I've been following Intel for over a decade, and I still get whiplash from the sentiment swings. In late 2024, the stock sits around $35–$40, down more than 50% from its 2021 highs. P/E ratio below 25? Sure. But when you strip out the foundry losses and look at core business, it's trading at a massive discount to peers like AMD and Nvidia. The question isn't whether Intel is cheap — it's whether the cheapness is justified. After digging through quarterly transcripts, talking to supply chain contacts, and running my own DCF, I'm convinced the market is pricing Intel as if its turnaround will fail. That's a bet I'm willing to take the other side of.
The Valuation Gap: How Cheap Is Intel Really?
Let's start with the raw numbers. Intel's enterprise value to EBITDA (EV/EBITDA) currently sits around 9x, while the semiconductor industry median is closer to 18x. Even AMD, which has its own challenges, trades at 30x+.
| Metric | Intel (INTC) | AMD (AMD) | Nvidia (NVDA) |
|---|---|---|---|
| P/E (TTM) | 22.5x | 45x | 65x |
| EV/EBITDA | 9.2x | 32x | 50x |
| Price / Book | 1.6x | 3.2x | 30x |
| Dividend Yield | 1.8% | 0% | 0.04% |
The discount is obvious. But here's a nuance most analysts skip: Intel's adjusted free cash flow (excluding foundry capex) is actually positive. The company generates billions from its CCG (Client Computing) and DCAI (Data Center and AI) segments. The problem is that the foundry business is burning cash — about $2–3 billion per quarter. If you back out that drag, Intel's core ops are worth far more than the market gives credit for.
Why P/E Ratios Lie About Intel
In 2023, Intel reported a net loss of $0.4 billion — but that was entirely due to restructuring and impairment charges. Non-GAAP earnings were $1.05 per share. For 2024, consensus expects around $1.20. At $38, that's a forward P/E of 31x — still not dirt cheap, but remember those estimates are depressed by the foundry drag. Once IFS (Intel Foundry Services) breaks even (target: 2027), earnings power could jump to $2.50–$3.00 per share. Suddenly the stock looks like a 12–15x forward P/E. That's value territory.
What the Market Gets Wrong About Intel
Three main misconceptions keep Intel undervalued:
1. "Intel lost the process lead forever." Yes, TSMC is ahead. But Intel's 18A process (due late 2024 to 2025) is showing promising early yields. I've spoken with two equipment vendors who told me Intel's gate-all-around (RibbonFET) implementation is actually more aggressive than TSMC's. If 18A delivers on density and performance, Intel could regain parity.
2. "AI is all about GPUs, and Intel has nothing." Intel's Gaudi AI accelerators have gained traction with cost-conscious hyperscalers. Gaudi 3 is sampling now, and while it won't beat Nvidia's H100 on raw specs, it offers a lower total cost of ownership. I know a mid-tier cloud provider that chose Gaudi over H100 because they could deploy 30% more nodes for the same budget.
3. "Intel is a dying dinosaur." This ignores the fact that Intel still owns 70%+ of the PC CPU market and a strong position in networking and automotive. The cash flow from these legacy businesses funds the turnaround. They aren't going bankrupt.
The Foundry Gamble: Intel's Biggest Catalyst
CEO Pat Gelsinger's bet is simple: become the world's second-largest foundry behind TSMC. The U.S. CHIPS Act gives Intel $8.5 billion in direct funding plus tax credits. But the real kicker is customer wins. Intel already has Amazon, Qualcomm, and potentially Nvidia (rumored) interested in 18A capacity. If Intel secures a marquee customer like Apple or Nvidia, the stock could double overnight.
But it's a gamble. Foundry margins are razor-thin until scale kicks in. I estimate Intel needs to reach $15 billion in foundry revenue to break even — they're currently at less than $2 billion. That's a long climb. However, the U.S. government's push for onshoring gives Intel a structural advantage that TSMC can't match for geopolitical-sensitive clients.
Risks That Keep Me Up at Night
I won't sugarcoat it. Intel undervalued doesn't mean risk-free. Here are the real dangers:
Execution risk on 18A. If yields disappoint, Intel loses credibility and customers delay orders. That's a 20–30% downside scenario.
Debt load. Intel's net debt is over $25 billion, and interest coverage is declining. If foundry losses persist longer than expected, they may need to cut the dividend or issue equity.
AI shift to custom silicon. More hyperscalers are designing their own chips (AWS Trainium, Google TPU). That reduces the total addressable market for Intel's foundry.
Competitor pricing warfare. AMD and Nvidia have strong incentives to keep pressure on Intel. They could slash prices to prevent Intel from gaining foundry customers.
How to Play Intel Undervalued as an Investor
If you're convinced Intel is undervalued, here's how I'd approach it — based on my own portfolio moves:
- Start small, add on weakness. Intel is volatile. I bought my first batch at $42, but added more when it dipped to $35. Dollar-cost average.
- Consider LEAPS calls. If you have a 2-year horizon, long-dated call options can capture upside while limiting capital at risk.
- Set a catalyst checklist. I track three milestones: 1) 18A tape-out success, 2) a major foundry customer announcement, 3) quarterly foundry loss reduction below $1 billion. When two of three happen, I'll add aggressively.
And don't forget the dividend. It's not huge, but it provides a floor. Intel hasn't cut it since 1992 — they'll do everything to maintain it.
FAQ: Intel Undervalued — What Others Are Asking
*This article reflects my personal analysis and is not financial advice. I hold a long position in Intel. Fact-checked against Intel quarterly reports, earnings transcripts, and industry data from Gartner and IC Insights.
