S&P 500 PEG Ratio Hits 30-Year Low: 4 AI Power Stocks to Buy Now

While everyone's debating whether $NVDA is worth its weight in GPUs, one YouTube creator is making a compelling case that the real AI trade isn't in the chips at all. It's in the pipes, the power boxes, and the cooling systems keeping those chips from melting into expensive paperweights.

The Traveling Trader's latest video lays out the case clearly: Elon Musk has publicly identified the next major AI bottleneck for 2027, and it's not model quality or data. It's infrastructure. Think transformers, wiring, liquid cooling, chillers, and networking. With 15 gigawatts of AI compute projected for 2027 and serious power-on constraints already emerging, the companies solving the physical layer of AI are sitting on a gold mine that the market hasn't fully priced in.

Here's the kicker on valuation: the S&P 500 PEG ratio is currently at its lowest level in 30 years, below even the depths of 2008 and the COVID crash lows of 2020. The forward PE is also back at Liberation Day bottom levels. If you're waiting for a clearer "buy the market" signal, history suggests you'll be waiting until after the train has already left the station.

So where specifically is the opportunity? The video highlights four names worth watching closely:

• $VRT (Vertiv): The liquid cooling leader, directly solving the thermal management crisis at hyperscale data centers
• $ETN (Eaton): Exposure to both transformers and cooling, two of Musk's explicitly named bottlenecks
• $GEV (GE Vernova): Power generation at scale, which is the upstream problem before any cooling even becomes relevant
• $BE (Bloom Energy): On-site fuel cells offering an alternative power path that bypasses grid constraints entirely

Meanwhile, Bloomberg's crew is a bit more cautious. Their G20 coverage flags that CSIS analyst Philip Locke sees real cracks beneath the US economic surface, arguing that productivity gains may reflect a weakening labor market rather than genuine efficiency improvements. And with inflation still running at roughly 3.5% against a 2% Fed target, the macro backdrop for rate-sensitive capital spending isn't exactly pristine.

The bull and bear case basically collide right here: structurally undervalued market plus a genuine multi-year infrastructure buildout versus sticky inflation, stretched consumer sentiment, and a Fed that might be closer to a hike than a cut. The AI infrastructure trade doesn't need a perfect macro environment to work. It just needs the data centers to keep getting built. And right now, every signal says they will.

Fifteen gigawatts doesn't cool itself.