PDD's business keeps growing profitably while its cash pile builds, as solid Q2 results show. Revenue rose 8% to RMB 112.4B, led by transaction services up 13%, and operating profit rose 8% to RMB 27.8B. Operating cash flow reached RMB 25.7B, and cash and short-term investments grew to RMB 456.4B from RMB 422.3B at end-2025. Net income fell 12%, and operating expenses rose 13%, mainly on sales and marketing. PDD, along with Meta, is a stock to buy on the dip. x.com  x.com 

PDD Holdings is too cheap at 5 to 6x earnings ex cash because the market prices a permanent margin collapse, likely as extreme as the old belief in endless growth. Its push into supply chain, first-party and merchant investment may make it bigger but less profitable, and Temu's durability is unproven. Yet this quarter's $15.4B revenue still yielded about $2.8B operating profit, and $77B of cash and investments backs much of its value. If earnings just stabilize, the multiple looks unjustifiable. … Show more

PDD Holdings carries governance and geopolitical red flags that its low valuation does not offset. Its founder stepped down and left the board at 41, the company fractured a regulator's hand and kept employees from speaking to regulators, growth slowed from 90% in 2023 to about 10%, it buys back no stock, and US investors face unknown treatment in a war. @fintergirl argues it trades below 4x operating earnings with 60% of its market value backed by cash. x.com 

AI-related numbers will prove effectively made up, setting up a 'biblical' rotation out of semiconductor stocks (SOXX) by the time it is done. Confidence that the capex bubble has burst is high. x.com 

PDD can outperform without innovating because it trades at about 4x EV to free cash flow. That answers @TYCFOR, who questioned the bullish case given little innovation beyond cutting prices by going direct to producers. x.com 

PDD Holdings can no longer grow faster than high single digits and even has typos in its quarterly report. Per @teechongyen, revenue rose 8% and operating profit 7.6%, while net income fell 12% on an "other loss", and the cash flow statement labeled a US dollar column as RMB. x.com 

US small caps are set for a long run of outperformance, since their 10-year return gap versus large caps has turned negative only in technology booms, each followed by big small-cap runs. After the space-race trough they reached an 18% annualized lead by 1983; after the dotcom trough they led by 8-9% into the mid-2000s. They were cheaper than now only at the dotcom peak. Though up about 20% in 2026, they sit near the widest gap in decades. The call fails only if AI keeps all money in ten names. … Show more

Alibaba's HK$80 billion share sale looks like a grab-the-money-and-run move before the AI party ends, with base rates set to go ballistic as in 1972-1973. @AKL_019 notes the HK$112.70 price is 39.5% below the 52-week high, and asks why it didn't raise money in late 2025 or early 2026. @stevehou notes Alibaba is selling stock for AI without new highs or market validation of its AI efforts. In capital markets, the stupid but practical reason is usually the right one, as with SpaceX's fundraising. … Show more

Cybersecurity stocks look increasingly bullish on higher timeframes as capital rotates out of AI infrastructure. A new buy signal has appeared on the Startup indicators for the group; the indicators' earlier sell signal caught a roughly 20% drop and their buy signals a roughly 62% rally. The signal can still be invalidated before the weekly close, and the AI trade is not close to over. x.com  x.com 

Amazon, Alphabet, Meta, Microsoft and Oracle are likely to keep spending their free cash flow rather than deliver it to shareholders, and the market knows it. If investors believed consensus FCF estimates, like those @rbrtrmstrng shared, the megacaps would trade materially higher; instead the market expects them to spend even more on AI, or the next big thing, by 2030. They just raised capex estimates about 30-50% across the board, and either way they'll find something else to spend FCF on. … Show more