March has handed AI investors something they rarely get: a pullback! Market have been sliding this week as investors fret over rising oil prices and concerns tensions in Iran could spill over into the global economy.
That’s led to some stocks with strong secular tailwinds selling off this week. We recently announced 5 new buy recommendations on the AI Investor Podcast by 24/7 Wall St.
Here’s a listing of our most recent buy recommendations, below I’ve provided more details on each stock and why we’re choosing to pul the trigger on buying them.
| Company | Ticker | 1-Year Return |
|---|---|---|
| Onto Innovation (NYSE:ONTO) | ONTO | +35.14% |
| Liberty Energy (NYSE:LBRT) | LBRT | +84.40% |
| Intel (NASDAQ:INTC) | INTC | +113.86% |
| AIXTRON (OTC: AIXXF) | AIXXF | +125.34% |
| NVIDIA (NASDAQ:NVDA | NVDA Price Prediction) | NVDA | +62.91% |
If you’re not familiar with the AI Investor Podcast, it’s hosted by 24/7 Wall St. Analyst Eric Bleeker.
Eric is famous for first calling NVIDIA’s AI opportunity back in 2009 and has been closely following the AI space ever since. He manages a portfolio of his favorite ideas (which was just expanded to $1 million of his personal capital on the last episode!). The best part: getting these recommendations is absolutely free.
So far, the average return of his recommendations is at 80%, significantly outpacing the market. Some of the biggest winners have included:
- Coherent: Up 204%
- Credo: Up 293%
- Vertiv: Up 111%
- Lumentum: Up 562%
- Ciena: Up 323%
- SK Hynix: Up 330%
- Lam Research: Up 165%
- Applied Optoelectronics: Up 335%
And all these recommendations have happened across the past 18 months. So, let’s dive into Eric’s most recent recommendations that he’s calling the ‘top 5 AI stocks to buy in March.’ If you’d like to listen to the full episode with Eric talking about all five of these companies, we’ve embedded it in Spotify and Apple Podcasts below.
#5: Liberty Energy
- Ticker: Liberty Energy (NYSE:LBRT)
- Market Cap: $4.51 billion
- Business: Oilfield services company pivoting toward distributed power infrastructure for AI data centers.
- Buy Amount: $5,000
Liberty is the most indirect AI play on this list. Its Liberty Power Innovations subsidiary has secured a 1 GW development agreement with Vantage Data Centers, with 400 MW firm for 2027, plus a separate 330 MW Texas data center agreement. The company is targeting 3 GW of distributed power deployment by 2029. CEO Ron Gusek framed it directly: “Liberty has evolved from a premier North American completions company into a diversified energy technology and power infrastructure platform… powerful growth engine with AI and cloud data center power demand.”
The key idea with Liberty: massive data center power needs are requiring companies to go off the grid. That is to say, companies like Google, Microsoft, and Amazon are building ‘behind-the-meter’ power options that rely on natural gas turbines and other energy sources. Liberty has long had expertise in providing off-grid energy to the fracking industry, and now is able to pivot that expertise to the data center industry.
The upside could be very large. Liberty has already begun winning massive contracts and is targeting 3 gigawatts of deployment by 2029. If it hits those targets, revenues could double. Liberty isn’t the most well-known name in the AI space, but it has major tailwinds behind it.
#4: Intel
- Ticker: Intel Corporation (NASDAQ:INTC)
- Market Cap: $229.5 billion
- Business: Semiconductor giant executing a foundry turnaround while rebuilding its AI data center business.
- Buy Amount: $10,000
Intel’s +113.86% one-year return is the second-best on this list, reflecting a genuine narrative shift.
The company introduced its first products on the Intel 18A process node, with CEO Lip-Bu Tan calling it “an important milestone… sharpen execution, reinvigorate engineering excellence, and fully capitalize on the vast opportunity AI presents.” The Data Center and AI segment delivered $4.74 billion in Q4 revenue, up 9% year over year.
We’re investing in Intel for two primary reasons:
- First, the company has a terrific opportunity with demand for CPUs exploding. Agentic AI usage is driving AI demand at torrid levels, and we believe this catalyst will persist across 2026 and 2027. Intel failed to meet demand last quarter due to a lack of inventory, which caused shares to step back. However, we’re confident the boom in CPU demand will last long enough for Intel to significantly cash in. This provides Intel with runway to continue building out its foundry business.
- And second, what we just referenced above. Intel’s future rests on becoming a viable foundry alternative to Taiwan Semiconductor. Our belief is that Apple will move some of its M-class chip to Intel, NVIDIA will likely use the company for a portion of its advanced packaging needs, and Intel will also find new customers in American companies like Microsoft and Amazon. Building chips is a difficult business that requires a ‘flywheel.’ That is to say, foundries need to get initial customers that then allow them to mass produce chips and improve the ‘yield’ and economics of production. To date, Intel has been unable to sign up initial customers, but we’re optimistic that Intel will begin signing major deals across the next 12 months.
#3: Onto Innovation
- Ticker: Onto Innovation (NYSE:ONTO)
- Market Cap: $9.79 billion
- Business: Semiconductor process control and inspection equipment maker critical to advanced chip manufacturing.
- Buy Amount: $10,000
Onto’s Dragonfly platform is becoming the inspection tool of choice for high-bandwidth memory manufacturing, the memory architecture powering AI accelerators. The company signed a $240 million-plus volume purchase agreement with a leading HBM manufacturer for Dragonfly through 2027. Q4 2025 was a record quarter with $266.87 million in revenue, and Q1 2026 guidance of $275 million to $285 million suggests momentum continues. CEO Mike Plisinski called it a “robust upcycle in semiconductor capital equipment spending.”
The bottom line is that ‘advanced packaging’ isn’t a topic most investors know about, but it’s quickly becoming one of the most attractive investing categories in AI. Onto Innovation is about to start seeing major growth from memory makers, and could soon benefit from other major trends like ‘co-packaged optics.’ Shares have dropped amidst market volatility in recent weeks, providing what we’re confident will be a strong entry point for investors who want to hold a company that looks poised to become a ‘secular winner’ in the AI buildout.
#2: AIXTRON
- Ticker: AIXXF
- Market Cap: $3.8 billion
- Business: A semiconductor equipment company that could see revenue skyrocket thanks to two incoming megatrends.
- Buy Amount: $10,000
Here’s a name that’s likely not familiar to many investors! AIXTRON is a German company that specializes in creating semiconductor equipment for compounds like Gallium Nitride (GaN) and Silicon Carbide (SiC).
Put simply, those two markets have stunk in recent years as AITXON’s largest customers were in the EV space. However, the next major datacenter shift from NVIDIA is 800V racks that will shift AIXTRON’s end market from EVs and into data centers in a massive way.
AIXTRON’s also a leading supplier to the optics industry. Demand for optics equipment is rabid, just this past week NVIDIA invested $4 billion into Lumentum and Coherent to assure supply. As optics companies race to build out more capacity, they’ll look to AIXTON to provide their vapor deposition equipment. I believe AIXTRON is a (largely) unknown company whose revenue could go from declining today to skyrocketing in the years ahead.
#1: NVIDIA
- Ticker: NVIDIA Corporation (NASDAQ:NVDA)
- Market Cap: $4.46 trillion
- Business: The dominant designer of AI accelerators and the infrastructure software stack powering the global AI buildout.
- Buy Amount: $25,000
NVIDIA just reported Q4 FY2026 revenue of $68.13 billion, up 73.2% year over year, with Data Center revenue of $62.31 billion, up 75%. Full-year revenue came in at $215.94 billion. Q1 FY2027 guidance is approximately $78 billion. Full-year free cash flow was $96.58 billion. Jensen Huang said on the earnings call: “Agentic AI inflection point has arrived. Grace Blackwell with NVLink is the king of inference today.”
Here’s the bottom line: NVIDIA shares are surprisingly cheap. I predicted weeks ago that the company could see $9 to $10 in EPS this fiscal year. At the time, Wall Street expected NVIDIA to make $7.76.
As of today, Wall Street estimates have now moved up to $8.25. Slowly but surely, they’re coming around to what I predicted weeks ago.
Why does this matter? Well, if NVIDIA delivers something like $9.50 in EPS this year, that will mean:
- The company (despite its size) will be growing earnings at nearly 100% pace
- And yet, shares trade at just 18.7X my estimate for this year’s profits. That’s below market averages.
Company after company in the AI space this week spoke of increasing visibility of strong AI orders not just in 2026, but into 2027 and even 2028. Wall Street is fretting that NVIDIA’s growth will come crashing back down to Earth next year. My bet: they won’t grow 100% again, but growth will be strong enough that the price NVIDIA shares trade at will look absolutely foolish.
We first recommended NVIDIA in September 2024, and that recommendation is up 50%. I like the chance for this recommendation to grow 50% across the next 18 months as well. NVIDIA offers great risk-reward at today’s prices.
The Bottom Line
If you’re looking for new buy ideas, 24/7 Wall St. is your number one place to find them. On the next episode of the AI Investor Podcast, I’m going to be releasing even more buy ideas. All you have to do is subscribe to make sure you don’t miss an episode. You can also join our email list to ensure you get updates whenever we issue new recommendations.