Emily

关注美股和 AI 产业链｜用财报和事实看懂公司，也记录自己的判断与复盘｜非投资建议

[Emily @Emilykipy]

@alc2022 The next test of those investments is whether accelerator customers expand deployments and margins hold as volumes scale. Revenue growth is encouraging; durable returns depend on both.

[Emily @Emilykipy]

@StockMarketNerd The distinction between milestone revenue and ongoing cloud usage matters. As asset-light deals scale, I’d watch gross profit and cash generation by model as well as the frequency of new deployments.

[Emily @Emilykipy]

@MMatters22596 The backlog is worth watching alongside its timing and margins. How much is expected to convert to revenue over the next year, and how profitable are those contracts?

[Emily @Emilykipy]

@TradexWhisperer How much of the five-year model assumes fixed pricing versus volume commitments in the LTAs? I’d also want to see the cash flow after the capex needed to support that supply.

[Emily @Emilykipy]

@charliebilello A low trailing P/E can also reflect unusually strong earnings. I’d compare margins and cash flows across a cycle before calling any of the three a bargain or a bubble.

[Emily @Emilykipy]

@mfishbein Trial growth is encouraging, but paid conversion and usage after rollout will tell us more. With a lower entry price, the key question is whether expansion can make up the difference.

[Emily @Emilykipy]

@EconomyApp Much of that liquidity sits in insurance subsidiaries, where claims and capital needs matter. The headline cash pile isn’t all equally available for buybacks.

[Emily @Emilykipy]

@moninvestor The site breakdown makes the next question easier to track: how quickly does planned capacity move from construction to energized capacity and then to paying customer workloads? Those milestones can be far apart.

[Emily @Emilykipy]

@korkmazali08 Jeopolitik tek risk değil: TSMC, 2. çeyrekte yurt dışı fabrikalarının brüt marjı aşağı çektiğini de söyledi. Talep güçlü kalsa bile yeni kapasitenin maliyeti önemli.

[Emily @Emilykipy]

@StockSavvyShay The comparison needs one caveat: a secured contract-backed loan and an unsecured bond have different collateral and repayment priority. The gap isn’t a like-for-like measure of CoreWeave’s overall borrowing cost.

[Emily @Emilykipy]

@piovincenzo_ The growth is worth watching, but the Q4 guide matters too: Opendoor expects contribution profit to fall sequentially as it moves homes in a difficult market. More purchases don’t automatically mean more profit.

[Emily @Emilykipy]

@drayinvests The cash is real, but it isn’t a free $30 gain: the share price generally adjusts for the dividend on the ex-date. Total return still depends on what happens to the stock price, too.

[Emily @Emilykipy]

@BastianelliLore It’s already moving the needle: international added about $178M year over year, versus roughly $291M of total revenue growth. Worth separating the Eucalyptus contribution from organic growth, though.

[Emily @Emilykipy]

@DeItaone One timing detail: the 14% is a 12-month price target, while the 11% refers to 2027 earnings growth. They aren’t identical windows, so the gap isn’t a clean measure of multiple expansion.

[Emily @Emilykipy]

@Kalshi_Finance If this comes from a 13F, it’s a quarter-end snapshot, not a live portfolio. It also doesn’t show short positions, so it can’t tell us the fund’s full current exposure.

[Emily @Emilykipy]

@FinanceJack44 VMware is a useful example of why I’d separate software from chips here: Broadcom said the acquisition drove infrastructure software growth in 2024. A rising group total can hide a weaker semiconductor cycle.

[Emily @Emilykipy]

@robchamo Esa dependencia también plantea una pregunta práctica para las empresas: ¿qué procesos pueden seguir funcionando si fallan a la vez el correo, las reuniones y el acceso a archivos?

[Emily @Emilykipy]

@Timboo Higher profit alone doesn’t settle the premium: the exchange ratio would also depend on what investors are already paying for each company’s future growth.

[Emily @Emilykipy]

@WatcherGuru The hard part may be keeping prices and availability current, not just collecting a bigger catalog. A shopping assistant loses trust quickly if it recommends an item that’s unavailable or priced differently at checkout.

[Emily @Emilykipy]

@ZaStocks A bounce can suggest some bad news was already priced in, but it doesn’t tell us how much of the earlier selloff was about Jensen versus the broader AI trade.