Summary The extra money collected overseas in H1 only made up for 70% of the shortfall at home. Overseas is growing, but it can't carry XPeng—the problem is that the domestic base is shrinking.

XPeng's overseas business really is growing, but for now it can't carry the company. The real problem is at home: the base is shrinking.

In H1, overseas revenue rose 62%, yet group revenue fell 3.8%. The market is fixated on the overseas number and getting excited—it's looking in the wrong place. What overseas is doing right now is plugging the hole at home; it isn't a second growth curve yet. If domestic bleeding isn't stopped, no matter how fast overseas runs, it can't fill the gap. Q2 group revenue has already returned to year-on-year growth, but the regional breakdown is only available on a half-year basis, so everything below is based on the H1 numbers.

The group revenue hole is domestic

XPeng's revenue this half-year was dragged down by the domestic business, with overseas merely chasing from behind.

Whether a company's revenue rises or falls depends on how much money came in and how much went out—pretty percentages don't tell the story. Overseas brought in an extra 3.16 billion yuan; domestic lost 4.46 billion yuan. The money overseas clawed back only covers 70% of what was lost at home. The remaining 30% is exactly the slice by which group revenue fell.

Overseas growth is much faster than the domestic decline, so why can't it catch up? Because the two bases are worlds apart. In H1 last year, domestic revenue was 5.7 times overseas. Overseas has a small base—no matter how fast it grows, the absolute money added is limited. Domestic has a big base—a small percentage drop means a lot of money lost. This is determined by the base numbers; it has nothing to do with how hard the overseas team works, and it won't disappear just because overseas sells a few more cars.

The situations on the two sides are also different. Every increment overseas requires opening markets one by one, building channels and service from scratch. Every decrement at home is leakage from territory already held. One side is slowly stacking bricks in new places while the other is losing chunks in old places—the patching speed naturally can't keep up with the leaking speed.

Domestic is XPeng's base. A shrinking base means the revenue that feeds this company is thinning. No matter how fast overseas goes, for now it's just filling in place. As for why domestic is shrinking, the interim report's regional data doesn't break down by model or price, so I won't guess on the company's behalf. The shrinkage is a fact; the amounts are right there. To evaluate XPeng, the order has to be reversed: look at domestic first, then overseas. Any new story the company tells externally must first pass the domestic test.

XPeng H1 2025 vs. H1 2026 domestic and overseas revenue comparison. Data from XPeng's 2026 interim report regional revenue.

H1 group revenue change breakdown: overseas +3.16 billion yuan, domestic -4.46 billion yuan. Author's calculation based on the interim report.

The bulk of revenue is still domestic

Overseas share surging shows XPeng's revenue structure has changed—it doesn't show the company has grown.

The rise in overseas share comes partly from selling more overseas and partly from collecting less domestically, which shrinks the denominator. The latter has nothing to do with overseas achievements. Overseas growth is real, but using the share to prove overseas success credits the domestic contraction to overseas as well. Whether the group has grown must be judged by total revenue; share can't answer that question.

The remaining three-quarters determine that when domestic moves, the group shakes. For overseas to pull the group back to growth, the money added must exceed the money lost domestically. In H1, overseas didn't cross that line. How high that line is depends on how much domestic falls—overseas doesn't get a say. Rising share and falling revenue are happening at XPeng simultaneously.

These three-quarters are also where XPeng most needs to hold steady. Framing overseas expansion as XPeng's main storyline is premature. The main storyline must be the part that carries revenue, and XPeng's part is still domestic. Miss this, and you'll read a company that's still contracting as one that's expanding. For investors, the most useful thing about the overseas share is the reminder of how heavy the domestic weight still is.

In the August delivery news, XPeng said overseas revenue contributed about a quarter and that Q2 overseas sales exceeded 20,000 units for the first time. These numbers are all correct, but they answer how fast overseas is growing. Why domestic revenue is falling—the news doesn't explain. The louder the overseas story, the easier the domestic problem is covered up. Export sales are counted on a wholesale basis and include knockdown kits; once cars are shipped, there's still a way to go before they're sold at the retail end. Using shipment volumes to prove overseas is carrying the company isn't enough evidence.

XPeng revenue by region. Data from XPeng's 2026 interim report.

No profit breakdown disclosed—can't see if it's profitable

Whether overseas actually makes money—outsiders can't see it now. XPeng hasn't disclosed the profit breakdown.

Revenue only shows how much was sold; profit shows whether the business is worth scaling. A growth curve can carry a company only through the money it retains. XPeng's interim report reports only one operating segment; management reports domestic and overseas as one business externally. Regional revenue is broken out, but regional costs and gross profit aren't broken out at all. Revenue split but costs not split—this is exactly how readers of financial statements end up remembering only the good half.

For overseas per-vehicle gross profit, the company only gave a qualitative statement. In the August news, XPeng said the average overseas vehicle price exceeds 40,000 euros, with both per-vehicle revenue and per-vehicle gross profit at leading levels among Chinese automakers going abroad. A high selling price shows the cars command a premium overseas—that's real skill. But average price is ultimately a selling price; there's still cost between it and earning more. Cars must be shipped and pass through tariffs—where these costs land by region, outsiders have no way of knowing. There's no number after "leading"; investors can't verify it.

Cars sold, how much money retained—the statements don't say. The revenue overseas clawed back has an account; the profit clawed back has no account. How much gross profit the domestic revenue lost and the overseas revenue gained each carry—the interim report doesn't break it out, and group gross margin mixes the two together