Two of Wall Street's favorite artificial intelligence trades have quietly landed in the Nasdaq 100's bargain bin, even after triple-digit gains this year.
華爾街最青睞的兩個人工智慧交易標的已悄然跌入納斯達克100指數的「打折區」，儘管它們今年已錄得三位數的漲幅。
Micron Technology, Inc.(NASDAQ:MU) trades at 6.2 times forward earnings — what investors pay for each dollar of profit expected over the next 12 months — making it the second-cheapest stock in the index behindStrategy Inc.
美光科技（NASDAQ:MU）目前的遠期市盈率為6.2倍——即投資者為未來12個月每1美元預期利潤所支付的價格——使其成為該指數中僅次於Strategy Inc.的第二便宜股票。
• Micron Technology stock is showing exceptional strength. What's behind MU gains?
• 美光科技股價表現出異常強勁的勢頭。MU上漲的背後原因是什麼？
The stock is down 23.3% month-to-date in July, on pace for its worst month since June 2022. It is still up 208% in 2026, after a run of nearly 2,000% from its 2025 lows to June highs.
該股7月至今已下跌23.3%，正朝著2022年6月以來最差單月表現邁進。儘管如此，其2026年迄今仍上漲了208%，此前從2025年低點到6月高點累計上漲近2,000%。
Sandisk Corp.(NASDAQ:SNDK) has fallen 35.7% this month and 24.7% this week alone, its worst weekly drop since March 2025. It sits at 8.1 times forward earnings and remains up 504% year-to-date.
Sandisk Corp.（NASDAQ:SNDK）本月已下跌35.7%，僅本周就下跌了24.7%，創下自2025年3月以來最差單周跌幅。其遠期市盈率為8.1倍，年內迄今仍上漲504%。
Chart: Micron, SanDisk Dip After Historic YTD Rally
圖表：美光、SanDisk在年內歷史性大漲後出現回調
Are These Chips Actually Cheap Here?
這些晶片股在此處真的便宜嗎？
Memory is the most cyclical corner of semiconductors. A new fabrication plant takes two to three years and tens of billions of dollars to build, so supply always arrives late to its own party.
記憶體器是半導體行業中周期性最強的領域。新建一座晶圓廠需要兩到三年時間，並投入數百億美元資金，因此產能總是姍姍來遲。
That is why a collapsing forward multiple has historically been a warning, not a discount. It falls when the market decides the earnings underneath it are about to fall further.
正因如此，遠期市盈率的大幅下滑在歷史上通常是一個警示信號，而非折扣機會。當市場判斷其背後的盈利即將進一步下滑時，該倍數就會下降。
Micron has lived this movie.
美光科技早已經歷過這一幕。
The stock peaked near $64 in May 2018 at roughly 4.5 times forward earnings, during the most profitable year in the company's history to that point. It then fell 57% by December as DRAM prices rolled over. Revenue dropped 23% the following year.
該股票在2018年5月達到歷史最高點，約為$64，當時市盈率約為未來收益的4.5倍，這也是該公司截至當時歷史上盈利最豐厚的一年。隨後，隨著DRAM價格見頂回落，該股到12月下跌了57%。次年營收下降了23%。
The multiple was never cheap. It was pricing the peak.
該估值倍數從未便宜過。它反映的是對峰值的定價。
The bull case is that this cycle breaks the pattern. High-bandwidth memory is contracted years out at locked prices and effectively sold out, and every prior bust started in the spot market.
看漲的理由是，本輪周期打破了以往的模式。高帶寬內存（HBM）已按鎖定價格簽訂多年期合約，且實際上已售罄，而此前每一次下行周期都是從現貨市場開始的。
If AI demand is structural, these stocks deserve higher multiples, not lower.
如果人工智慧的需求具有結構性，那麼這些股票理應享有更高的估值倍數，而非更低。
Wall Street fell in love with these names on the way up. The question is whether a bargain multiple is enough to bring anyone back.
華爾街曾在這些股票上漲過程中對其青睞有加。問題在於，一個看似便宜的估值倍數是否足以吸引投資者重返。 (Benzinga)