did we overlook something now revealed? Yes, five things

Li's pre-arranged selling plan was adopted on 2025-09-10. It already existed when v1 said "he sells none."
v1's 16.0% founder stake counts 43.2M shares Li only votes for other people (Management tab).
Shopee's value-added-services revenue (mostly logistics fees) fell 9.0%, because shipping subsidies are now subtracted from it. Part of Shopee's price-war spending hides inside net revenue.
Garena's deferred revenue (cash collected from players but not yet counted as revenue) barely grew: +$16.9M in Q2 vs +$254.8M in Q1 and +$102.2M in Q2 2025. Garena revenue grew +33.5% while bookings grew +15.5%, so revenue is being pulled forward from past bookings.
From Q3, Garena's adjusted EBITDA will exclude the deferred-revenue effect. The 2026-11-10 numbers will not compare cleanly with earlier quarters.

the growth engine is working. The question is whether management keeps feeding its profits back into Shopee and Monee, or sends them elsewhere.

v1 errors corrected in this version

The "8 beats" gate used mismatched numbers. v1 compared reported (GAAP, standard accounting) earnings per share with Wall Street's adjusted estimates. On a like-for-like adjusted basis, Q1 2026 was a beat ($0.81 vs $0.77) and Q2 a miss ($0.84 vs $0.86). The record is 2 beats and 6 misses over 8 quarters, not "7 straight misses." The gate still fails.
The 2026-03-03 "unattributed capitulation" was the Q4 2025 earnings day (filing 0001193125-26-088151). v1's own timeline said so.
The breaker "loan-loss charges > Monee adjusted EBITDA" could never work. Monee's adjusted EBITDA is measured after those charges. Replaced on the Trade Plan tab.
Scenario margins were too high, and scenario revenue too low. v1's bear assumed 6.5% net margin and $29.5B of 2027 revenue. Actual trailing margin is 6.0%, and trailing revenue is already $27.7B.
"$0 salary, so infinite skin in the game" was unsupported. The annual report (20-F) shows $3.13M of total cash pay for directors and executives as a group in 2025.
Buyback "$832M remaining" missed a $14.0M Q4 2025 purchase. The correct figure after Q1 was about $817.6M.
Take rate "13.7%" never stated its basis. It is total Shopee segment revenue (including goods Shopee sells itself) ÷ GMV. On that basis Q2 is 14.59%.
v1 skipped the framework's 10 priced downside scenarios (Stage 10). They are on the Red Team tab now.
Ownership was 16.1% / 57.7%. The 20-F shows 16.0% / 57.6%. Minor.
v1 compared GAAP scenario earnings with adjusted consensus. This version puts both on the same basis (Valuation tab).

none of these errors changes the direction of the call by itself. Together they made v1 look more certain than it was, especially on the founder's stake and the earnings record.

§3 · 16-GATE CHECKLIST · data as of 2026-10-26

The Checklist: 16 hard gates, re-graded (v1 → now)

In plain English

The framework's 16 yes/no tests. Two moved the wrong way this quarter: insider buying and selling, and operating efficiency. None moved up. Several passes have new caveats. Each card shows the July grade and today's grade.

Pass

9 v1: 11

Fail

4 v1: 3

Watch

3 v1: 2

PASS

1 · Growth 30–50%+

v1 PASS → now PASS

Q2 2026 revenue $7,787.8M, +48.1% vs Q2 2025. 2026 Wall Street estimate $31.19B, +36.0% (Q2 filing 2026-08-11; stockanalysis).

PASS*

2 · Free cash flow positive

v1 PASS* → now PASS* (caveat worse)

Q2 operating cash flow $1,506.0M minus spending on buildings and equipment $356M [Q1: $139M] = $1,150M. Last four quarters: $4,423M [v1: $4.67B]. *If you also subtract new loans Monee made (about $1,794M in Q2), cash left over is about −$644M [Q1: −$142M].

PASS

3 · Flywheel exists

v1 PASS → now PASS

Advertising take rate +90 basis points or more. Ad revenue +70% or more. Instant-delivery orders +~80% with cost per order −~20% (Q2 call). New leak: profits now also fund Sea Cloud, which sits outside the loop.

FAIL

4 · Moat ≥ 9/10

v1 FAIL (7.5–8) → now FAIL (7/10)

Vietnam: Shopee 61% → 53%, TikTok Shop 33% → 44% (Q&Me). Region: TikTok Shop 27% → 39% (Cube, first half 2025 → first half 2026). Thailand: TikTok at 48% of combined gross merchandise value (Cube).

PASS

5 · Founder-led

v1 PASS → now PASS

Forrest Li, founder and CEO since 2009, 57.6% of the votes.

PASS

6 · Skin in the game ≥ 5x

v1 PASS → now PASS (v1 overstated it)

Li's own economic stake is about 47.9M shares (7.8%) ≈ $3.5B at $74 [v1: $10.96B], plus 13.0M exercisable options. Even if the entire $3.13M group cash pay went to him, that is over 1,100x. The framework ideal is 1,000x.

FAIL

7 · Insider activity

v1 PASS → now FAIL ▼

9 insiders sold 2,908,286 shares (≈$344.0M) from 2026-07-10 to 2026-10-23, with zero open-market buys. Li alone: $147.3M.

WATCH

8 · Institutions

v1 WATCH → now WATCH

Q2 fund filings (13F): 418 added, 305 cut. JPMorgan +8.19M shares. Cutters: Tiger −18.5%, Sands −45.5%, D1 −58.6%. William Blair exited. Baillie Gifford 40.95M → 37.15M.

FAIL

9 · 8 straight earnings beats

v1 FAIL → now FAIL (basis corrected)

Like-for-like adjusted basis: 2 beats, 6 misses in 8 quarters. Q2: $0.84 vs $0.86 estimate. Q2 GAAP diluted earnings per share: $0.70.

PASS

10 · Explainable to an 8-year-old

v1 PASS → now PASS

A phone mall, delivery bikes, a piggy bank and a game. Now also a rented-out computer room (Sea Cloud), which makes the story less simple.

PASS

11 · Price-to-sales band

v1 PASS → now PASS

Price-to-sales 1.64x [v1: 2.70x]: $45.33B market value ÷ $27.72B of yearly sales. That is in the framework's "below 2 = overly pessimistic" band.

WATCH

12 · Dilution

v1 WATCH → now WATCH (improving)

First-half 2026 diluted shares 634.49M vs 636.23M a year earlier (−0.3%). Basic shares +3.4%. Q2 buyback 4.7M shares.

FAIL

13 · Current ratio ≥ 2

v1 FAIL → now FAIL (worse)

Short-term assets ÷ short-term bills: 24,052.8 ÷ 16,094.4 = 1.49 [v1: 1.58]. Customer deposits ($4,285M) and escrow ($3,330M) inflate the short-term bills. Sea is part bank.

PASS

14 · Net margin positive

v1 PASS → now PASS (shrinking)

Q2 net margin 5.9% [Q2 2025: 7.9%]. Last four quarters: 6.0% [v1 cited 7.0% for 2025]. Tax rate 35.0% vs 25.2%.

PASS

15 · Cash > debt

v1 PASS → now PASS

$11.86B of cash and investments vs $2.22B debt excluding leases ($4.13B including leases) at 2026-06-30.

WATCH

16 · Operating efficiency

v1 PASS → now WATCH ▼

Operating margin 8.35% [Q2 2025: 9.27%]. Sales and marketing +64.5% vs revenue +48.1%. Monee sales and marketing +139.8%. Stock-based pay $131.9M, the 6th straight quarterly decline.

Tally: 9 PASS / 3 WATCH / 4 FAIL [v1: 11 / 2 / 3]. Moved down: gate 7 (insiders, PASS → FAIL) and gate 16 (efficiency, PASS → WATCH). Nothing moved up.

the stock still passes the growth, cash and valuation tests. It now fails the "insiders act like owners" test. That cuts straight at the reason v1 trusted management with a premium.

§7 · FINANCIALS · data as of 2026-10-26 · filings through Q2 2026

Financials: the honest pictures

In plain English

Sales are compounding fast with no acquisitions. Cash from operations is strong. But profit grew far slower than sales this quarter, the tax bill jumped, and once you count new loans, the cash left over went negative. Both are true: the mall and the games make lots of cash, and the lender is absorbing it.

Q2 2026 summary

Revenue $7,787.8M, +48.1% year over year.
GAAP net income (profit under standard accounting rules) $458.1M, +10.6%. Net margin 5.9% [Q2 2025: 7.9%]. Worked example: $458.1M ÷ $7,787.8M = 5.9 cents of profit per dollar of sales.
Gross margin (sales minus direct costs, as a share of sales) 45.6%. Operating margin (profit from operations ÷ sales) 8.35%.
Adjusted EBITDA (operating profit before interest, taxes, depreciation and some non-cash items) $917.2M: +10.6% year on year and −11.3% vs Q1. Most of the drop vs Q1 is Garena's deferred-revenue swing. A Wall Street estimate of about $874M is UNVERIFIED (search-summary source).
Earnings per share: GAAP $0.72 basic, $0.70 diluted. Adjusted earnings per share $0.84 vs a $0.86 estimate.
Stock-based pay $131.9M (6th straight decline). Depreciation and amortization $114.7M.

Revenue by quarter ($M)

+48.1% year over year in Q2 2026 · no acquisitions

Net income by quarter ($M)

Tripled from Q3 2024 to Q2 2026, but only +10.6% in the last year

Segment revenue by quarter ($M)

Shopee · Monee · Garena · Other services + goods sold directly

Operating cash flow by quarter ($M)

Last four quarters' free cash flow $4,423M · but about −$644M in Q2 once new loans are counted

Basic shares outstanding (M)

Basic +3.4% · diluted −0.3% in the first half of 2026 · Q2 buyback 4.7M shares · Q2 2026 is the first-half weighted average

Eight-quarter detail (Q2 2024 rolled off; Q2 2026 added)

All figures in $ millions. Segment columns are service revenue. "Other + goods" = other services plus goods Sea sells itself. Earnings per share = GAAP diluted, vs Wall Street's estimate. Sources: Sea quarterly filings (6-K), stockanalysis, Yahoo Finance.

Quarter	Revenue	Shopee	Monee	Garena	Other + goods	Gross margin %	Operating margin %	Net income	Earnings per share, actual / estimate	Operating cash flow	Basic shares (M)
Q3'24	4,328.2	2,767.2	615.7	497.8	447.5	43.0	4.7	153.3	0.24/0.59	1,170.1	578
Q4'24	4,950.4	3,192.5	733.3	519.1	505.5	44.6	6.2	237.3	0.39/n-a	1,021.0	580
Q1'25	4,841.1	3,118.9	787.1	495.6	439.5	46.2	9.4	403.1	0.65/0.93	756.9	590
Q2'25	5,259.5	3,312.2	882.8	559.1	505.4	45.8	9.3	414.2	0.65/1.03	1,615.7	592
Q3'25	5,986.0	3,778.7	989.9	653.0	564.5	43.4	8.0	375.0	0.59/1.11	1,175.4	598
Q4'25	6,851.9	4,336.1	1,131.9	701.0	682.9	43.8	8.3	397.1	0.63/0.91	1,476.4	602
Q1'26	7,097.5	4,503.1	1,241.8	696.6	656.0	44.3	8.4	427.9	0.67/0.746	1,057.9	612
Q2'26	7,787.8	4,931.9	1,402.8	746.6	706.5	45.6	8.35	458.1	0.70 GAAP / 0.86 est (adjusted actual 0.84)	1,506.0	611.5*

Basis warning. The "estimate" figures are Wall Street's adjusted estimates, while the "actual" figures are GAAP. The two do not match by design (checklist gate 9). Like-for-like adjusted actuals: Q1 2026 $0.81 vs $0.77 (beat), Q2 2026 $0.84 vs $0.86 (miss).
Q2 2026 detail: Other + goods = $48.8M other services + $657.7M goods. *Basic shares are the first-half weighted average, not Q2 alone.
Rule of thumb. Revenue nearly doubled from Q3 2024 ($4,328.2M) to Q2 2026 ($7,787.8M). Net income tripled ($153.3M → $458.1M), but most of that came in 2025. From Q2 2025 to Q2 2026, revenue rose 48% and net income only 11%.

the easy margin gains from cost-cutting are over. Further profit growth has to come from scale.

Q2 2026 segment detail

Shopee

Segment revenue $5,587.6M (+48.2%); service revenue $4,931.9M (+48.9%).
Core marketplace revenue (commissions, ads, transaction fees) $4,255.5M, +65.6%. Value-added services (mostly logistics fees) $676.4M, −9.0%, due to "higher revenue net-off against shipping subsidies."
Gross merchandise value (GMV) $38.3B (+28.4%). Orders 4.2B (+27.5%).
Take rate (Shopee revenue ÷ GMV) 14.59%: 5,587.6 ÷ 38,300 [Q1 ≈13.7%; Q2 2025: 3,771.1 ÷ 29,830 = 12.64%]. Core marketplace ÷ GMV: 11.11% [Q2 2025: 8.61%].
Adjusted EBITDA $255.4M (+12.2%) = 0.67% of GMV [Q1: 0.60%; Q2 2025: 0.76%].
Shopee sales and marketing $1,258.1M (+56.6%) = 3.28% of GMV [Q2 2025: 2.69%].

Shopee charges sellers more per sale, but it hands much of that back as shipping subsidies and marketing. Profit per dollar of goods sold went down vs last year.

Monee

Revenue $1,402.8M (+58.9%). Adjusted EBITDA $288.0M (+12.8%). Adjusted EBITDA margin 20.5% [Q1: 22.2%; Q2 2025: 28.9%].
Monee sales and marketing $293.9M, +139.8%.
Loans $11.1B (+62.5%, slowing from +71.3%). Loans 90+ days late: 1.0%.
Loan-loss charges (money set aside for expected bad loans): $555.2M group / $553.7M segment, +71.5%. That is 39.5% of Monee revenue [Q1: 37.1%].
Loss reserve: $1,114.8M total, of which $1,074.0M is current, = 11.1% of on-book loans. Reserve growth vs Q1 was +13.3–13.4% vs on-book loans +13.6%, a ratio of about 0.98–1.0x [v1, Q1: 2.7x].

Monee is growing fast and paying up for it. Late loans are not rising, but the cost of growth is.

Garena

Bookings $763.5M (+15.5%). Revenue $746.6M (+33.5%). Adjusted EBITDA $429.8M (56.3% of bookings).
Quarterly active users 666.3M (flat). Paying users 68.1M (+10.2%). Average bookings per user $1.15.
Deferred-revenue build only +$16.9M [Q2 2025: +$102.2M; Q1 2026: +$254.8M].

Garena revenue grew faster than what players actually spent this quarter. That gap cannot last.

Group costs

Sales and marketing +64.5%. Research and development +5.6%.
Tax rate 35.0% (tax $250.6M ÷ pre-tax profit $715.9M) [Q2 2025: 25.2%].

Why GAAP profit and operating cash differ (Q2 2026)

Net income $458.1M vs operating cash flow $1,506.0M. The gap is $1,047.9M:

Loan-loss charges $555.2M. Non-cash: money set aside, not spent.
Stock-based pay $131.9M.
Depreciation and amortization $114.7M.
Deferred-revenue build $16.9M.
Other working-capital items about +$229M (not broken out here; derived).

cash from operations looks much better than profit, mostly because the loan-loss charges are a set-aside, not cash spent. But the loans themselves cost cash (next table).

Cash flow

Measure	Q2 2026	Note
Operating cash flow	$1,506.0M	First half $2,563.9M − Q1 $1,057.9M
Capex (spending on buildings and equipment)	$356M	Q1 $139M; 2.6x jump; not explained; last four quarters $792.8M
Free cash flow (operating cash flow − capex)	$1,150M	14.8% of revenue; last four quarters $4,423M [v1: $4.67B]
New loans made (net)	≈$1,794M	First half $2,855M − Q1 $1,061M
Bank-adjusted free cash flow (free cash flow − new loans)	≈ −$644M	[Q1: −$142M]; first half ≈ −$786M

if you treat Sea as part bank (and you should), it currently spends more cash on loans and equipment than it brings in. The gap is funded by customer deposits and the cash pile.

Balance sheet (2026-06-30)

Cash $3,529.3M + short-term investments $5,569.6M = $9.10B. Adding long-term investments of $2,756.1M gives $11.86B [v1: $12.71B]. Restricted cash of $2.46B is excluded.
Debt $4.13B in total: bank borrowings $1,224.3M (non-current part up from $510M to $908M); convertible notes $996.3M, which matured 2026-09-15 (conversion price about $477 per share, far above the stock, so cash repayment is presumed UNVERIFIED until the Q3 filing); leases $1,905.9M.
Debt excluding leases: $2.22B [v1: $1.96B]. Net cash about $9.63B.
Customer money on the same balance sheet: deposits payable $4,285M, escrow $3,330M.
Current ratio 1.49 [v1: 1.58].

Sea is not at risk of running out of money. After the September note repayment, gross cash should fall by about $1.0B, with net cash roughly unchanged.

Capital-structure six-pack (Stage 2)

Item	Now	v1
Price	$74.01 (close 2026-10-26)	$111.14
Shares	612,484,522 (566,956,729 Class A + 45,527,793 Class B) as of 2026-03-31 (20-F, the latest primary count). About 608M after the Q2 buyback UNVERIFIED estimate. Diluted ≈634.5M.	612.48M / 636.2M diluted
Market cap (price × shares)	$45.33B (612.48M × $74.01). Matches Yahoo Finance. No split or share sale found.	$68.07B
Cash and investments	$9.10B cash and short-term; $11.86B including long-term investments	$12.71B
Debt	$4.13B including leases; $2.22B excluding leases	$1.96B excluding leases
Enterprise value (market cap + debt − cash)	$40.36B (debt including leases, minus cash and short-term investments). $35.70B on a narrower definition (debt excluding leases, minus all investments).	≈$56B
Short interest (shares sold short)	10.24M = 3.12% of shares, about 2026-09-30. Prior month 10.94M. Days to cover 2.9.	18.41M / 5.62%

Data note: Yahoo Finance's 566.96M "shares outstanding" counts Class A only, so it is wrong. Media market caps built on it (for example, the Motley Fool's $41.96B) are understated.

Filings since 2026-07-12 (SEC EDGAR, company ID 1703399)

6-K, 2026-08-11: Q2 results. 6-K, 2026-08-25: annual-meeting notice. 6-K, 2026-10-20: Li's memo.
13G/A, 2026-10-21: Baillie Gifford stake update.
Form 144 notices of planned insider sales on 2026-08-11 (three filings), 08-14, 08-17, 08-18, 09-09, 09-11, 09-15 (plus an amendment) and 10-01.
About 100 Form 4 insider-trade reports.
No new stock or debt offering.

§8 · §10 · MANAGEMENT & EARNINGS TRACKER · data as of 2026-10-26

Management: 7/10 v1: 8/10

In plain English

The founders are still in charge and still very rich in Sea stock. But v1 overstated how much stock Li personally owns. Insiders sold heavily this summer. And the October memo opened two new uses of capital without numbers. Management is still strong; we just trust it a notch less.

CEO skin-in-the-game score, corrected

Score = insider's own stake value ÷ annual cash pay · framework: 5x weak → 1,000x ideal

v1: 16.1% × $68.07B = $10.96B. That count = 58,550,224 Class A + 45,527,793 Class B = 104.08M shares

Of which 43,182,662 Class A are voting proxies (Li votes them for others) and ~13.0M are exercisable options

Li's own: 45,527,793 Class B + 2,367,562 Class A ≈ 47.9M shares (7.8%) × $74.01 ≈ $3.5B

$3.5B ÷ $3.13M (all directors' and officers' 2025 cash pay, as if it all went to Li) ≈ 1,100x or more → still PASS

Arithmetic for the Class A count: 58,550,224 − 43,182,662 proxy shares − about 13.0M options ≈ 2.37M Class A owned.
About $4.3B with options at intrinsic value. The red-team review's $7.70B uses all 104.08M shares CONFLICT resolved toward the narrower count UNVERIFIED pending a direct read of the 20-F footnote.
Cash pay: the annual report (20-F) shows $3.13M total for all directors and executive officers in 2025, with no per-person split. So the "$0 salary" pledge is NOT confirmed for everyone.

Li is still a huge owner. He is just half as huge as v1 said.

Votes

Class B shares carry 15 votes each (20-F, answered). Li holds 57.6% of the votes on about 7.8% of the economic ownership (super-voting control).

outside shareholders cannot block anything, including a Monee listing.

Insider selling (Form 4 reports, 2026-07-10 to 2026-10-23): 9 insiders, 2,908,286 shares, ≈$344.0M, zero open-market buys

Insider	Role	Shares sold	Value	Average price / dates
Forrest Li	Founder, CEO	1,139,734	$147.3M	$129.28, 2026-08-11 to 08-28
Gang Ye	COO	1,121,208	$126.1M	$112.45, 2026-07-08 to 09-18
Chen Jingye	Chief product officer, Shopee	300,000	$31.5M	2026-09-09 to 09-22
Wang Yanjun	General counsel	96,820	$10.5M	—
Chris Feng	President	91,709	$11.4M	August
Tony Hou	CFO	65,000	$7.0M	—
Zhao Feng	—	56,420	$5.8M	—
Heng / Kuok	Directors	—	$2.4M / $2.0M	—

Context

Li's sales were under a pre-arranged trading plan (Rule 10b5-1), adopted 2025-09-10 through an entity he controls. All sales were at $120 or higher. None came after 2026-08-28.
The sale is 1.1% of the 104M-share count, or 2.4% of his own 47.9M.
He also exercised 1.47M options at $15 on 2026-03-25. Net, he holds about 0.33M more shares than in March.
Wider window (2026-03-18 to 10-22): $539.5M sold, adding David Ma's $117.2M. By month: August $231.5M, September $64.0M, October $6.1M.
The only offset: President Chris Feng exercised 1.0M options at $15 on 2026-10-21 and kept the shares (Form 4 0001193125-26-429381).

the selling was planned and done at higher prices, so it is not panic. But zero buys at $74, after a −44% fall, says insiders are not calling this a bargain either.

Scores (v1 → now)

Trait	Score	Evidence
Capital allocation	8.5 → 7.0	Sea Cloud opened with no budget. The Monee listing was floated with no structure. The capex jump is unexplained. Credit: a $416.8M buyback in Q2 and no stock or debt sales.
Ethics / governance	6.5 → 6.0	A Monee carve-out would be decided under 15-vote shares. The Ye private-jet allegation is still unchecked.
Long-term orientation	9.5 → 9.5	Sea Cloud is a long-term bet, for better or worse.
Operations	8.5 → 8.5	Unchanged.
Structure / talent	7 → 7	Unchanged.
Iteration tempo	9 → 9	Unchanged.

Bench & board (unchanged)

Chris Feng (President) built Shopee.
Tony Hou is CFO.
Board includes Jessica Tan (ex-Ping An), Silvio Savarese (AI) and Khoon Hua Kuok.

Red flags

Still no group chief technology officer on the leadership page. Sea Cloud makes this gap more relevant, not less.
Founder super-control is now active, not theoretical.

Earnings Tracker: promises vs kept (Q2 2026 graded)

In plain English

We track what management promised and whether it delivered. Q2 kept every explicit guide. The miss was in what management did not say: no budget for Sea Cloud and no Monee structure. The 2026-11-10 print has a clear pass/fail bar, below.

Promise	Status	Evidence
"$0 cash pay until self-sufficiency" (Sept 2022)	KEPT originally	Current status unclear: the 20-F shows $3.13M group cash pay in 2025.
Profitability sustained	KEPT	9 straight GAAP-profitable quarters, including Q2 2026.
2025 profit trajectory	KEPT	2025 net income $1.6B vs $447.8M in 2024.
2026 Shopee gross merchandise value (GMV) ~+25%	ON TRACK	First half +29.3% ($75.6B vs ≈$58.5B). Repeated on 2026-08-11 with a currency warning and tougher second-half comparisons.
2026 Shopee adjusted EBITDA ≥$1B, second half > first half (replaced v1's "≥ 2025 dollars")	PENDING	First half $478.6M (223.2 + 255.4). The second half needs ≥$521.4M, about $261M per quarter.
2026 Garena double-digit bookings growth	KEPT year to date	+20.1%, +15.5%. Not repeated on the Q2 call.
$1B buyback	ON PACE	Q4 2025 $14.0M + Q1 $168.4M + Q2 $416.8M = $599.2M used; ≈$400.8M left (derived; the tracker review's ≈$415M missed the Q4 2025 purchase). No expiry date found.
Convertible notes	PRESUMED REPAID	At the 2026-09-15 maturity. Confirm in Q3.
Free Fire India relaunch	BROKEN	No relaunch.
Q2 earnings "~mid-August"	KEPT	2026-08-11.
Like-for-like adjusted earnings per share vs estimate	2 BEATS / 6 MISSES	Q1 2026 beat ($0.81 vs $0.77). Q2 2026 miss ($0.84 vs $0.86). Over 8 quarters.
Revenue vs estimate	BEAT	Q2 beat consensus. Not separately quantified in this refresh.

Q2 2026 grade: B− on delivery, D on disclosure

Every explicit target held, and revenue and GMV beat. But profit growth was a fifth of sales growth, and nothing was disclosed on Sea Cloud or Monee. The market waited two months, then punished the memo. Sea gives no formal quarterly guidance, only qualitative annual guides. That is part of why the stock whipsaws on prints.

Q3 2026 (2026-11-10) PASS / FAIL bar

Wall Street expects revenue of $8.07B (+34.9%; range $7.69–8.32B) and adjusted earnings per share of $0.97 (5 analysts, Yahoo Finance). The tracker and red-team reviews proposed slightly different bars. This table merges them, using the stricter fail line where they differ.

Metric (Q3 2025 base)	PASS	WATCH	FAIL
Shopee GMV ($32.2B)	≥$39.9B (+24%)	+22–24%	<$39.3B (below +22%)
Shopee adjusted EBITDA	≥$261M (on pace for $1B)	$240–261M	<$240M, or the $1B target withdrawn
Take rate	≥14.3%	—	Cut of more than 30 basis points vs Q2
Value-added-services revenue	Back to growth	—	A second straight decline and Shopee sales and marketing above 3.3% of GMV (a subsidy war)
Monee loans 90+ days late	≤1.2%	1.2–1.5%	>1.5%
Monee annualized loan-loss charge	≤26% of average on-book loans	—	>26%
Reserve growth ÷ on-book loan growth	≤1.5x	—	>1.5x
Monee adjusted EBITDA	≥$288M	—	Down vs Q2
Garena bookings ($840.7M)	≥$925M (+10%)	—	<$841M (down vs a year ago)
Group sales-and-marketing growth vs revenue growth	Slower than revenue (resets the breaker)	—	Faster than revenue for a third straight quarter
Group GAAP net income	≥$431M (+15%)	—	Net-income growth below one-third of revenue growth for a third straight quarter
Sea Cloud / capex	Spending quantified; capex below 30% of operating cash flow	—	2027 capex guide above $1.5B with no outside revenue or contracts disclosed, Q3 capex above $500M, or open-ended spending
Monee listing	Structure disclosed; Sea keeps control	—	Sea sells more than 25%, or loses control
Capital return	Q3 buyback ≥$200M, plus a new authorization of $1B or more	—	Program lapses with no replacement

Comparison warning: from Q3, Garena's adjusted EBITDA excludes the deferred-revenue effect. Compare Garena on bookings, not EBITDA, until Sea restates history.
Currency warning: in Q3 the US dollar averaged +9.1% vs the Indonesian rupiah and +8.6% vs the Philippine peso compared with a year earlier. Reported growth will understate local-currency growth. Ask for constant-currency figures.

if Shopee hits about $261M of profit, credit holds and Sea Cloud gets a number, the "broken" label can come off. If the capex number is large and open-ended, the bear case moves to the front.

§11 · SENTIMENT & POSITIONING · data as of 2026-10-26

Sentiment & Positioning

In plain English

The October drop was not short sellers. It was long holders leaving after the memo, plus Indonesia's market crash and higher US rates. Wall Street analysts are still overwhelmingly bullish, so the framework's contrarian "everyone hates it" signal is only half there.

What moved the stock (2026-08-11 close $131.51 → $74.01, −43.7%)

Window	SE move	Driver	Evidence
2026-08-11 → 10-01	$131.51 → $95.27 (−27.6%)	Post-print fade, insider selling under pre-arranged plans, emerging-market and peer de-rating, rates	GRAB −27.5% and CPNG −16.7% since 2026-08-11. US 10-year yield 4.57% (07-10) → 5.29% (09-30). Insiders sold $231.5M in August and $64.0M in September.
DBS downgrade: Buy → Hold, target $148 → $105 (downside case $71)	About −0.1% to −2.1% on the day	DBS says quick-commerce spending cuts Shopee's 2026 adjusted EBITDA 15–20%, and cites PDD's cross-border push from 2027	CONFLICT on the date 2026-10-01 per itiger (−2.1% to $95.27); 2026-10-02 per ad-hoc-news and the price data (−0.08%, $95.27 → $95.19). The stock then rose +2.4% to 10-15. The price impact was immaterial either way.
MSCI Indonesia warning, 2026-10-16	10-16: −3.83%. 10-19: −5.65% ($88.47)	Indonesia country risk	Indonesia fund (EIDO) −14.9% on 10-16 and −3.3% on 10-19. CONFLICT on 10-19 the tracker review found no catalyst; the red team points to Indonesia; the technical review cites a Dow Jones Market Talk on Shopee currency headwinds plus a Seeking Alpha "insider selling" headline.
Li's memo, 2026-10-20	−9.1% to $80.43 (Yahoo Finance; StockStory says −9.4% to $80.17) on 18.9M shares (4.1–4.4x the 50-day average, depending on the window). 10-21: −3.10% on 15.55M.	The main driver, and specific to Sea	EIDO +7.2% and QQQ −0.1% that day, so it was not the market. Tickeron blamed DBS; that is wrong, because the DBS note was weeks earlier.
2026-10-22 → 10-26	$77.94 → $74.01 (−5.0%)	Follow-through. Dow Jones: Sea Cloud "could weigh on profit," with break-even in 3–4 years (headlines only).	CoreWeave −10.1% and QQQ −2.9% on 10-22.
October insider sales	—	$6.1M total	Negligible as an October cause.

Since 2026-10-15: SE −24.1%, EIDO −9.9%. From the 2026-10-16 close of $93.77: −21.1%.
Read: roughly a third of the fall is emerging-market and peer drag, and roughly a third is the memo plus Indonesia. The rest is follow-through selling.

the biggest single cause was something management said, not something the business reported. Management can also un-say it at the print by putting numbers on Sea Cloud.

Analysts

Downgrades

DBS: Buy → Hold at $105 (October 1 or 2). Cut to $85 on 2026-10-23 (single source, stockanalysis).
New Street: downgrade to Hold/Neutral, target $70 CONFLICT UNVERIFIED dated 2026-10-26 by stockanalysis and 2026-10-22 by one secondary mention.

Still rated Buy

Bernstein $150 (10-20), BofA $130 (10-21), Barclays $156 (10-22).
After Q2: Benchmark $140 → $175 and Barclays $122 → $156 (08-12). Arete $152 (09-07).

Consensus

Ratings: 4 Strong Buy / 22 Buy / 2 Hold / 0 Sell [v1: 22 / 5 / 2 / 0].
Average price target $150.04 (low $70, high $195, 28 analysts). Stale: spot is 51% below it.
The 2027 earnings-per-share estimate drifted from $5.25 (90 days ago) to $5.10 (−2.9%).

the analyst downgrade cycle has room to run. That is a near-term headwind, but it is also how bottoms usually form.

Funds (13F filings, Q2)

418 funds added, 305 cut.
Biggest adds: JPMorgan +8.19M shares, T. Rowe +2.23M.
Growth funds trimmed: Tiger −2.86M, Sands −2.50M, D1 −2.37M; William Blair exited.
Baillie Gifford: 40.95M → 37.15M shares (6.55%) per its 2026-09-30 filing (13G/A).
Q3 filings are due by about 2026-11-14.

Short interest

18.41M → 10.24M shares (−44%), 3.12% of shares, about 2026-09-30. Days to cover 2.9.

short sellers were covering, not piling in. There is no squeeze fuel, and no evidence of a bear raid.

Contrarian gate (Stage 7)

The framework buys "unanimous despair with an executing roadmap."

Despair: price yes (−64% from the peak, very oversold). Analysts no (26 of 28 still say Buy).
Executing roadmap: yes on the business, no on capital allocation.
Retail sentiment (X, Reddit, Substack) was not sampled in this refresh UNVERIFIED.

half the setup for a starter position exists. The other half depends on the print.

§9 · §12 · MOAT, VALUATION & SCENARIOS · data as of 2026-10-26

Moat & Competition: 7/10 v1: 7.5–8/10

In plain English

A moat is what stops competitors from taking your customers. Shopee's moat is real: its own delivery network, its scale and its lending data. But TikTok Shop keeps crossing it. Shopee is still the bigger and more profitable player, just by less each quarter. Further down this tab: at $74 you pay about $1.64 for every $1 of Sea's yearly sales, the lowest-but-one in the peer group, and the re-run base case is $109.

Market	Shopee	TikTok Shop (incl. Tokopedia)	Source / period
Southeast Asia, big three	55% [61% a year earlier]	39% [27%]	Cube, first half 2026 vs first half 2025
Southeast Asia, GMV (gross merchandise value) ratio	—	TikTok + Tokopedia GMV = 65.7% of Shopee's	Momentum Works, 2025
Vietnam	53% [61%]	44% [33%]	Q&Me, May 2025 to April 2026 (secondary summary)
Indonesia	54%	38%	2025 (not refreshed for 2026)
Thailand	~52%	48% of combined	Cube, Q2 2026

TikTok Shop's first-half 2026 GMV was $34.7B, +80% (Momentum Works). Shopee's Indonesia GMV grew +19% and Vietnam +10% (Cube).

Why the moat score fell half a point

TikTok took about 12 points of combined share in one year. Thailand is near a flip.

Why it did not fall further

Shopee raised its take rate about 195 basis points while losing share.
Advertising take rate +90 basis points.
Shopee is profitable. TikTok Shop's profitability is undisclosed.
SPX instant delivery is scaling with cost per order down about 20%.

customers and sellers are splitting their time, but Shopee still has pricing power. Kill line (unchanged): regional share below 50%, or a take-rate cut. Not hit: 55% and a rising take rate.

Peers (2026-10-26; Yahoo Finance figures as reported)

Ticker	Price	Market cap	Price / sales	Enterprise value / sales	Forward price / earnings	Revenue growth	Operating margin	Off 52-week high	v1 values
MELI	$1,668.72	$84.6B	2.40	2.62	29.7	+49.8%	6.7%	−31%	Forward P/E 46.1, P/S 2.95
CPNG	$13.75	$24.7B	0.70	0.68	49.6	+3.9% UNVERIFIED, looks wrong	−6.3%	−57.5%	P/S 0.96
GRAB	$2.71	$11.1B	2.97	1.87	20.5	+21.9%	2.1%	−56.6%	Forward P/E 41.9, P/S 4.53
SE	$74.01	$45.3B	1.64	1.46	14.5 (next 12 months)	+48.1%	8.35%	−54.9%	P/S 2.70

Forward price-to-earnings = price ÷ next year's expected earnings per share. Price-to-sales = market value ÷ yearly sales.
MELI's free-cash-flow data on Yahoo Finance is not usable, so no price-to-free-cash-flow is shown.
Since 2026-07-10: SE −33.4%, MELI −9.9%, GRAB −31.0%, the Nasdaq-100 fund (QQQ) −1.7%.
Read: SE has the lowest forward price-to-earnings and the second-lowest price-to-sales in the group, with growth level with MELI. The whole group de-rated, and SE de-rated the most.

the market prices Sea as if its profits are at risk, not its sales.

Valuation

In plain English

At $74 you pay about $1.64 for every $1 of Sea's yearly sales, and about 14.5 times next year's expected profit. The earnings multiple is the lowest in Sea's peer group, and the sales multiple the second-lowest. Our re-run base case is $109, but it assumes profit margins recover only modestly.

Lens	How it works (worked example)	SE now	v1
Price / sales (last four quarters)	Market cap $45.33B ÷ revenue $27,723M	1.64x	2.70x
Forward price / sales	$45.33B ÷ 2026 estimate $31.19B / 2027 estimate $38.33B	1.45x / 1.18x	—
Enterprise value / sales	$40.36B ÷ $27,723M (or $35.70B on the narrower definition = 1.29x)	1.46x	2.43x
Price / free cash flow	$45.33B ÷ $4,423M (stockanalysis: 10.78x)	10.2x	14.6x
Enterprise value / EBITDA	On GAAP EBITDA ≈$2.68B: 15.0x. On adjusted EBITDA over the last four quarters, $3,612.9M (874.3 + 787.1 + 1,034.3 + 917.2): 11.2x	11.2–15.0x	—
Trailing price-to-earnings	$74.01 ÷ GAAP diluted earnings per share over the last four quarters, $2.59 (0.59 + 0.63 + 0.67 + 0.70)	28.6x	43.8x
Forward price-to-earnings	$74.01 ÷ 2026 adjusted estimate $3.92 / 2027 estimate $5.10	18.9x / 14.5x	27.0x
PEG (price-to-earnings ÷ growth rate)	Yahoo Finance	0.73	1.43

Revenue per share is $45.26 (27,723 ÷ 612.48). So 1.0x sales = $45, 1.5x = $68 and 2.0x = $91.

a return to just 2x sales, the bottom of the framework's "left for dead" band, would be about $91.

Sum-of-parts check

DBS values Monee alone at $18–22B (12–15x its estimated 2027 adjusted EBITDA of about $1.5B). The share price implies about $6B for Monee (BigGo, 2026-10-21) UNVERIFIED secondary source.

if a Monee listing priced near DBS's range, it would highlight hidden value. That is the bull reading of the same news the market sold.

Scenarios re-run (exit at the end of 2027 · GAAP earnings per share · diluted shares)

🐂 Bull · 2027 exit

$203 (+174%)

2026 revenue $31.5B; 2027 ×1.25 = $39.4B
Net margin 11%
640M shares → $6.77 a share × 30
v1 target: $211

😐 Base · 2027 exit

$109 (+47%)

2026 revenue $31.0B; 2027 ×1.20 = $37.2B
Net margin 7.5%
640M shares → $4.36 a share × 25
At a 20x multiple, base = $87
v1 target: $124

🐻 Bear · 2027 exit

$46 (−38%)

2026 revenue $30.0B; 2027 ×1.10 = $33.0B
Net margin 5.0%
645M shares → $2.56 a share × 18
v1 target: $52

base case: $37.2B × 7.5% = $2.79B of net income. $2.79B ÷ 640M shares = $4.36 per share. $4.36 × 25 = $109.

What moved vs v1

Revenue inputs went UP. Actual growth runs about 6 points above v1's path (2026 revenue $31.0B vs $29.8B).
Margin inputs went DOWN. Trailing GAAP net margin is 6.0% and falling. The drivers: the 35% tax rate, loan-loss charges +71.5%, sales and marketing +64.5%, and a new Sea Cloud cost base.
Share count is roughly flat. Buybacks offset dilution, with about $401M of authority left.

Same-basis sanity check (fixing v1's error)

Base 2027 GAAP earnings per share is $4.36.
Consensus adjusted 2027 earnings per share is $5.10. Subtract stock-based pay of about $0.84 per share per year (≈$0.21 per quarter) and you get about $4.26 in GAAP terms.
So the base case is in line with Wall Street, not above it.

Skew

Bull +174% vs bear −38%.
Base vs bear: +47% vs −38% = 1.24 to 1 [v1: 1.7 to 1].
Probability-weighted value (PM judgment: bull 15%, base 50%, bear 35%) ≈ 0.15 × 203 + 0.50 × 109 + 0.35 × 46 = ≈$101.
The red team's bear range is $46–55 (Red Team tab).

the stock is much cheaper than in July, but the upside-to-downside balance is worse. In July the main risk was the multiple. Now the risk is the margin itself.

Macro stamp (Stage 1; rates as of 2026-10-22/23)

Gauge	Reading
Fed funds rate (the Fed's policy rate)	3.75–4.00%. A hold is about 90% likely at the 2026-10-28 meeting. December has about 71% odds of a hike (CNBC/FedWatch, 2026-10-23).
10-year Treasury yield	4.96% (2026-10-22), about 5.00% at the 2026-10-23 close. 2-year 4.65%. Inflation-adjusted 10-year 2.64%.
Prices	Headline inflation (CPI, the consumer price index) 3.7%. Brent oil about $114.
Growth	Purchasing-manager surveys (ISM, above 50 = expanding): manufacturing 54.5, services 54.9. Atlanta Fed real-time GDP estimate (GDPNow): +3.5%. Last monthly jobs gain: +29K.
Stock fear gauge (VIX)	19.10 [2026-10-23: 20.15]
Credit	Junk-bond spread over Treasuries: 3.40%, +60 basis points in 20 days. Lowest-rated (CCC) bond yield: 13.40%, +228 basis points. Investment-grade spread: 0.89%.
Liquidity	Contracting: tightening signals outnumber easing ones 6 to 2.
Framework rules	"Never initiate unprofitable small caps late in a liquidity cycle" does not apply: SE is a $45B GAAP-profitable large company. The Stage-9 late-cycle test ("only names that completed a 60–85% correction") is met on depth: −64.1% from $199.30 to $71.53. It is not met on structure: the chart has made no higher low.

Macro: YELLOW. A roughly 5% 10-year yield squeezes what investors will pay for growth. It is a headwind, not a veto for this name.

First-principles KPIs: not Wall Street's

A KPI (key performance indicator) is a number tracked to judge progress. These eight are the ones that tell you whether the business engine is healthy. All are checked at the 2026-11-10 print.

KPI	v1 value	Now	Healthy reading	Check by
1. Delivery cost per order (SPX), vs a year ago	−6% Asia / −21% Brazil	Instant delivery about −20% (different basis; from the Q2 call summary)	Still falling	2026-11-10
2. Take rate and share, together	13.7% / 53–59%	14.59% / 55% (Cube)	Take rate flat or up while share holds at 50% or more	2026-11-10
3. Reserve growth ÷ on-book loan growth	2.7x (Q1)	0.98x (improved)	Below 1.5x	2026-11-10
4. Bank-adjusted free cash flow (operating cash flow − capex − new loans)	−$142M (Q1)	−$644M (worse)	Back toward zero as loan growth slows	2026-11-10
5. Garena deferred-revenue build (bookings minus recognized revenue)	+$254.8M (Q1)	+$16.9M (drained)	Positive and rising; it is the gaming engine's fuel gauge	2026-11-10
6. Stock-based pay in dollars	$133.3M	$131.9M (6th straight decline)	Keeps falling	2026-11-10
7. NEW: Shopee adjusted EBITDA ÷ GMV	—	0.67% (Q1 0.60%; Q2 2025 0.76%)	Rising toward Li's 2–3%	2026-11-10
8. NEW: net-income growth ÷ revenue growth	—	0.22 (10.6% ÷ 48.1%)	0.33 or more	2026-11-10

two of the eight got worse this quarter (cash after loans, and Garena's fuel gauge), one got clearly better (credit reserves), and the new profit-versus-sales gauge sits below its healthy line. The print decides which way the rest break.

§13 · RED TEAM · DOWNSIDE SCENARIOS RE-GRADED · data as of 2026-10-26

Red Team: downside scenarios re-graded

In plain English

The red team's job is to argue the bear case as hard as possible. Its verdict: the bear case got stronger on the facts, because profit is lagging and spending is rising. But the price already reflects much of it. The re-run bear is 26–38% below today, vs 53% below in July.

v1 did not publish the framework's 10 priced downside scenarios. Below, v1's 9 ranked risks plus its bear case are mapped to scenarios 1–10, with five new ones added.

Pricing base for the scenarios

2027 consensus revenue $38.33B. Trailing GAAP net margin 6.0% (1,658 ÷ 27,723). 634.5M diluted shares.
"Current path" 2027 GAAP earnings per share ≈ $3.62 (38.33B × 6.0% ÷ 634.5M). $74 is about 20.4x that.
Grades compare with v1 directionally. v1 gave no probabilities.

Implied price by scenario vs today's $74.01

Bars show each scenario's implied price; the dashed line is today's close. Ranges are drawn at their midpoint. Scenario 15 folds into #1 and has no separate price.

#	Scenario	Grade vs v1	Dated evidence	Implied price
1	TikTok Shop takes share for good	WORSE	Cube (2026-10-13): Shopee 61% → 55%, TikTok 27% → 39%. Thailand TikTok 48%. Vietnam Shopee 53% / TikTok 44%. Offset: take rate +195 basis points.	$34B revenue × 5% ÷ 634.5M = $2.68 × 16x ≈ $43
2	Monee credit cycle	WORSE (mild)	Annualized loss charge 21.7% → 23.6%. Charges = 39.5% of revenue. Mix shift to Brazil and off-Shopee loans. Monee marketing +139.8%. Rupiah near its record low. Offsets: 1.0% late loans, reserves keeping pace (0.98x), book growth slowing.	Monee profit (~$1.15B run rate) wiped out ≈ −$0.9B after tax → $1.35B → $2.13 × 15–18x ≈ $32–38
3	A reinvestment print repeats the March selloff	HAPPENED (differently)	Earnings day was +14.6%. The punishment came 2026-10-20 from the memo. Adjusted EBITDA growth +9.3% / +10.6% vs revenue +46.6% / +48.1%.	$3.62 × 18x ≈ $65
4	Valuation multiple shrinks (macro)	HAPPENED	Price-to-sales 2.70 → 1.64. Forward price-to-earnings 27.0 → 14.5. 10-year yield 4.57% → 4.99% (peak 5.29% on 09-30). Cause mostly emerging-market and Sea-specific: QQQ −1.7%.	$3.62 × 15x ≈ $54
5	Currency translation	WORSE	Q3 average vs a year earlier: the dollar up 9.1% vs the rupiah, 8.6% vs the peso, 6.9% vs the Taiwan dollar, 3.0% vs the baht. The Brazilian real is 5.7% stronger. Indonesia is ≈40% of Shopee's Southeast Asian GMV.	Revenue −6%, net income −8% → $3.33 × 20x ≈ $67
6	Free Fire concentration	UNCHANGED	Bookings growth 20.1% → 15.5%. Users flat. Deferred-revenue build collapsed to $16.9M. The 20-F says Free Fire is a "significant majority" of Garena. New titles announced, not launched.	Garena profit −40% (≈ −$0.52B after tax) → $2.73 × 18x ≈ $49
7	Indonesia regulation	WORSE (country risk, not rules)	MSCI downgrade warning 2026-10-16. Jakarta index ≈ −40% for the year. KPPU logistics complaint vs Tokopedia (filed 2026-04-15, pending). Tokopedia deal monitoring runs to 2027-06-17. No new anti-Shopee rule.	Net income −20% → $2.90 × 17x ≈ $49
8	Governance	WORSE	Li sold $147.3M. A Monee carve-out dilutes outside holders' share of the fastest-growing segment, decided under 15-vote shares.	15% conglomerate discount ≈ $63
9	Analysts all bullish, so only downgrades are left	PARTLY HAPPENED	DBS to Hold. Low target now $70 [v1: $101]. 2027 earnings estimate −2.9%. Spot 51% below the average target.	Estimate cut −20% to $4.08 × 14.5x ≈ $59
10	v1 bear valuation ($52)	Revenue leg FALSIFIED, margin leg CONFIRMED	v1 bear 2027 revenue $29.5B vs trailing $27.7B already. v1 bear margin 6.5% vs actual 5.9% (Q2) and 6.0% (trailing).	Mechanical re-run: $3.03 × 18x = $54.5. Updated: $35.6B × 5.5% ÷ 634.5M = $3.09 × 16x = $49
11	NEW: Sea Cloud capex cycle	NEW, medium	Q2 capex $356M (2.6x Q1), no budget. The AI-cloud comparable CoreWeave is −37.9% since 2026-07-10.	Capex rises to ~$2.5B a year plus ~$0.3B of cloud losses ≈ −$0.53B net income → $2.71 × 18x ≈ $49. Free cash flow before loans falls from $4.42B to about $2.7B.
12	NEW: Monee listed at a low fintech valuation	NEW, low-medium	Nubank −16.95% on 2026-10-26 (cause UNVERIFIED), about −30% for the year (secondary). The memo says "could."	Floating 20%: about $0.84B of Monee after-tax profit × 20% = $0.17B a year = $0.26 per share × 20x ≈ −$5 → $69. The bigger risk is losing the Shopee-Monee data and funding link (flywheel step 4).
13	NEW: Indonesia capital flight / rupiah crisis	NEW, medium-low	The rupiah moves from 18,081 to 20,000+ and spreads into consumer credit. Indonesia is SeaBank's home market.	Scenarios 2 + 5 + part of 7 ≈ $45–50
14	NEW: confusing Q3 print	NEW, medium	Garena's adjusted EBITDA definition changes. Q2 group adjusted EBITDA already fell 11.3% vs Q1.	A −10% to −15% drop on a misread → $63–67, temporary
15	NEW: Thailand flips to TikTok	NEW (folds into #1)	TikTok at 48% of combined Q2 GMV (Cube).	Watch for a country-level take-rate cut.

Spread: of 15 scenarios, 5 price at $49 or less. The median implied price is about $54–59.

most bad outcomes land between $45 and $65. Today's $74 is above all of them. The cushion is real, but thinner than the cheap multiples suggest.

Bear case: stronger or weaker? STRONGER on facts, cheaper to own

Stronger, because

v1's bear rested on revenue slowing to +15% / +12%. That leg is dead: revenue is +48%.
The bear moved to the profit line, and it is winning there: GAAP net income +10.6% on +48% revenue; adjusted EBITDA growing at about a fifth of the revenue rate; net margin 5.9%, already below v1's bear 6.5%; tax rate up 10 points; bank-adjusted free cash flow about −$644M.
TikTok took about 12 points of share in a year.
The 2026-10-20 memo turned "harvest a self-funded duopoly" into "open a capital-hungry AI venture with no budget, and maybe carve out the best-growing asset."
Three v1 business or insider breakers and two price breakers fired.

Weaker, because of price

At $74.01, the market already discounts most of the margin bear: 1.64x sales, 1.29–1.46x enterprise value to sales, about 20x current-path 2027 GAAP earnings and 14.5x consensus.
The re-run bear ($46–55) is −26% to −38% away, vs −53% at v1.
Credit is steady: 1.0% late loans, 11.1% reserve coverage, reserves matching book growth.
Short interest halved.
Li's sales were planned and done at $120 or higher, and he is net positive on shares after his option exercise.

Red-team call

Honor the breaker. No stock adds before 2026-11-10. Re-entry depends on the Sea Cloud capex number, not on chart levels.

Audit of v1 (red-team findings)

See the "v1 errors corrected" list on the What Changed tab. Two further audit items:

The ladder was mis-built. All three zones traded through within 23 trading days. The $75 stop sat only $2.05 below the $77.05 low. That is 0.2 of a typical weekly price swing (average weekly range $10.45), so it was within normal noise. The 200-week average was both the biggest buy zone and an exit trigger, so the plan bought at the level where it would later sell.
Tactical calls missed: the "compression into July inflation/Fed" call and the "Q2 repeats March" call.

the rules failed partly because they were built badly, not only because the stock fell. v2's rules (Trade Plan tab) fix the spacing and the double-duty levels.

§15 · §16 · TRADE PLAN, RISKS, VALIDATORS & BREAKERS · data as of 2026-10-26 close ($74.01)

Trade Plan: exit honored, re-entry by puts

In plain English

The chart is broken. Price fell below every major average and a key multi-year average, on heavy volume. It is also very oversold and sitting on a historically busy price area. That could be a bottom, but nothing confirms it yet. So: no stock buys until after 2026-11-10. Before the print, the only tool is selling cash-secured puts at lower prices. The position size is half of v1's.

RSI (14-day momentum)

17.7 d / 31.35 w

Daily: second-lowest since 2023 · weekly not yet below 30

Fibonacci level being tested

0.886 = $71.45

Of the $55.00 → $199.30 advance · low so far $71.53

Trend

Below all averages

20-day $90.61 · 50-day $102.48 · 200-day $99.69 · 200-week $93.98

SE weekly: candles · daily Bollinger bands (20, 2) and 20/50/200-day averages · v2 zones

Weekly candles from 2024-11-01 through 2026-10-26 · indicators computed on DAILY data, sampled weekly · daily RSI 17.7 · hover or touch-drag for open/high/low/close

20-day avg50-day avg200-day avgBollinger bandsbuy zones A/B/C (puts only before 11-10)trim bands

Decoder for this tab (each term is in the Glossary)

20/50/200-day average: the average closing price of the last 20, 50 or 200 trading days (simple moving average). The 200-week average covers about 4 years.
RSI-14: a 0–100 momentum gauge. Below 30 is oversold; below 20 is extreme.
ATR-14: the average daily price range over 14 days.
Bollinger bands: a channel two standard deviations around the 20-day average.
MACD: a trend-momentum indicator built from two moving averages.
Fibonacci retracement: a level at a set fraction (0.618, 0.786, 0.886) of a past move.
High-volume node: a price range where many shares changed hands.
Cash-secured put: you get paid now to agree to buy at a lower price.
Days to expiry, implied volatility (the option market's expected swing) vs realized volatility (the actual swing), delta (roughly the chance an option ends in the money), covered call (you get paid to agree to sell shares you own at a higher price).

PM rulings layered on the technical plan

The v1 exit is honored (Verdict tab). The v1-plan position is exited, not averaged.
Each v2 price line has one action. Weekly close below $55 = stop all adds and re-underwrite. It is a ladder stop, not a thesis exit. Thesis exits come only from the business breakers below. This removes v1's "stop adding" vs "exit" contradiction.
CONFLICT the tracker review proposed different levels (zones $65–70 and $55–60, no-chase $95, repair trigger weekly close above $83, invalidation weekly close below $64). The technical review's levels below are used. The tracker's "weekly close above $83 = repair" is kept as a secondary confirmation only.
The technical review measured ATR-14 at $3.66; the tracker review measured $3.74. Immaterial.
All profit and loss figures below grade the v1 plan hypothetically. They are not anyone's actual account.

Verdict on the v1 exit rules

v1 breaker	Status	Evidence (dated)
Weekly close < $75	FIRED	Week ending Fri 2026-10-23 closed $74.63. Follow-through 10/26: close $74.01, intraday low $71.53 (new 52-week low).
Loss of the 200-week average on >2x average weekly volume	FIRED	The 200-week average was $93.87 on 10/23 (now $93.98; v1 said ~$90.09; it has risen as 2022 lows roll off). Weekly close $74.63 is 20.5% below it. Week volume 57.56M = 2.79x the 20-week average (20.62M) and 2.73x the 10-week. First daily close below it: 10/19 at $88.47.
Daily death cross with price < $85	NOT YET · ~5 trading days out	50-day 102.48 vs 200-day 99.69, a 2.79 gap, shrinking ~0.65–0.77/day (6.48 → 2.79 over 8 sessions). Projection: the cross lands on trading day +5 (≈2026-11-02) at flat $74, at −0.5%/day, and even at +1%/day (≈$78). It will very likely fire before the 11/10 print. The last cross was a golden cross on 2026-08-26; the prior death cross was 2025-12-11.

The v1 exit rule fires: 2 of 3 conditions are met and the third is near-certain within days. As written, "exit, don't average" applies now.

Was the rule wrong? It worked as written. The design had three flaws:

Flaw 1: the 200-week average was both v1's Zone-2 "big dog" add level (40% of the tranche) and an exit trigger. The plan bought at the level it would later exit on, which guaranteed a whipsaw.
Flaw 2: the $75 hard stop sat only 3.8% below the bottom of Zone 3 ($78). That is less than one daily ATR (5% today) and inside the $72.5–75 high-volume node. It was not a structural level.
Flaw 3: all three zones sat inside one 22% band ($78–100). One news event filled all three in 3 sessions (10/19–10/21), which breaks the framework's "ladder in tranches" principle in spirit. There was no time spacing.

Bottom line: honor the rule. No adds at market and no averaging down in stock before 2026-11-10. Re-entry is conditional.

Structure now (2026-10-26)

Ranges: close $74.01. 52-week high $164.03 (2025-10-28), low $71.53 (2026-10-26): 54.9% off the high, 3.5% above the low. Cycle high $199.30 (2025-09-11): drawdown to the low −64.1%. From the all-time high of $372.70 (2021-10-19): −80.8%. Post-Q2 swing high $131.94 (2026-08-11): the full round trip from there is −45.8% in 2.5 months.
Moving averages (daily, all above price): 20-day 90.61 (price −18.3% below), 50-day 102.48 (−27.8%), 100-day 102.35, 200-day 99.69 (−25.8%). The 200-day is falling (it was ~107 in late August). 200-week 93.98 (price −21.2% below). 100-week 124.61. 50-week 103.33.
Bollinger (20, 2): upper 107.51, lower 73.71. Price is sitting on the lower band.
RSI-14: daily 17.73, the second-lowest reading since 2023; the only lower one was 17.0 on 2023-08-24. The analog cuts against calling a bottom: after the August 2023 RSI-17 (~$38), SE chopped for another 5 months and bottomed 11% lower at $34.35 on 2024-01-17. Weekly 31.35, not yet below 30.
MACD: daily −7.20 vs signal −5.37, histogram −1.83 and still widening. No bullish divergence. Weekly −5.24 vs −1.98, a bearish cross.
Volatility: ATR-14 $3.66 = 4.95% of price. Realized volatility: 20-day 44.3%, 30-day 40.6%, 60-day 50.4%.
Volume: climax 10/20 18.92M (4.34x the 50-day), 10/21 15.55M (3.37x). For comparison, the 2026-03-03 capitulation was 34.8M. Follow-through: 10/22 1.91x, 10/23 1.44x, 10/26 1.92x. 10/26 closed at 78% of its range ($71.53–74.71), a hammer-like candle but a single bar.
Unfilled gaps overhead: $86.43–87.88 (10/19 low to 10/20 high) and $93.09–93.49 (10/16 low to 10/19 high).
Volume profile since 2024-01 ($2.5 bins): high-volume nodes 72.5–75 (105M), 82.5–85 (106M), 87.5–90 (110M), 95–97.5 (104M), 65–67.5 (91M), and a 52.5–60 shelf (247M combined). Low-volume pockets: 60–62.5 (36M) and 47.5–52.5.
Short interest: 3.12% of float. No squeeze fuel.
Price-to-sales at spot: market cap $45.33B ÷ $74.01 = 612.5M shares implied (Yahoo Finance's 567M share count is wrong; use 612.5M). Revenue over the last four quarters $27.72B gives 1.64x, below the framework's 2–4x "left for dead" band. (PM note: $27,723M ties to the Q2 6-K sum of the last four quarters.)

Pro-capitulation

Volume climax
RSI 17.7
Price at the 0.886 Fibonacci level of the full advance ($71.45)
The high-volume node at $72.5–75
A −64% drawdown, inside the framework's 60–85% correction band
Price-to-sales 1.64

Against

A lower low on 10/26, no higher low
The 200-week lost on 2.8x volume
A death cross ~5 sessions away
The catalyst (Sea Cloud capex and the Monee listing) unresolved
An all-or-nothing print 10 sessions out
The August-2023 RSI analog that kept falling

Capitulation or broken? Broken chart, capitulation not confirmed. Framework Stage 9 "late-bull" test: the depth is met (−64%), but "wave count done" is not. Classification: ACCUMULATE ONLY VIA PUTS OR AFTER THE PRINT; DO NOT BUY STOCK AT MARKET.

the selling looks exhausted, but exhausted selling and a finished bottom are different things. The plan waits for the print to tell them apart.

Fibonacci grid and 200-week average: what changed vs v1

Level	v1 (7/10)	Now (10/26)	Status
Full advance $55.00 (2024-08-05) → $199.30 (2025-09-11): 0.618	110.12	110.12 (same swing)	Was support in v1; now overhead resistance
Same: 0.786	85.88	85.88	Broken 10/20 (gapped down through it)
Same: 0.886	not cited	71.45	Being tested (low $71.53)
Same: 1.0	55.00	55.00	Next major support
Secular $34.35 (2024-01-17) → $199.30: 0.786	n/a	69.65	New support
200-week average	90.09	93.98 (rising ~0.1/week)	Moved +$3.89; lost on 10/19 daily and the 10/23 weekly
New recovery swing $77.05 (2026-03-03) → $131.94 (2026-08-11): 0.5 / 0.618 / 0.786 / 1.0	n/a	104.50 / 98.02 / 88.80 / 77.05	All broken
Recovery swing extensions 1.272 / 1.618	n/a	62.12 / 43.13	1.272 is the next measured target
Decline $131.94 → $71.53 retraces 0.236 / 0.382 / 0.5 / 0.618 / 0.786	n/a	85.79 / 94.61 / 101.73 / 108.86 / 119.01	New trim grid
Decline $199.30 → $71.53 retraces 0.236 / 0.382 / 0.5 / 0.618	(v1 used 0.382 recovery 123.75 etc.)	101.68 / 120.34 / 135.42 / 150.49	v1's recovery grid is invalidated because the swing extended down

v1 ladder, trims and no-chase graded

v1 call	Grade	Detail
No-chase $116	CORRECT	The 8/11 gap reached $131.94 and fully round-tripped. Anyone who chased $116–131 is now down 36–44%.
Trim 1 $123–127	CORRECT, HIT	8/11 high 131.94, close 131.51, 8/12 high 131.90. The best call in v1.
Trim 2 $138–145, Trim 3 $152–155, Runner $170–175	RETIRE	Never reached. The swing math is invalidated.
Zone 1 $95–100 (30%)	FILLED · RETIRE	Late Sep/early Oct (lows: 9/18 99.54, 9/25 99.25, 10/1 93.70). Now −22 to −26%. Now overhead resistance: 200-week 93.98, high-volume node 95–97.5. (PM note: the tracker review dated the first Zone-1 fill to 2026-07-23, low $99.30; different touch thresholds. Immaterial.)
Zone 2 $86–91 (40%)	FILLED · RETIRE	10/19 (low 87.88), gapped through on 10/20 (open 85.52). Now −14 to −19%. The 200-week/0.786 overlap failed on volume.
Zone 3 $78–83 (30%)	FILLED · RETIRE	10/20–10/21 (lows 79.89 / 75.99). Now −5 to −11%.
Hard invalidation $75	FIRED	Weekly close 2026-10-23.

a full v1 ladder executed blended at ≈$88.8 (0.3 × 97.5 + 0.4 × 88.5 + 0.3 × 80.5). At $74.01 that is about −16.7%.

v1 called $77–92 "the fat pitch" and treated the 200-week/Fibonacci overlap as near-certain support, without a stop that sat clearly apart from the buy zones. That is what v2 fixes.

New plan: half-size, puts first

Gate

The current tranche is about 50% of v1's tranche size: the exit rule fired and the moat is still below 9, so this stays a conviction-spec position. No stock buys before the 2026-11-10 print (Yahoo Finance date, pre-market US per Sea's pattern; confirm with investor relations). Zones A–C are stock buy zones only after the print. Before the print, accumulate only through cash-secured puts at Zone B/C strikes. A Zone A put is allowed only if it expires before the print (11/6).

$69–72Zone A. 0.886 of the full advance 71.45; secular 0.786 69.65; 52-week low 71.53; lower edge of the 72.5–75 node. Price-to-sales 1.52–1.59x.25% of tranche

$62–66Zone B. 1.272 extension 62.12; 65–67.5 node; Aug-2024 breakout gap 60.75–64.86. Price-to-sales 1.37–1.46x.35%

$55–58Zone C. 100% retrace of the advance (Aug-2024 low 55.00); 52.5–60 shelf (247M shares traded). v1's bear case was $52. Price-to-sales 1.22–1.28x.40%

< $55New hard invalidation. Weekly close below $55 (the 2024–25 advance fully erased). Stop all adds and re-underwrite toward v1's bear case ($52). Not an exit by itself.STOP ADDS

Price-to-sales uses last-four-quarters revenue of $27.72B and 612.5M shares.

Post-print path: "pass"

The print passes the Management-tab bar and the stock reclaims and holds a daily close above $86.43–87.88 (gap fill + 0.236 of the decline at 85.79 + the old 0.786 at 85.88). Buy only on the first pullback to $78–82, and at most 25% of the tranche there. No-chase line: $88. Above it, do not start buying: the next resistance cluster is $93–95 (200-week, 0.382 of the decline, gap, node), so reward-to-risk vs Zone A is poor.

Post-print path: "fail"

The print fails and price breaks $69 on a weekly close. Skip Zone A and let Zone B/C fill only through puts. (PM note: a "fail" print that also triggers a business breaker overrides the Zone B/C put plan. In that case, stop selling puts.)

Time spacing (new rule): fill at most one zone per calendar week, and never two zones on the same news event.
Death cross: once it confirms (~11/2), require a weekly close back above the 200-day average (≈$99 and falling) before any tranche size returns to v1 levels.

New trim bands (replace v1's T1–T3)

Trigger	Action	Why
$85–88	T1: sell 20–25% of the trade tranche	0.236 of the decline 85.79, old 0.786 85.88, gap 86.43–87.88, 87.5–90 node
$93–95	T2: sell 25%	200-week 93.98, 0.382 of the decline 94.61, gap 93.09–93.49, 92.5–97.5 node
$100–104	T3: sell 25%	Daily 200-day / 50-day averages (99.7 / 102.5 today, both falling), 0.5 of the decline 101.73, 0.236 of 199 → 71 at 101.68
$108–111	Runner	0.618 of the decline 108.86, 0.618 of the advance 110.12, 0.382 of the recovery 110.97. Keep a core above that only if the print shows reacceleration
Price-to-sales ≥ 4.0 (≈$181 on current revenue)	Trim regardless of chart (kept from v1)	Valuation rule. Irrelevant this cycle

Caveat on the blended v1 cost: for a holder at the v1 blend (~$88.8), T1 and T2 are at or below cost. T2 is the first band that gets back to breakeven.

Wheel angle (TradingView options chain, close 2026-10-26, spot $74.01)

Implied volatility by expiry (at the money): Nov-06 56.1% (10 days, expires before the print) · Nov-13 77.5% (17 days, first post-print) · Nov-27 67.9% · Dec-18 63.6% · Jan-15 59.5%.
Seller edge vs realized: Dec-18 implied 63.6% vs realized 30-day 40.6% / 60-day 50.4% = +13 to +23 volatility points. In v1 the edge was +7 to +15. The edge is wider now.
Implied earnings move: comparing Nov-06 and Nov-13 volatility gives an earnings-day one-standard-deviation move of ≈11.6% (expected absolute move ≈9.2%). The Nov-13 $74 straddle is $10.65 (14.4% over 17 days).
Open interest is not in this data feed.

Print date	Earnings-day move
2024-11-12	+10.5%
2025-03-04	+7.1%
2025-05-13	+8.2%
2025-08-12	+19.1%
2025-11-11	−8.2%
2026-03-03	−16.5%
2026-05-12	+13.1%
2026-08-11	+14.6%

the average absolute move on the last 8 prints is 12.2%. The coming print is priced at or below that history, so there is no edge in selling options specifically to bet on the print.

Cash-secured puts by zone

Mid price; yield = mid ÷ strike; assigned basis = strike − mid; annualized = yield scaled to a year.

Zone	Contract	Mid (bid/ask)	Delta	Yield / days	Annualized	Assigned basis	Note
A	Nov-06 70P	1.145 (1.01/1.28)	−0.26	1.6% / 10d	60%	$68.86	Expires before the 11/10 print. The only Zone-A put that respects the gate.
A	Nov-27 70P	3.925 (3.60/4.25)	−0.35	5.6% / 31d	66%	$66.08	Spans the print; would place a Zone-A strike across the event.
A	Dec-18 70P	4.90 (4.65/5.15)	−0.36	7.0% / 52d	49%	$65.10	Spans the print; would place a Zone-A strike across the event.
B	Dec-18 65P	2.995 (2.84/3.15)	−0.25	4.6% / 52d	32%	$62.01	Preferred. Tight market, basis at the 1.272 extension.
B	Nov-27 65P	2.35 (1.86/2.84)	−0.22	3.6% / 31d	43%	$62.65	Wide market.
B	Jan-15 65P	4.00 (3.75/4.25)	−0.27	6.2% / 80d	28%	$61.00	—
C	Dec-18 60P	1.715 (1.51/1.92)	−0.16	2.9% / 52d	20%	$58.29	—
C	Jan-15 60P	2.495 (2.30/2.69)	−0.19	4.2% / 80d	19%	$57.51	—
C	Dec-18 55P	0.91 (0.80/1.02)	−0.09	1.65% / 52d	11.6%	$54.09	—
C	Jan-15 55P	1.35 (1.05/1.65)	−0.12	2.5% / 80d	11.2%	$53.65	—

Dec-18 $65 put: you are paid $2.995 per share now. If SE is below $65 on 2026-12-18 you buy at $65, so your real cost is $65 − $2.995 = $62.01, which is 1.37x sales. If it stays above $65, you keep the $2.995: 4.6% on the cash set aside, over 52 days.

Mechanics

Nov-27 markets are 13–40% wide; Dec-18 and Jan-15 are tight (for example 65P 2.84/3.15). Use limit orders at mid.
Roll at 21 days to expiry or at 50–60% of max profit.
Flow note (Yahoo Finance volumes, 10/26): 1,100 contracts in Nov-20 65P, 2,038 in Dec-18 130C. Direction unknown.

v1 wheel trades graded (hypothetical, per the v1 plan)

Wins: Sep-18 95P @4.13 expired out of the money (9/18 close 101.79): full credit, 4.3%. Aug-21 100P @4.53 / 90P @1.80 / 85P @1.24 all expired out of the money (8/21 close 117.53): full credit.
Loss: Dec-18 80P @3.75 is now 10.35 mid (10.00/10.70), a mark-to-market loss of −$6.60/share (−176% of the credit). Assigned basis $76.25 vs spot $74.01 is −2.9%. It is already past any normal stop-loss. Since the exit rule fired: buy to close, or roll down and out to the Jan-15 70P or 65P for a credit, if available. Estimate: close at 10.35 vs the Jan-15 70P at 5.775 is a −4.58 debit, not a credit roll. Plainly: take the loss or accept assignment at $76.25. Rolling cannot be done for a credit.

Covered calls at the new trim bands (for assigned or held shares; yield = mid ÷ spot)

Band	Contract	Mid (bid/ask)	Delta	Yield / days	Annualized
T1	Dec-18 85C	3.425 (3.05/3.80)	0.33	4.6% / 52d	32%
T1	Dec-18 87.5C	2.775 (2.57/2.98)	0.29	3.7%	26%
T1	Jan-15 85C	4.475 (4.25/4.70)	0.37	6.0% / 80d	28%
T1	Jan-15 87.5C	3.825 (3.60/4.05)	—	—	24%
T2	Dec-18 92.5C	1.895	0.22	2.6%	18%
T2	Dec-18 95C	1.51 (1.23/1.79)	0.18	2.0%	14%
T2	Jan-15 95C	2.39 (2.17/2.61)	0.23	3.2%	15%
T3	Dec-18 100C	1.05 (0.76/1.34)	0.13	—	10%
T3	Jan-15 100C	1.745 (1.57/1.92)	0.18	—	11%
Runner	Jan-15 110C	0.92 (0.74/1.10)	0.10	—	5.7%

Holders at the v1 blend (~$88.8) should not sell T1 calls: they lock in a loss. T2 at $95 is the lowest strike above that cost.
Do not sell Nov-13 or Nov-27 calls into the print if you want the upside. Nov-27 85C is 2.50 mid but a 1.90/3.10 market.

The plan in one paragraph

Sell the v1-plan shares. Before the print, the preferred trade is the December $65 put. It pays about $3 per share and buys at about $62 if assigned, which is 1.37x sales. The pre-print $70 put is acceptable only in the Nov-06 expiry. After 2026-11-10, buy stock only if the print passes and price reclaims about $87. Even then, buy only on a pullback to $78–82, and never above $88. A weekly close under $55 stops all buying until the case is rebuilt.

Risks: ranked by what kills the thesis

In plain English

This is the checklist for the next quarter. Each line says what to watch, what counts as good or bad, what to do, and when we will know. v1 broke rules partly because some rules were badly specified. Those are rewritten here.

#	Risk	What it is · check by
1	TikTok Shop takes share for good (was #1)	Share fell to 55% and Thailand is near a flip. Kill signal unchanged: regional share below 50%, or a take-rate cut. Check: 2026-11-10 (take rate). The next Cube or Momentum Works share report is undated.
2	NEW: capital allocation (Sea Cloud, Monee listing)	Spending with no budget, and a possible carve-out under super-voting control. Check: 2026-11-10.
3	Profit not keeping up with sales (was #3)	Net income growing at a fifth of the revenue rate; tax rate 35%. Check: 2026-11-10.
4	Monee credit cycle (was #2, now lower)	Reserves are now keeping pace with the book, but the mix is riskier. Check: 2026-11-10; Indonesia's buy-now-pay-later cap from 2027.
5	Indonesia country risk and currency (was #5 and #7, merged)	MSCI warning, rupiah near its record low. Check: MSCI decision (date unknown); Q3 currency disclosure on 2026-11-10.
6	Valuation and macro (was #4)	The 10-year yield is about 5%, and December has 71% odds of a rate hike. Check: Fed meetings on 2026-10-28 and in December.
7	Free Fire concentration (was #6)	Bookings slowing, deferred revenue drained. Check: 2026-11-10 (bookings).
8	Governance (was #8)	Raised by the Monee carve-out question.
9	Analyst downgrade cycle (was #9)	26 of 28 still say Buy, so there is room to cut.

Validators: signs the thesis is strengthening

Validator	Current	Check by
Shopee GMV ≥ +24% with take rate ≥14.3%	+28.4%, 14.59%	2026-11-10
Shopee adjusted EBITDA ≥$261M a quarter (on pace for $1B)	$255.4M	2026-11-10
Monee loans 90+ days late ≤1.2%, and reserve growth ÷ loan growth ≤1.5x	1.0%, 0.98x	2026-11-10
Group net income growth ≥ one-third of revenue growth	10.6% vs 48.1% (fails today)	2026-11-10
Sea Cloud spending quantified, with capex below 30% of operating cash flow	Not disclosed; Q2 capex 24% of operating cash flow	2026-11-10
New buyback authorization of $1B or more	About $401M left	2026-11-10
Garena bookings ≥ +10% (compare on bookings; the EBITDA definition changes)	+15.5%	2026-11-10
Q3 fund filings: growth-fund selling stops	Q2: 418 added / 305 cut	~2026-11-14
Insiders: no new sales below $100; any open-market buy	0 buys	Rolling (Form 4s)
Brazil: share rising, loan book growing without late-loan spikes	Brazil 15–16% of Monee loans	2026-11-10

Breakers: rewritten, each names its action

Breaker	Action	Status	Check by
Shopee regional share below 50%, or any take-rate cut of more than 30 basis points vs the prior quarter	EXIT	Not triggered (55%; take rate up)	2026-11-10
Loans 90+ days late above 2%	EXIT	Not triggered (1.0%)	2026-11-10
REPLACES "loan-loss charges > Monee adjusted EBITDA": (a) reserve growth ÷ on-book loan growth above 1.5x for two straight quarters, or (b) Monee adjusted EBITDA down vs the prior quarter for two straight quarters, or (c) annualized loan-loss charge above 28% of average on-book loans	EXIT	Not triggered (0.98x; $288.0M up vs Q1; 23.6%)	2026-11-10
Subsidy war (clarified): group sales and marketing growing faster than revenue for a third straight quarter, while GMV growth slows and value-added-services revenue falls again	EXIT	Two of three quarters already (letter triggered in Q2)	2026-11-10
NEW: open-ended Sea Cloud: a 2027 capex guide above $1.5B with no outside revenue or contracts, or Q3 capex above $500M	EXIT	Pending disclosure	2026-11-10
NEW: Monee carve-out against minority holders: Sea sells more than 25%, or loses control or consolidation	EXIT	Pending	When announced
Free Fire ban in a top-3 market (Indonesia, Brazil)	EXIT	Not triggered	Rolling
Jakarta rule that hits local platforms rather than cross-border rivals	EXIT	Not triggered	Rolling
Forrest Li sells stock outside the disclosed 2025-09-10 plan, or below $100	EXIT	Plan sales at $120 or more triggered v1's broader rule on 2026-08-11 (honored, Verdict tab)	Rolling (Form 4s)
Price: weekly close below $55	STOP ADDING, re-underwrite	$74.01 (not an exit)	Weekly
Price: death cross confirmed (~2026-11-02)	CAP SIZE	Pending. Half of the v1 tranche until a weekly close above the 200-day average (≈$99)	~2026-11-02

from now on, the chart decides how much to buy, and the business decides whether to own it at all.

Next review

Interim check right after the Q3 print on 2026-11-10 (date unconfirmed by Sea).
Full refresh to v3.0 within a week of that print.
Death-cross check about 2026-11-02.
Q3 fund filings by about 2026-11-14.

§14 · §17 · CHANGELOG, OPEN ITEMS & SOURCES · data as of 2026-10-26

Changelog: a living document

In plain English

Every version of this report, newest first. Below that: what we still do not know and when we expect to find out, then every public source used. The two biggest unknowns are new: how much Sea Cloud will cost, and how a Monee listing would be structured.

v2.0 — 2026-10-26 (Liquid Wheel Research · quarterly refresh)
Stage-12 quarterly refresh after Q2 2026 (filed 2026-08-11) and the 2026-10-20 anniversary memo. Tracker graded every v1 claim (business claims mostly KEPT; price-path and entry calls mostly MISSED; capital-allocation and insider claims BROKEN). Ten v1 errors corrected, including the earnings-beat basis (now 2 beats / 6 misses like-for-like) and Li's own stake ($3.5B, not $10.96B). 8-quarter table rolled to Q3 2024–Q2 2026. Checklist re-graded to 9 PASS / 3 WATCH / 4 FAIL (insiders PASS → FAIL, efficiency PASS → WATCH). Moat 7/10, management 7/10. Scenarios re-run to $46 / $109 / $203; 15 priced downside scenarios added. v1 exit rule (weekly close below $75) FIRED and honored. New half-size plan: puts only before the 2026-11-10 print (Dec-18 $65 and $60; Nov-06 $70 only), post-print zones $69–72 / $62–66 / $55–58, no-chase $88, trims $85–88 / $93–95 / $100–104, stop-adds below $55. Breakers rewritten so each names its action. Verdict: 6.0/10 — IMPAIRED, ON WATCH. Classification: CONVICTION-SPEC (moat 7 < 9). Data as of close 2026-10-26 ($74.01).

Next scheduled review: interim check after Q3 2026 earnings (2026-11-10, unconfirmed); full v3.0 within a week of that print.

v1.0 — 2026-07-12 (Liquid Wheel Research · 5-agent deep-dive team)
Initial deep-dive: 16-gate checklist (11 PASS / 2 WATCH / 3 FAIL), three-engine flywheel map, 8-quarter financial longitudinal with GAAP-NI-vs-OCF bridge and bank-adjusted FCF, skin-in-the-game score (max), earnings tracker, dilution-adjusted FY2027 scenarios ($52 / $124 / $211), entry ladder ($95–100 / $86–91 / $78–83) with $116 no-chase line, wheel/CSP strikes, first-principles KPIs, known-unknowns ledger. Verdict: 7.5/10 — thesis INTACT, entry ON WATCH. Classification: CONVICTION-SPEC (moat 7.5–8 < 9 gate); size below the 10% Tier-1 cap. Data as of close 2026-07-10 ($111.14).

Next scheduled review: Q2'26 earnings (~mid-Aug 2026) — the reinvestment-vs-margin print is the live catalyst.

Known-unknowns ledger

Answered since v1

Earnings-estimate basis. The "actual" figures in common data feeds are adjusted, not GAAP: Q2 $0.84 vs GAAP $0.70, Q1 $0.81 vs GAAP $0.67. The record is 2 beats / 6 misses.
Class B votes. 15 per share (20-F, filed 2026-04-17).
Li's ownership. 104.08M shares headline (16.0% economic, 57.6% of votes), of which about 47.9M are his own (Management tab).
Executive cash pay. $3.13M in 2025 for the whole group of directors and officers (20-F Item 6.B). The per-person split is undisclosed.
Q2 earnings date. 2026-08-11, before the market opened.
2026-03-03 catalyst. The Q4 2025 earnings release.
Short interest. 10.24M shares / 3.12%, about 2026-09-30.
Whether Li sells. Yes: 1,139,734 shares under a pre-arranged plan, 2026-08-11 to 08-28.

Still open (carried from v1)

Does Sea carry first-loss on the $1.1B of loans placed with partners? The 20-F is silent on keyword search.
Free Fire's exact share of Garena. Only "significant majority" is disclosed.
2025 capex is still derived. Q1 2025 capex is blank on stockanalysis.
Constant-currency growth. Undisclosed.
The Kaesang private-jet allegation. Not checked.
Sea's own valuation history (price-to-sales at past peaks and troughs). Not sourced.
The hedge-fund holder count for Q2.
Q4 2025 GMV and Shopee's 2025 adjusted EBITDA. Needed for exact full-year paths.
Online-shopping penetration and runway figures.

New this quarter

Sea Cloud: capex and running-cost budget, revenue model, AI-chip (GPU) sourcing, and whether Q2's $356M capex jump is related. Check: Q3 call, 2026-11-10.
Monee listing: venue, size, timing, and whether Sea keeps control and consolidation. Check: 2026-11-10. DBS expects nothing before 2027.
Convertible notes: confirm the $996.3M cash repayment on 2026-09-15 and post-maturity cash and debt. Check: Q3 filing.
Buyback: about $401M left (derived; one unverified source shows $599.73M used to date, which fits). Expiry date unknown. Any new program? Check: 2026-11-10.
Current share count after the Q2 and Q3 buybacks.
Restated Garena adjusted EBITDA under the new definition.
How much is left on the Li and Ye trading plans.
Analyst actions: the New Street downgrade (date and target conflict) and DBS's $85 cut, both single-sourced.
Cause of the 2026-10-19 drop (−5.65%): three competing explanations.
Q3 earnings date: 2026-11-10 per Yahoo Finance. No company notice yet; last year's came 10-27.
Indonesia's 40% debt-to-income cap on buy-now-pay-later from 2027: confirm the regulation and its start date.
Nubank's −17% on 2026-10-26: a Brazil credit read-across to Monee?
2026 regional market shares from Momentum Works and Metric. Indonesia's 54% / 38% is a 2025 figure.
Q2 call transcript, verbatim. Only a summary was read. Within the call, Monee growth was cited as both 52% and 62.5%.
The Baillie Gifford 2026-10-21 filing: position read, but the narrative was not.
CPNG's +3.9% revenue growth on Yahoo Finance looks wrong.
The US insider-reporting start date for foreign issuers (inferred as March 2026, statute not verified).
The 2026-10-20 closing price: $80.43 (Yahoo Finance) vs $80.17 (StockStory).

most of the open questions get answered on one day, 2026-11-10. That is why the plan buys no stock before it.

Sources

Public documents and URLs only. Tap a group to open it.

Company filings SEC EDGAR · Sea Limited · company ID 1703399

Q2 2026 results (6-K 0001193125-26-344596, 2026-08-11): cdn.sea.com · Sea Second Quarter 2026 Results
Q1 2026 results (6-K Ex-99.1): sec.gov · d78490dex991.htm
Li's memo (6-K Ex-99.1, 2026-10-20): sec.gov · d181814dex991.htm
Annual-meeting notice (6-K, 2026-08-25): sec.gov · 000119312526364238
2025 annual report (20-F, filed 2026-04-17): sec.gov · ef20067274_20f.htm
Baillie Gifford 13G/A (2026-10-21): sec.gov · primary_doc.xml
2021 convertible-note terms (424B5): sec.gov · ny20000593x4_424b5.htm
Filing index: data.sec.gov · CIK0001703399.json
About 1,325 Form 4 insider-trade line items, 2026-03-20 to 2026-10-23 (for example, a Li sale: sec.gov · ownership.xml). Feng option exercise: Form 4 0001193125-26-429381.
Prior quarterly filings used in the 8-quarter table: 0001193125-25-118807, -25-178935, -25-275616, -26-088151, -26-219378.
Q3 2025 release: businesswire.com · Sea Limited Reports Third Quarter 2025 Results
Q4 2025 figures: seekingalpha.com/pr/20421038
Buyback program (November 2025): businesswire.com · Sea Limited Announces Share Repurchase Program

Earnings call transcript summary

Q2 2026 call transcript summary (Motley Fool): fool.com · SE Q2 2026 earnings call transcript

Market data prices, estimates, options

stockanalysis.com: statistics · forecast · quarterly financials · quarterly cash flow
Yahoo Finance: SE daily and weekly prices 2021 to 2026-10-26; estimates, price targets, recommendations, short interest. Peers: GRAB, MELI, CPNG, BABA, CRWV, KWEB, QQQ, EEM, EIDO, NU, GOTO.JK. Currencies and rates: IDR, PHP, THB, VND, MYR, TWD, BRL, 10-year Treasury yield, VIX.
TradingView options chains (OPRA), SE expiries 2026-11-06 to 2027-01-15, close 2026-10-26.
CNBC / CME FedWatch rate odds (2026-10-23).

News and analysis articles and wire items

DBS downgrade: itiger.com · ad-hoc-news.de
DBS on Monee (BigGo): finance.biggo.com
Dow Jones (via TradingView), 2026-10-20: Sea plans new cloud business, weighs financial-services arm IPO
Dow Jones Market Talk, 2026-10-19: Shopee's earnings could be hurt by currency weakness
Dow Jones Market Talk, 2026-10-22: Sea's cloud segment could weigh on profit · Sea Cloud could break even in three to four years
StockStory, 2026-10-20: Sea (SE) stock trades down, here is why
Seeking Alpha headline, 2026-10-19: Insider selling weighs on Sea Ltd
Yahoo Singapore: Shopee parent launches cloud business, weighs listing of fintech arm
TradingKey, 2026-10-20: tradingkey.com
Motley Fool, 2026-10-23: Big changes could be coming to Sea Limited stock · This all-star stock just plunged
Tickeron (attribution disputed): tickeron.com
GuruFocus, 2026-10-26: gurufocus.com
Quiver Quant: quiverquant.com
MarketBeat: marketbeat.com
Insider trackers: massive.com · secform4.com

Competition market-share research

Cube, 2026-10-13: cube.asia · Shopee growth strategy
Cube, 2026-06-02: cube.asia · Shopee, Lazada and TikTok Shop in 2026
China Digital Retail Report, TikTok Shop Watch #6 (2026-08-19): chinadigitalretailreport.substack.com
KrASIA: kr-asia.com
DealStreetAsia (Momentum Works 2026): dealstreetasia.com

Indonesia / MSCI wire reports

The Star / Reuters, 2026-10-16: thestar.com.my
Euronews: euronews.com

Prior version & framework

SE Deep-Dive v1.0 (2026-07-12) and its sources (Bloomberg, SCMP, Momentum Works, digitalinasia, McKinsey, BPS Indonesia, stocktitan, Insider Monkey, Glassdoor).
Liquid Wheel Research · Deep-Dive Framework, Stages 1–12.

Disclosure & disclaimer: This report is education and personal research, not financial advice. The author may hold positions in SE shares and/or options. All profit-and-loss figures on this page grade the published v1 plan hypothetically. Numbers from SEC filings, company releases, and market data as of 2026-10-26 close — verify before acting. Past performance doesn't guarantee future results. Do your own research — that's rather the point.

§18 · GLOSSARY · data as of 2026-10-26

Glossary

In plain English

Every term, abbreviation and ratio used in this report, defined in one sentence, with Sea's current value where one exists. Anywhere on the page, a word with a dotted underline is in this list: hover over it, or tap it on a phone, to see the definition. Tags: CONFLICT means two research reviews disagree; UNVERIFIED means the figure has a single source or is derived.

Term	Plain-English definition	Sea's value
10b5-1 plan (pre-arranged trading plan)	A plan set up in advance that sells an insider's shares automatically on preset terms, so the timing is not a fresh decision.	Li's plan was adopted 2025-09-10 and sold 1,139,734 shares at $120–132.
10-year Treasury yield	The interest rate on 10-year US government debt. It sets the bar every stock must beat.	About 5.00% on 2026-10-23.
13F / 13G filing	Reports where big funds disclose their stakes, either every quarter (13F) or when they cross an ownership threshold (13G).	Baillie Gifford's 2026-09-30 filing shows 37.15M shares (6.55%).
200-week moving average	The average of the last 200 weekly closing prices, about four years. Long-term buyers often defend it.	$93.98, about 21% above today's price.
20-F	A foreign company's annual report to the US Securities and Exchange Commission (SEC).	Sea's 2025 20-F was filed 2026-04-17.
6-K	A foreign company's interim filing with the SEC, used for quarterly results and news.	Sea's Q2 results and the 2026-10-20 memo were 6-Ks.
Adjusted EBITDA	Profit before interest, tax, depreciation and some non-cash items such as stock pay, as defined by the company. Roughly cash operating profit.	$917.2M in Q2 2026 (+10.6%).
Adjusted (non-GAAP) earnings per share	Earnings per share with items such as stock-based pay removed. Wall Street estimates are usually on this basis.	$0.84 in Q2 2026 vs a $0.86 estimate.
Allowance for credit losses (loss reserve)	Money Monee has set aside on its balance sheet for loans it expects to go bad.	$1,114.8M, 11.1% of on-book loans.
Annualized yield (options)	The option premium divided by the strike, scaled up to a full year so trades of different lengths compare fairly.	The Dec-18 $65 put yields 4.6% in 52 days, about 32% a year.
Assignment / assigned basis	Being required to buy the shares when a put you sold finishes below its strike. Your cost is the strike minus the premium you were paid.	$62.01 on the Dec-18 $65 put.
ATR-14 (average true range)	The average daily price range over the last 14 trading days, in dollars.	$3.66, about 5% of the price.
Bank-adjusted free cash flow	Operating cash flow minus spending on buildings and equipment minus new loans made. The honest cash number for a company that is part bank.	About −$644M in Q2 2026 (Q1: −$142M).
Basis point	One-hundredth of a percentage point. 100 basis points = 1 percentage point.	The take rate rose about 195 basis points (1.95 points) in a year.
Bollinger bands	A price channel drawn two standard deviations above and below the 20-day average. Touching the lower band means the price is stretched to the downside.	Upper $107.51, lower $73.71, where the stock sits.
Bookings (Garena)	Money players spent in Garena's games in the quarter, before accounting spreads it out as revenue.	$763.5M in Q2 2026 (+15.5%).
Breaker	A pre-written rule that says the investment case is broken. In v2 each breaker names its action: exit, or stop adding.	Several v1 breakers fired; see the Trade Plan tab.
Buyback	The company buying its own shares, which shrinks the share count so each remaining share owns more.	$416.8M in Q2; about $401M of the $1B program left.
Capex (capital expenditure)	Spending on buildings, servers and equipment.	$356M in Q2 2026, 2.6x Q1.
Capitulation	A burst of panicked selling on huge volume, when holders give up. It often marks a bottom, but not always.	Not confirmed yet: the 2026-10-26 low was a new low.
Carve-out	Listing part of a subsidiary on the stock market while the parent keeps the rest.	A possible Monee listing.
Cash-secured put (CSP)	You are paid now to promise to buy 100 shares at the strike price, with the cash set aside in case you have to.	The plan's preferred trade is the Dec-18 $65 put at about $3.00.
CCC bond yield	The yield on the lowest-rated junk bonds, a gauge of stress at the weakest borrowers.	13.40%, +228 basis points.
Class A / Class B shares	Two share types. Class B carries 15 votes each; Class A carries one.	Li controls 57.6% of the votes.
Conglomerate discount	The lower price investors pay for a company that runs many unrelated businesses.	Scenario 8 assumes 15%.
Consensus	The average Wall Street analyst estimate.	Q3 2026: revenue $8.07B, adjusted earnings per share $0.97.
Constant currency	Growth with exchange-rate swings removed, so you see the real business change.	Sea does not disclose it.
Convertible notes	Bonds the holder could swap for shares at a set price.	$996.3M matured 2026-09-15 with a conversion price near $477, so cash repayment is presumed.
Conviction-spec	The framework's sizing tier for a company whose moat scores below 9/10. Positions are smaller than the top tier.	Sea's tier; the new position is half of v1's size.
Covered call	You are paid now to promise to sell shares you own at a higher strike price.	For example, the Jan-15 $95 call at $2.39.
CPI (consumer price index)	The main US inflation measure.	Headline 3.7%.
Current ratio	Short-term assets divided by short-term bills. The framework wants 2 or more.	24,052.8 ÷ 16,094.4 = 1.49.
Days to cover	Shares sold short divided by average daily trading volume: how many days short sellers would need to buy back.	2.9 days.
Death cross / golden cross	The 50-day average crossing below (death cross) or above (golden cross) the 200-day average. Traders read it as a trend change.	A death cross is expected about 2026-11-02.
Deferred revenue	Cash collected for things not yet delivered. It is counted as revenue later.	Garena's build was only +$16.9M in Q2 2026 (Q1: +$254.8M).
Delta (options)	Roughly the probability an option finishes in the money.	The Dec-18 $65 put is −0.25: about a 1-in-4 chance of assignment.
Diluted shares	The share count including options and convertible securities that could become shares.	About 634.5M.
Dilution	Growth in the share count, which shrinks each share's slice of the company.	Diluted shares fell 0.3% year on year in the first half of 2026.
DTE (days to expiry)	Days until an option expires.	The Dec-18 contracts had 52 days left on 2026-10-26.
Duopoly	A market dominated by two big players.	Cube, first half 2026: Shopee 55%, TikTok Shop 39% of the big three.
Effective tax rate	Tax paid divided by pre-tax profit.	35.0% in Q2 2026 vs 25.2% a year earlier.
EIDO / QQQ	Exchange-traded funds that track Indonesian stocks (EIDO) and the Nasdaq-100 (QQQ).	QQQ −1.7% since 2026-07-10.
Enterprise value (EV)	Market cap plus debt minus cash: the price to buy the whole company, debts and all.	$40.36B (or $35.70B on a narrower definition).
Enterprise value / sales	Enterprise value divided by yearly revenue.	$40.36B ÷ $27.72B = 1.46x.
Escrow and deposits payable	Cash Sea holds for buyers, sellers and bank depositors. It sits on Sea's balance sheet as a short-term bill.	Deposits $4,285M, escrow $3,330M.
Fed funds rate	The Federal Reserve's main policy interest rate.	3.75–4.00%.
Fibonacci retracement	A price level at a set fraction (0.236, 0.382, 0.5, 0.618, 0.786, 0.886) of a past move. Traders watch them as support or resistance.	The 0.886 level of the $55 → $199.30 move is $71.45.
Flywheel	A loop where each step makes the next one stronger, so the business speeds itself up.	Sea: games → mall → delivery → lending → profits back in.
Form 4 / Form 144	SEC reports of an insider's trade (Form 4) or a planned sale (Form 144).	About 100 Form 4s since v1.
Forward P/E	Share price divided by next year's expected earnings per share.	$74.01 ÷ $5.10 (2027 estimate) = 14.5x.
Free cash flow	Operating cash flow minus spending on buildings and equipment: the cash left over.	$1,150M in Q2 2026; $4,423M over the last four quarters.
Free Fire	Garena's flagship mobile game, built for cheap Android phones.	A "significant majority" of Garena's revenue and profit (20-F).
FX (foreign exchange)	Changes in exchange rates. A stronger US dollar shrinks the dollar value of sales made in local currencies.	The dollar averaged 9.1% stronger vs the Indonesian rupiah in Q3.
GAAP	US generally accepted accounting principles: the standard rules for reported profit.	Q2 2026 GAAP diluted earnings per share: $0.70.
Garena	Sea's game-publishing arm.	Q2 2026 bookings $763.5M (+15.5%).
GDPNow	The Atlanta Fed's real-time estimate of current-quarter economic growth.	+3.5%.
GMV (gross merchandise value)	The total value of goods sold on Shopee, before Shopee's cut.	$38.3B in Q2 2026 (+28.4%).
Gross margin	Revenue minus direct costs, as a share of revenue.	45.6% in Q2 2026.
High-volume node	A price range where many shares changed hands. It tends to act as a floor or a ceiling later.	$72.5–75 (105M shares traded since 2024).
High-yield (junk-bond) spread	The extra yield risky company bonds pay over US government bonds. A rising spread signals credit stress.	3.40%, up 0.60 points in 20 days.
Implied volatility (IV)	The yearly price swing the options market is pricing in. Higher means options pay more.	Dec-18 at-the-money options imply 63.6%.
Investment-grade spread	The extra yield safer company bonds pay over US government bonds.	0.89%.
IPO / listing	Selling a company's shares to the public for the first time.	Li's memo said Monee "may well ring that bell."
ISM surveys	Monthly business surveys. Above 50 means the sector is growing.	Manufacturing 54.5, services 54.9.
KPI (key performance indicator)	A number tracked to judge progress.	Eight first-principles KPIs on the Valuation tab.
Ladder / tranche	Buying in several planned pieces at different prices instead of all at once. A tranche is one planned block of money.	v2 zones A/B/C at 25/35/40% of a halved tranche.
Leases (as debt)	Rent owed on long-term property leases, counted as debt under accounting rules.	$1,905.9M.
Loans 90+ days late (NPL90+)	Non-performing loans: loans more than 90 days behind on payments, as a share of the loan book.	1.0% at Monee.
MACD	A momentum indicator built from two moving averages. Below its signal line and falling means downward momentum.	Daily −7.20 vs signal −5.37, still falling.
Market cap	Share price times shares outstanding: what the stock market says the company is worth.	$74.01 × 612.48M = $45.33B.
Mid price (options)	Halfway between the bid (what buyers pay) and the ask (what sellers want). Use limit orders near it.	Dec-18 $65 put: 2.995 (2.84 / 3.15).
Moat	What protects a business from competitors taking its customers.	7/10 for Sea (the framework's full-size bar is 9).
Monee	Sea's fintech arm: consumer loans, digital banks and wallets (formerly SeaMoney).	Q2 revenue $1,402.8M; loan book $11.1B.
MSCI	The company whose stock indexes guide trillions of dollars of fund money.	Warned on 2026-10-16 it may drop Indonesia from emerging-market indexes.
Multiple	Any price ratio, such as price-to-earnings. "The multiple shrank" means investors pay less per dollar of earnings or sales.
Net cash	Cash and investments minus debt (excluding leases).	About $9.63B.
Net margin	Net income divided by revenue: cents of profit per dollar of sales.	$458.1M ÷ $7,787.8M = 5.9% in Q2 2026.
No-chase line	A price above which you do not start buying, because the reward no longer justifies the risk.	$88 in v2 (v1: $116).
Off-book loans	Loans Monee arranges but places with partner lenders, so they are not on Sea's balance sheet.	$1.1B of the $11.1B book.
Open interest	The number of option contracts currently open at a strike. Not in this refresh's data feed.
Operating cash flow	Cash generated by day-to-day operations.	$1,506.0M in Q2 2026.
Operating margin	Operating profit divided by revenue.	8.35% in Q2 2026 vs 9.27% a year earlier.
Oversold	A price that fell so far, so fast that a bounce is statistically more likely. It is not a buy signal on its own.	Daily RSI 17.7.
PEG	Price-to-earnings divided by the expected earnings growth rate. Below 1 suggests growth is cheap.	0.73.
P/E (price-to-earnings)	Share price divided by earnings per share.	Trailing: $74.01 ÷ $2.59 = 28.6x.
P/FCF (price-to-free-cash-flow)	Market cap divided by free cash flow.	$45.33B ÷ $4.42B = 10.2x.
PM	Portfolio manager: the person who makes the final sizing and trade calls. "PM" marks that judgment.
Price gap	A price range the stock jumped over without trading, for example between one day's low and the next day's high. Gaps often get "filled" later.	Unfilled overhead: $86.43–87.88 and $93.09–93.49.
Price target	An analyst's 12-month price estimate.	Consensus average $150.04; low $70.
Probability-weighted value	Each scenario's target times its odds, added up.	0.15 × 203 + 0.50 × 109 + 0.35 × 46 ≈ $101.
Provisions (loan-loss charges)	The quarterly expense to top up the loss reserve for loans expected to go bad. A set-aside, not cash spent.	$555.2M in Q2 2026 (+71.5%).
P/S (price-to-sales)	Market cap divided by yearly revenue: what you pay for each $1 of sales.	$45.33B ÷ $27.72B = 1.64x.
QAU / QPU / ABPU (Garena)	Quarterly active users, quarterly paying users, and average bookings per user.	666.3M / 68.1M / $1.15.
Realized volatility (RV)	How much the stock actually moved, annualized.	30-day 40.6%, 60-day 50.4%.
Red team	A reviewer whose job is to argue the bear case as hard as possible.	Bear range $46–55.
Restricted cash	Cash Sea holds but cannot freely use, such as customer funds.	$2.46B, excluded from cash here.
Revenue per share	Yearly revenue divided by shares outstanding. Multiply by a price-to-sales ratio to get a price.	27,723 ÷ 612.48 = $45.26.
Roll (options)	Closing one option and opening another with a later date or a different strike.
RSI-14 (relative strength index)	A 0–100 momentum gauge over 14 days. Below 30 is oversold; below 20 is extreme.	Daily 17.7, the second-lowest since 2023.
Sales and marketing (S&M)	Spending on ads, promotions and subsidies to win buyers and sellers.	+64.5% in Q2 2026 vs revenue +48.1%.
Sea Cloud	Sea's new AI cloud business for outside customers, announced 2026-10-20.	No budget, revenue or timeline disclosed.
SEC	The US Securities and Exchange Commission, which receives company and insider filings.	Sea's company ID is 1703399.
Short interest	Shares sold short by traders betting the price falls.	10.24M (3.12%), down 44% since v1.
SMA (simple moving average)	The average closing price over a set number of trading days (20, 50 or 200).	50-day $102.48; 200-day $99.69.
SPayLater	Monee's buy-now-pay-later credit, increasingly used outside Shopee.
SPX Express	Shopee's in-house delivery network.	Instant-delivery orders +~80%, cost per order −~20%.
Stock-based compensation (SBC)	Pay given in shares. A real cost that dilutes owners.	$131.9M in Q2 2026, the 6th straight decline.
Straddle	Buying a put and a call at the same strike: a bet on a big move either way. Its price shows the move the market expects.	Nov-13 $74 straddle: $10.65 (14.4%).
Strike	The price at which an option lets you buy or sell.
Sum-of-parts	Valuing each business separately and adding them up.	DBS puts Monee alone at $18–22B.
Super-voting	Shares with more votes than their economic weight.	Li: 57.6% of votes on about 7.8% own economic stake.
Support / resistance	Price areas where buying (support) or selling (resistance) has tended to show up before.
Take rate	Shopee's revenue as a share of GMV: Shopee's cut of every dollar sold.	14.59% in Q2 2026 vs 12.64% a year earlier.
TAM (total addressable market)	The whole market a company could sell into.	About $190B a year for the big three Southeast Asian platforms (derived).
Tier-1	The framework's top sizing tier. It requires a moat score of 9/10 or better.	Sea is not eligible (moat 7).
Trailing twelve months (TTM)	The last four reported quarters added together.	Revenue $27,723M.
Trim	Selling part of a position into strength.	v2's first trim band is $85–88.
Validator	A pre-written sign that the investment case is getting stronger.
Value-added services (VAS)	Shopee's logistics and other fees charged to sellers.	$676.4M, −9.0%, because shipping subsidies are netted off.
VIX	A measure of expected US stock-market swings: the "fear gauge."	19.10.
Weekly close	The Friday closing price. Rules use it so one bad day does not trigger them.	$74.63 on 2026-10-23.
Year over year / quarter over quarter	Comparing with the same quarter last year, or with the previous quarter.	Q2 2026 revenue +48.1% year over year.