SINGAPORE, 20 AUG 2026 — Sea Limited reported second-quarter revenue of US$7.8 billion, up 48.1 per cent year on year, with net income of US$458.1 million. Shopee's quarterly gross merchandise value reached US$38.3 billion, and management pointed to US$1 billion of adjusted EBITDA for the year.
The number that describes the company best is not in the e-commerce segment. It is the loan book.
The quarter
Gross profit rose 47.3 per cent to US$3.5 billion. Shopee contributed US$5.6 billion of GAAP revenue, Monee's digital finance revenue grew 58.9 per cent to US$1.4 billion, and Garena's revenue rose 34 per cent to US$747 million on bookings of US$764 million, with Free Fire averaging more than 100 million daily active users.
The results were reported on 11 August.
Revenue up 48 per cent, net income up 11
The gap between those two figures is the first thing to look at. Revenue grew more than four times as fast as net income, and gross profit tracked revenue closely, which locates the divergence below the gross line.
That is the signature of a company spending to grow: sales and marketing, credit provisions, content and platform investment, all rising with the top line. This is not a warning sign on its own; a business capturing share deliberately trades near-term margin for position. It does, however, mean the headline growth rate overstates how much is reaching the bottom.
Adjusted EBITDA at US$917.2 million against US$7.8 billion of revenue is the more useful operating figure, and management's US$1 billion Shopee target for the year is the specific claim to hold them to.
Garena deserves a sentence of its own, because it has been written off repeatedly and keeps not dying. Bookings of US$764 million against revenue of US$747 million means the business is converting close to what it sells in the period rather than living off deferred balances, and a single title averaging more than a hundred million daily users is a distribution asset most consumer companies would pay a great deal for. The concentration risk in that one title is obvious and has been obvious for five years.
The lending business is the one to watch
Monee's loans outstanding rose 62.5 per cent to US$11.1 billion, with revenue up 58.9 per cent. Non-performing loan ratios are described as stable.
An eleven-billion-dollar consumer loan book inside a technology group is a bank-sized exposure without a bank's balance sheet or, in most of its markets, a bank's supervision. Growing the book at 62.5 per cent while holding non-performing loans stable would be a good result, but it is also the hardest claim in the release to verify from outside.
The arithmetic that makes this worth watching is the same one we applied to Skyro reaching break-even in Philippine consumer lending. A book growing at that rate has a falling average loan age, and young loans have not had time to default. A stable ratio during rapid growth is partly a statement about credit quality and partly a statement about the denominator, and only the next few quarters separate the two.
This is not an allegation. These are the typical dynamics of a fast-growing lender, and they explain why vintage-level performance data is more revealing than a simple portfolio ratio.
There is a structural advantage here that a standalone lender cannot replicate, and it cuts both ways. Lending into purchases made on your own marketplace means you can see what the borrower bought, how often they buy, whether they return goods and whether they have paid before. That is better underwriting data than a credit bureau holds in most of these markets. It also means the credit book and the commerce book fail together: a downturn that reduces basket sizes reduces repayment capacity in the same population at the same moment.
What the three segments now are
The group is frequently described as an e-commerce company with a games business attached. That has stopped being accurate.
In this model, Shopee provides the distribution and the customer relationship. Monee monetises that relationship with credit, at margins e-commerce cannot reach, while Garena supplies cash and a second acquisition channel through Free Fire's hundred million daily users.
Read that way, the interesting comparison is not with regional marketplaces. It is with the Chinese platforms that built payments and lending on top of commerce, and with Grab, whose financial services arm follows the same logic from a different starting point. The regional question over the next two years is whether Southeast Asia ends up with two of these or one.
The regulatory position is the unpriced risk
A consumer lending book of this size across multiple jurisdictions attracts supervision, and the supervision is arriving at different speeds in each market.
Indonesia, the Philippines, Vietnam and Thailand have each tightened rules on digital lending in the past two years, on interest caps, collections practice, data use in credit scoring and licensing. The relevant risk to a business like this is rarely a single adverse ruling; it is a gradual narrowing of what the product may charge and how it may collect, applied unevenly across the markets that supply the growth.
This risk does not appear in a quarterly release, but it is the one an investor in the business is actually carrying. Regional regulators have watched consumer credit expand rapidly on platforms they do not directly supervise, and the historical pattern is that they act after a consumer-harm episode rather than before one.
What we could not establish
Anything below the portfolio-level credit figures. The company did not disclose vintage analysis, charge-off rates, provisioning policy, or a geographic breakdown of the loan book. Any of these would show whether the stable NPL ratio reflects strong underwriting or just rapid growth.
Also unestablished: the split of Shopee's revenue between commissions, advertising and logistics; how much Monee lends into Shopee purchases as against general-purpose credit; Garena's dependence on Free Fire specifically; the cost of funding the loan book; segment-level cash generation as against adjusted EBITDA; and the competitive position by market rather than in aggregate.
What to watch
Watch whether the US$1 billion Shopee EBITDA target is met and how. Hitting it through take-rate increases is a different business from hitting it through logistics efficiency, and only one of those is durable against a competitor willing to spend.
Then watch the loan book's growth rate rather than its size. A deceleration would be the first honest test of the credit quality claim, because that is when the average loan age stops falling and the ratio has to stand on its own.
Finally, watch for regulatory action in any single market. The first meaningful cap on digital lending rates in Indonesia or the Philippines would reprice this segment across the region, and it would arrive with little notice.
