字号·· | 护眼
华尔街日报

股市“便宜货”里暗藏大量AI股票The Stock Market’s Bargain Bin Is Secretly Stuffed With AI Stocks

点「原文对照」整页切到原文,或双击某段只看那段的原文。

插画中,一个人蜷缩在一个标有“FOR SALE”(待售)的纸箱里,四周漂浮着绿色的股票行情符号。Illustration of a person cowering in a cardboard box labeled "FOR SALE" while surrounded by floating green stock ticker symbols.

突然之间,“买入英伟达躺平”已不再是一句让人心安的咒语。直到几周前,网络股票聊天室里充斥着催促你买入并持有英伟达股票以致富的信息。如今,随着该公司首席执行官黄仁勋因性丑闻辞职,且这家芯片制造商的股价自10月初以来下跌逾30%,一些投资者正逃离人工智能交易,寻找避风港。

Suddenly, “NVDA and chill” is no longer a soothing mantra. Until a few weeks ago, online stock chat rooms were full of messages urging you to buy and hold Nvidia shares to get rich. Now, with the company’s chief executive, Jensen Huang, resigning over a sex scandal and the chip maker’s stock down by more than 30% since early October, some investors are fleeing the artificial-intelligence trade and looking for shelter.

一个答案可能是价值股,即市场“廉价货架”上的公司。遗憾的是,你不能只买名字里带有“价值”字样的任何基金。这种投资风格所需的耐心储备,比你想象的要深厚得多。

One answer might be value stocks, the companies in the market’s bargain bin. Unfortunately, you can’t just buy any fund with “Value” in its name. This style of investing requires a deeper reservoir of patience than you might realize.

价值投资是一辈子的事业,而非临时的交易。从人工智能仓促冲进价值基金,未必能像近期业绩所暗示的那样快速见效——尤其是因为许多价值基金依然由亚马逊、苹果和微软等这类与人工智能相关的公司主导。原因如下。

Value is a lifelong investment, not a temporary trade. Rushing out of AI into a value fund might not pay off as soon as recent results would suggest—especially since many value funds remain dominated by such AI-adjacent companies as Amazon.com, Apple and Microsoft. Here’s why.

价值股历来被定义为股价相对于每股净资产(衡量净值的一个指标)较低的股票。一些投资者也将其定义为相对于盈利、销售额或自由现金流较便宜的股票。

Value stocks have historically been defined as those with low prices relative to their book value, a measure of net worth. Some investors also define them as being cheap relative to earnings, sales or free cash flow.

用像每股净资产这样过时的指标,似乎理应导致对市场对未来最热门期望的敞口极小。但许多价值基金却塞满了与人工智能相关的股票。

It might seem obvious that using an antiquated measure like book value would result in minimal exposure to the market’s hottest hope for the future. But many value funds are chock-full of AI-related stocks.

“我们谈论‘价值’的方式,大概是华尔街在大庭广众之下犯下的最大错误,”芝加哥投资公司 Distillate Capital Partners 联合创始人托马斯·科尔说,该公司管理资产 25 亿美元。他说,鉴于专利、商标、软件和数据等无形资产的重要性日益增加,“在标普 500 指数至少有一半由无形资产驱动的世界里,主要依赖市净率是讲不通的。”

“The way we talk about ‘value’ is probably the biggest mistake Wall Street makes in broad daylight,” says Thomas Cole, co-founder of Distillate Capital Partners, an investment firm in Chicago that manages $2.5 billion. Given the growing importance of such intangible assets as patents, trademarks, software and data, he says, “relying mainly on price-to-book-value in a world where at least half the S&P 500 is driven by intangibles doesn’t make sense.”

例如,半导体制造商在工厂和生产设备上投入巨资,推高了其账面价值,往往使其有资格进入价值指数。但由于这些股票搭乘了人工智能热潮的顺风车,按其他衡量标准来看,它们可能极其昂贵。

The makers of semiconductors, for instance, have invested massively in factories and production equipment, boosting their book value and often making them eligible for value indexes. But because these stocks have been riding the AI boom, they can be extravagantly expensive on other measures.

这扭曲了罗素 1000 价值指数等流行基准。今年前 10 个月,该指数的科技板块上涨 88.5%,帮助该基准指数实现 24.5% 的回报。同期,罗素 1000 成长指数下跌 0.2%。

That has distorted popular benchmarks, such as the Russell 1000 Value Index. In the first 10 months of this year, that index’s technology sector gained 88.5%, helping to power the benchmark to a 24.5% return. Over the same period, the Russell 1000 Growth index fell 0.2%.

这是罗素公司 1987 年创建这些指数以来,任何可比时期中价值股跑赢成长股幅度最大的一次。其中很大一部分差距由米科技和闪迪等“价值”股推动,这两家数据存储设备制造商今年前六个月分别上涨 305% 和 858%。

That was the widest outperformance by value over growth in any comparable period since Russell created these indexes in 1987. Much of that gap was driven by such “value” stocks as data-storage device makers Micron Technology and Sandisk, which gained 305% and 858%, respectively, in the first six months of this year.

根据 FactSet 数据,这两家公司按账面价值衡量曾经很便宜,但到 2026 年中,按市销率计算,它们的估值至少是整体市场的三倍,按市现率计算也贵得多。

Formerly cheap as measured by book value, by mid-2026 Micron and Sandisk were at least three times as costly as the overall market based on price to sales, according to FactSet, and substantially more expensive based on price to cash flow.

自 6 月罗素基准指数再平衡以来,这两家公司连同其他科技明星股已转入成长指数——正好赶上近期人工智能股票的回调。这也让“价值”看起来表现更好。

Since the rebalancing of Russell’s benchmarks in June, those two companies, along with other technology stars, have moved into the growth index—just in time to catch the recent stumble in AI stocks. That, too, has made “value” look better.

这些设计上的怪癖使得很难找到一只包含在多个指标上都便宜的股票的价值基金。(富时罗素产品管理总监凯瑟琳·吉本说,罗素正在考虑拓宽其定义,纳入更多标准。)如今的大多数价值股,已不太像爷爷奶奶过去持有的那些股票了。

These design quirks make it tricky to find a value fund that includes stocks that are cheap on several measures. (Russell is considering broadening its definitions to include additional criteria, says Catherine Yoshimoto, director of product management at FTSE Russell.) Most of today’s value stocks aren’t much like the ones that grandma and grandpa used to own.

几十年来,市净率便宜的股票跑赢了大盘。据剑桥大学和伦敦商学院的埃尔罗伊·迪姆森、保罗·马什和迈克·斯通顿统计,从1927年到2025年,美国大盘价值股的年化平均回报率为12.3%,而大盘成长股为10.2%。这种超额回报被称为“价值溢价”。许多价值投资的拥护者——我自己几十年来也是其中之一——从未充分承认过价值溢价是多么善变且难以预测。

For decades, stocks that were book-value bargains outperformed. From 1927 through 2025, large U.S. value stocks returned an annualized average of 12.3%, compared with 10.2% for big growth stocks, according to Elroy Dimson, Paul Marsh and Mike Staunton of Cambridge University and London Business School. That excess return is called “the value premium.”What many advocates for value investing—and I have been one myself for decades—have never fully admitted is how fickle and unpredictable the value premium is.

市净率低的股票在过去二十年的大部分时间里表现挣扎。据迪姆森、马什和斯通顿统计,2008年至2025年期间,在美国,价值股年化跑输成长股3.1个百分点。

Low price-to-book-value stocks have struggled for most of the past two decades. In the U.S. from 2008 through 2025, according to Dimson, Marsh and Staunton, value underperformed growth by 3.1 percentage points annualized.

事实上,研究人员报告称,从1984年底到2024年底,价值股累计跑输成长股,跨度长达四十年。

In fact, the researchers report, value cumulatively underperformed growth stocks from the end of 1984 through year-end 2024, a stretch of four decades.

这些数字严峻地提醒我们,很难判断价值股近期的反弹将持续多久——甚至是否会持续。尽管自AI交易在10月6日左右见顶以来,价值股已跑赢成长股约15个百分点,但这一优势可能会消退。

These numbers are stark reminders of how hard it is to say how long—or even whether—value’s recent rebound will last. Even though value stocks have outperformed growth by roughly 15 percentage points since the AI trade peaked around Oct. 6, that edge could fade.

管理着1.08万亿美元资产的维度基金顾问副总裁韦斯·克里尔说:“我们在短期内没有发现价值溢价的任何可预测性。”因此,试图择时这部分市场是徒劳的。但如果你长期持有,当上涨行情到来时,你将从中获益。

“We don’t find any predictability in the value premium” in the shorter term, says Wes Crill, a vice president at Dimensional Fund Advisors, which manages $1.08 trillion. So trying to time this part of the market is pointless. But if you hold it long term, you will benefit from the upside when it arrives.

一个可能的避风港是小盘价值股——可以说是市场上最受忽视的角落,也肯定是与人工智能热潮最远的地方。研究人员发现,这一板块在长期内往往能跑赢更广泛的市场。

One possible shelter is small value stocks—arguably the most neglected corner of the market and surely the most remote from the AI boom. Researchers have found that this segment tends to outperform broader markets over the long run.

在这个领域中,年费较低的几只交易所交易基金包括:先锋晨星小盘价值基金、先锋美国小盘价值基金、先锋美国小盘价值投资组合以及iShares标普小盘600价值基金。

A few exchange-traded funds in this area with low annual expenses are Vanguard Morningstar Small-Cap Value, Avantis U.S. Small Cap Value, Dimensional U.S. Small Cap Value Portfolio and iShares S&P Small-Cap 600 Value.

请记住,便宜的小盘股总价值不足美国股市总市值的5%。但如果人工智能股票继续下跌,把你的一小部分钱投在那里,或许能让你多几分安心——前提是你要有乌龟般的耐心。

Just remember that cheap, little stocks constitute well under 5% of the total value of the U.S. stock market. But a smidgen of your money there could give you some extra peace of mind if AI stocks continue to stumble—so long as you have the patience of a tortoise.

如果你持有广泛市场指数基金,并且能长期持有不动,那么对冲短期人工智能崩盘可能得不偿失。

And if you own a broad market index fund and can stay put for the long term, hedging against a short-term AI collapse is probably more trouble than it’s worth.

阅读媒体原文 ↗