DBS CIO: Nvidia's 17x P/E Shows AI Rally Is Not a Bubble
nashnova research
DBS Group CIO Hou Wey Fook argues that Nvidia's forward P/E of just 17x — paired with an expected 70% earnings growth next year — is nowhere near the 100x Cisco carried before the dot-com bust, meaning the AI tech rally is far from bubble territory.
Why does a 17x multiple rule out a bubble?
Nvidia's 12-month forward P/E — the valuation multiple based on expected earnings over the next year — sits at roughly 17x, with next-year earnings projected to grow 70%.
Hou compared this to Cisco's roughly 100x multiple before the dot-com crash: "If the flagship AI stock trades at a teens multiple, how can you call it a bubble?"
This means → by the valuation yardstick, the market is paying for Nvidia's AI growth at a price far below historical bubble levels, and tailwinds in semis and AI remain intact.
How should investors ride AI gains while managing risk?
Hou recommends a "barbell" allocation: growth-oriented tech on one end, investment-grade fixed income on the other.
Hedge funds and gold sit in the middle as diversifiers, keeping overall portfolio volatility in check while preserving AI exposure.
In plain terms = don't go all-in on tech, and don't hide in bonds either — load both ends of the barbell and add a shock absorber in the middle, so you keep pace on the way up without capsizing on the way down.
What has been happening with Nvidia's stock and capital flows?
Nvidia's share price hit an intraday record last Friday. Market cap stands at roughly $5.6 trillion, just short of becoming the first public company to breach $6 trillion — up about 25% year to date.
On the funding side, Nvidia and SoftBank each completed their final $10 billion tranche into OpenAI's latest round, fulfilling $30 billion commitments apiece.
A weaker-than-expected US September payrolls report led markets to pare bets on a Fed rate hike in October, lifting tech broadly and giving Nvidia an added push.
How long can the "no bubble" narrative hold?
Whether a 17x forward P/E can keep supporting the "no bubble" story ultimately depends on Nvidia delivering its expected 70% earnings growth next year.
This means → the multiple itself is not expensive, but it is built on a high-growth assumption — if growth misses, 17x could start to look pricey.
In plain terms = the current price already has next year's rapid growth "baked in." The story survives only if earnings keep up with the bill.
市场有风险，内容仅供研究参考，不构成投资建议。