Broadcom is one of the largest companies in the world, its AI chip revenue is growing over 200% a year, and Morningstar thinks it's worth $650 a share while the market has been paying closer to $360. That gap — a roughly 44% discount to fair value as of late June — is the kind of thing that makes a stock interesting. Not the fact that it went up.

That's the lens for the best stocks to buy in August 2026 below. Each one comes with the actual figures: what the company earns, what it's growing at, what you pay for it, and where independent analysts think fair value sits. You can disagree with any of them, but you'll be disagreeing with numbers instead of vibes.

What the Market Looks Like Right Now

The S&P 500 gained nearly 8% in the first half of 2026 and sits around 7,410. Its trailing price-to-earnings ratio — the share price divided by the last year of profit per share — is roughly 30, against a 10-year average nearer 25. On forward estimates, which use next year's expected profits, it's closer to 21.

Strategists still see room. The average year-end 2026 target compiled by Bloomberg is 7,555, with a range from 7,000 to 8,100; Goldman Sachs lifted its own target to 8,000 and raised its 2026 earnings forecast to $340 per share, implying 24% growth. The uncomfortable part is where that growth comes from: AI-related investment is estimated to drive roughly 40% of S&P 500 earnings growth this year. One theme is carrying the index, and Bank of America has warned that speculation is "hitting extreme levels," the kind of setup that has historically preceded a valuation snapback.

So the honest framing is that these are good businesses at a moment when a lot of good news is already priced in.

The AI Infrastructure Names

These four are the same underlying bet with different risk profiles. Read that as a warning, not a shopping list.

Broadcom (AVGO) — Morningstar rates it 4 stars against a $650 fair value estimate, and it carries a Zacks Rank #2 (Buy). Fiscal 2026 earnings are projected to grow 72.1% on 66% revenue growth. CEO Hock Tan guided AI semiconductor revenue to $16 billion in Q3 FY2026, up over 200% year on year, with a path past $100 billion in AI revenue in FY2027. Of the analysts covering it, 44 rate it a buy and none rate it a sell. Earnings land September 8.

Micron (MU) — The cheapest-looking stock here and the most cyclical. Q3 fiscal 2026 revenue hit a record $41.5 billion, roughly quadruple a year earlier, on a memory shortage that sent chip prices vertical. Full-year revenue consensus is near $130 billion. Morningstar's $455 fair value estimate implies a fiscal 2026 adjusted P/E of about 7. The stock has traded near $293. A single-digit multiple is the market saying loudly that it expects these earnings to fall — memory has boomed and busted repeatedly. If you buy Micron, you're taking the other side of that.

Nvidia (NVDA) — 88% earnings growth on 82% revenue growth. Morningstar rates it 4 stars with a $240 fair value estimate against a share price near $187. Worth knowing before you buy: the stock has fallen an average of 2.8% after each of its last four earnings reports despite beating on revenue and profit every time. Beating expectations is no longer enough. Next report is August 26.

Celestica (CLS) — 70% earnings growth on 55% revenue growth. Celestica builds the servers, switches and racks that data centers physically need. Margins are thinner than a chip designer's because assembly is a lower-value job, but the orders are tied to the same spending wave.

The Cash-Flow Half of the List

Eaton (ETN) — Zacks Rank #2, with 2026 earnings growth of 10.6% on 15.9% revenue growth. Eaton makes electrical distribution and power management equipment, and every data center needs it before a single chip runs. The stock has dropped roughly 15% month-to-date, which is precisely why it's worth a look in August rather than in June. Morningstar's published view is that Eaton is well positioned for outsize electrical and aerospace demand.

GE Aerospace (GE) — Jet engines are a razor-and-blades business: the engine sells at thin margins and the decades of mandatory servicing that follow do not. Morningstar raised its fair value estimate to $347 from $307 on a steeper aftermarket margin trajectory, and currently rates the stock 3 stars — meaning roughly fairly valued, not a bargain. Analyst consensus is Strong Buy with targets clustering between $360 and $405.

Brookfield Infrastructure Partners (BIP) — The income pick, and the cheapest thing on this list on a straight valuation basis. Zacks Rank #2, a distribution yield near 4.5%, trading at 10.5x forward earnings — a discount to its own 15-year median and nearly 40% below its sector. Adjusted earnings are projected to grow 9% in 2026 and over 11% the year after. It owns toll roads, pipelines, ports and data centers, so the cash flows are contracted and often inflation-linked.

AppLovin (APP) — 56% earnings growth on 50% revenue growth at a forward P/E of 24.4. Its software decides which ads appear inside mobile apps, and margins have widened as revenue shifted from its own games toward pure software. It has also attracted multiple short-seller reports, which is a genuine risk factor rather than a footnote.

The Two Diversifiers

Neurocrine Biosciences (NBIX) — 60% earnings growth on 32% revenue growth at a 16.9 forward P/E. Its drugs are already approved and selling, which puts it in a different category from clinical-stage biotechs that either double or collapse on a single trial readout. Healthcare has lagged the broad market in 2026, which is why something growing this fast is still available at under 17 times earnings.

Alamos Gold (AGI) — 61% earnings growth, 31% revenue growth, a 10.7 forward P/E. Gold miners are a leveraged bet on the gold price: when gold rises, mining costs stay roughly flat and the extra revenue falls to the bottom line. It's on the list mainly because it doesn't move with the other nine.

What to Actually Do With a List Like This

Count the AI exposure in that list: four direct plays, and Eaton is the same bet one step removed. Buy all ten in equal slices and roughly half your money moves as a single thing on any given day. That isn't a flaw in the screen — screens surface whatever is growing fastest, and at the moment one theme is growing fastest — but ten tickers is not ten bets. If this would be a first portfolio, the usual shape is an index fund carrying most of the weight and individual names as a small slice on top. Our guide to what to invest in as a beginner gets into how to size that.

After that, the things that actually move returns are unglamorous. Position size is the big one, and it tends to get set by enthusiasm rather than by any rule — the stock you've read the most about quietly becomes the biggest holding. Whatever your number is, 3% or 5%, the name you're excited about should get the same as the one you're lukewarm on. If Nvidia ends up a good chunk of an account, one bad quarter takes back years of good picking.

Entries are the second thing. Nobody times a market trading at 30 times trailing earnings. Splitting a $3,000 buy into three $1,000 buys across three months won't feel like a strategy, but it means a rough September costs a third of what it otherwise would. That's all dollar-cost averaging is, and it earns its keep in exactly this kind of market.

The third takes about five minutes and almost nobody bothers. Write down now what would tell you the story has broken: memory prices rolling over at Micron, data center orders slowing at Eaton, a patent loss at Neurocrine, another short report on AppLovin. Then read the quarterlies against your own notes. It's much easier to decide what would change your mind before you own something than after, when there's a loss involved and every reason to wait one more quarter.

All of which assumes you'd notice the story breaking, and that assumes some vocabulary. If forward P/E or limit orders or cost basis are still fuzzy, that's worth sorting out before money goes behind a pick. Our stock market quiz runs 64 questions scored by topic — stocks, risk, index funds, fees, taxes, order types, strategy, fraud — with an explanation for every answer, which makes it a fast way to find the gaps.

The obvious caveat: none of this is aimed at you personally. Nobody writing a stock list knows your tax situation, your timeline, or how you'd sleep with a position down 30%. These are just ten companies whose numbers hold up in August 2026 based on multiple sources. What belongs in your account is a different question, and a private one.

Valuation data from Morningstar, market and strategist consensus figures via Bloomberg, rankings from Zacks, plus company filings, as of early August 2026. All of it changes frequently.