· 方法论

批评用传统模型建模CPO放量期收入,分享如何识别光子学超级周期中的纯正概念股。

涉及标的:

中文翻译

请停止尝试为2025-2026年CPO/SiPH放量期的收入建立模型了……当我上个月在约115美元(现已四舍五入到200美元)查看$TSEM时,前瞻市盈率压缩至约16-18倍(在成长情景下低至10-12倍)。同样的逻辑也适用于$AEHR/$SIVE/$SOI/Win Semi等其他标的。这是2026年上半年。量产的放量将出现在2026年下半年。我们正处在一个大规模超级周期的最早期阶段,而这些是我在光子学下一代架构变革中最纯正的概念股配置:用于测试、CW激光器、衬底和晶圆代工的公司,在光子学超级周期的新范式转换中占有一席之地。像$LITE或$AAOI这样我去年长期持有的公司已经跨越了多个周期。然而,最大的回报来自于提前预判哪些公司将在未来收入/TAM增长中获得最大收益(这部分尚未反映在当前股价中)。不是用过去2年的历史回报来计算公允价值。而且当我们看一个庞大的新型光子学TAM(lightcounting的牛市情景预测为1100亿美元以上),主要由使用CW激光器作为示例的架构驱动,或在良率方面遇到挑战。许多这类公司很可能会出现大幅重新定价。尤其是当我们看到下一代架构变革开始发生时,时间点大约在2026年下半年。

英文原文

Please stop trying to model 2025-2026 revenue for future CPO/SiPH ramp… When I looked at $TSEM back at ~$115 last month (round to $200 now). The forward p/e compressed to rates like ~16-18 (down to 10-12 in growth scenarios) Same applies to $AEHR / $SIVE / $SOI /Win Semi and other names. This is H1 2026. Volume ramp hits H2 2026. We’re at the very beginning of a massive supercycle and these are my more pure play exposure picks for the next architectural changes in photonics: For testing, cw lasers, substrates, and foundries in next paradigm shift in the photonics supercycle. Companies like $LITE or $AAOI that I’ve longed last year cover multiple cycles. However, the most returns comes from anticipating what benefits the most Mc wise relative to future revenue/TAM growth (not priced into current earnings). Not looking back at 2 year historical returns to calculate fair value. And when we’re looking at massive new photonics TAM ($110B+ bull case from lightcounting), largely driven from architecture that use CW lasers as an example or struggle with yields. A lot of these companies are likely going to re-rate hard. Especially when you look at the start of the next architectural changes, happening around H2 2026.

在 X 上查看原推 ↗