When the media tout warnings about the AI bubble, they include companies trading at low price-to-earnings multiples. When the forward P/E is even lower, then calling them bubble stocks is a misnomer.
Micron Technology (MU) fell by 22% last week and is down by 45% from the June 2026 peak. Its forward P/E of 3.3 times appears too good to be true. The firm has multi-year strategic customer agreements that are set until 2030 in some cases. Yet SCAs are not 100% future revenue. A customer would likely renegotiate or cancel the SCA if demand dried up.
At its stock price peak, Oracle (ORCL) reported RPOs (remaining performance obligations) worth $455 billion in 2025. That rose to $664 billion in Q1/2027. Still, fulfilling those bookings relies on OpenAI to demand even more compute.
Last week, OpenAI raised around $11 billion, far short of its original target of up to $30 billion. The days of easy access to credit appear over.
Markets unexpectedly re-priced the risk of the Democrats initiating scrutiny for data center approvals. Voters supported more regulations. That would give communities more say on electricity, water, and land used by proposed data centers.
Compute demand is not easing. Rental rates continued to climb. Slowing the data center build-out will only raise the cost of using AI chatbots. For now, the demand for AI continues to grow. That should help lift chip stocks like Micron, Samsung (SSNLF), and SK Hynix (SKHY).
Related Stories
