--- title: "May Retail Sales: Dark Before Dawn?" type: "Topics" locale: "en" url: "https://longbridge.com/en/topics/41909685.md" description: "Dolphin Research's earlier concerns unfortunately materialized, with May retail sales weaker than Apr. YoY growth in total retail sales fell into negative territory at -0.6%, the first decline in the post-pandemic era.Even before the Sep-24 'big reversal', domestic retail had not been this weak. From a channel perspective, online retail shows signs of bottoming and a rebound, but offline spending deteriorated more sharply (both goods and dining), further slowing the headline growth. By category..." datetime: "2026-06-16T09:00:28.000Z" locales: - [en](https://longbridge.com/en/topics/41909685.md) - [zh-CN](https://longbridge.com/zh-CN/topics/41909685.md) - [zh-HK](https://longbridge.com/zh-HK/topics/41909685.md) author: "[Dolphin Research](https://longbridge.com/en/news/dolphin.md)" --- # May Retail Sales: Dark Before Dawn? Dolphin Research’s earlier concerns have unfortunately materialized: May’s retail sales came in even weaker than Apr., marking the first negative YoY print (-0.6%) in the post-Covid period. Even before the Sep-24 ‘reversal’, domestic retail sales had not looked this soft. From two angles: by channel, online retail shows signs of bottoming and a tentative rebound, while offline consumption deteriorated more sharply across both goods and dining. This divergence pulled the headline growth rate further down. By category, weakness has spread beyond those hit hardest by state subsidies—appliances, furniture, autos—to nearly all segments, including staples like grain & oil and daily necessities, and discretionary such as alcohol & beverages, apparel, and cosmetics. The softness is no longer confined to a few categories. That said, the worst may be now. After May, last year’s subsidy-driven peak and high base roll off, easing YoY pressure. With online retail reacting faster and already showing a bottoming pattern this month, it could provide some leading indication. **1) May retail turned negative** According to the NBS, total retail sales in May fell 0.6% YoY, worse than Apr., extending the step-down trend that began in Feb. We had flagged that Mar–May would face a tough base given last year’s subsidy tailwinds, and the outcome unfortunately came in even weaker than expected. By consumption type, both goods and dining growth slowed in tandem, with goods seeing weaker absolute growth and dining posting a sharper sequential deceleration. Specifically, goods retail contracted 0.7% YoY vs. -0.1% in Apr., while dining growth dropped from +2.2% to just +0.6%. Fortunately, the base for both goods sales and dining eases notably after Jun. Subsidy benefits faded meaningfully after Jun last year, and dining growth also plunged thereafter amid the food-delivery price war. With the base rolling off, retail sales growth may start to bottom from Jun. **2) Online bottoms, offline keeps weakening** Despite the overall slowdown, the NBS data show an improvement in online retail. May online physical goods GMV rose 2.6% YoY, a clear rebound vs. Apr.’s 0.2%. Because this year’s 618 promotion window is similar to last year and platforms have not rolled out high-profile subsidy campaigns amid tighter scrutiny of cut-throat competition, the rebound does not look like a front-loaded promo effect. The recovery appears more fundamental than seasonal. Hence, the main drag likely came from offline. By our estimate, offline goods retail fell 4.5% YoY in May vs. -1.2% in Apr.; given periodic historical revisions, these calculated rates are for reference only. (Note: from 2026, the NBS adjusted the total online retail definition, changing ‘online retail sales’ to ‘online goods & services retail sales’, materially expanding coverage of online services. The statistical scope for online physical goods remains largely unchanged and is comparable with historical data.) **3) Softness spreading from subsidy-linked & autos to broad categories?** In recent months, autos and subsidy-linked home appliances, furniture, and 3C products have been the biggest drags. A change this month is that while these remain the weakest categories in absolute terms, their marginal deterioration vs. last month was limited (smartphones aside). Meanwhile, above-scale sales growth in staples such as grain & oil and daily necessities, and in non-subsidy discretionary like tobacco & alcohol and cosmetics, all saw notable MoM slowdowns. The weakness is shifting from structural to broad-based. The downtrend is becoming more pervasive. From this angle, appliances and furniture show signs of peak pressure from subsidies, still under the most strain but with scope for marginal improvement. Smartphones and digital devices, however, may still be heading toward their trough given memory price hikes and a later subsidy impact last year. As for apparel, F&B and daily necessities, they are likely in the early-to-mid stages of the spillover. While the absolute hit should be limited, their growth troughs may arrive later. Risk disclosure and statement: [Dolphin Research disclaimer and general disclosure](https://support.longbridge.global/topics/misc/dolphin-disclaimer) ## Comments (10) - **amos · 2026-06-16T15:55:52.000Z**: Follow - **相信概率 · 2026-06-16T10:49:15.000Z**: It's just started, and I was still thinking about dawn😹 - **石来运转 · 2026-06-16T09:20:06.000Z · 👍 4**: It's even more difficult in the short term now, with AI laying off people every day. The Boston University prediction of falling into an artificial layoff trap might come true. Layoffs save costs for companies, but those laid off are also customers. Without income, customers naturally reduce related - **Dolphin Research** (2026-06-17T01:52:29.000Z): AI layoffs will affect residents' income and consumption capacity, further suppressing the recovery of offline and discretionary consumption. Current consumption is already weak, and pressure on the i - **inron** (2026-06-19T02:17:40.000Z): Mainly, the capital has made a huge profit, and this portion of wealth has been transferred. - **维港自由艇** (2026-06-19T05:56:39.000Z): Haha, the impact of AI hasn't even arrived domestically yet, but employment has long been paralyzed... - **RichR · 2026-06-16T09:14:24.000Z · 👍 1**: Which country's data are you talking about??? - **Dolphin Research** (2026-06-16T10:58:19.000Z): Look at the data objectively, okay? - **千年伯爵 · 2026-06-16T09:10:30.000Z · 👍 1**: Unfortunately, it came true as predicted. - **Dolphin Research** (2026-06-16T10:56:52.000Z): Changed it.