How to Bet on AI Without Single-Stock Risk — SMH, DTCR, and Three Names
How to Bet on AI Without Single-Stock Risk — SMH, DTCR, and Three Names
You want AI exposure without betting it all on one ticker
This is the question I get the most these days. "I want AI in my portfolio, but I don't want to be all-in on a single name." Two infrastructure ETFs are doing the work this year, and three individual names are riding alongside them. Here's how I think about the basket.
This is a list-format breakdown — one ticker at a time, kept tight.
1. SMH — the headline semiconductor ETF
Up about 27% YTD. Exactly what the name says — a basket of semis. The value of SMH for me is simple: when you can't pinpoint which chip company is going to be next quarter's winner, you take the whole "AI infrastructure / chips" slice in one move.
The single biggest contributor this year is Micron, up over 50% YTD on its own. The memory cycle is finally lining up with AI datacenter demand, and that combination is doing real work in the ETF.
2. DTCR — data center and digital infrastructure
Up almost 30% YTD. Global X's Data Center & Digital Infrastructure ETF. I've been talking about this one since last year, and the construction is what makes it interesting. It blends data center REITs with semiconductors — names like APLD sit next to Micron, AMD, Broadcom, TSMC, and Nvidia in a single fund.
The macro is the part most people miss. Global data center revenue is projected to expand from $416 billion to $624 billion, driven by generative AI, mobile connectivity, smart grids, and grid-based infrastructure. SMH is the chip layer. DTCR is the chip layer plus the building plus the power. Wider surface area, same direction.
3. APLD (Applied Digital)
Up close to 30% YTD. A datacenter / HPC infrastructure name, also a holding inside DTCR. Direct beneficiary of AI cloud and GPU hosting demand.
4. IREN
Up 21% YTD. Started life as a Bitcoin miner, has been pivoting weight toward AI/HPC hosting. The mining DNA gives it more volatility than SMH or DTCR, but the AI infrastructure exposure is real.
5. NBIS (Nebius)
Up 77% YTD. One of the steepest year-to-date moves I've seen in AI infrastructure. European AI cloud and GPU hosting operator. Too big to ignore, but too narrow to fit cleanly into an ETF — which is why people are running it as a single position alongside the funds.
How I'd actually combine these
If diversification is the priority, one of SMH or DTCR carries the load. If you want sharper AI exposure, the cleaner combo in my view is DTCR plus one or two individual names from APLD, IREN, and NBIS. If you don't want to touch single names at all, SMH alone still captures the chip cycle.
The risk is concentrated and obvious. One soft datacenter revenue guide can shake this whole basket. AI monetization clarity from next week's mega-cap prints will probably set the May tone for every ETF in this category.
More in this Category
If a Stop-Out Scares You, Your Size Is Too Big — Thinking in 10,000 Trades
If a Stop-Out Scares You, Your Size Is Too Big — Thinking in 10,000 Trades
If a stop-out scares you, the position is too big. How thinking in terms of the next 10,000 trades — not the next one — changes position sizing and trading psychology.
Why Top Traders Average Only 2-5% Per Month — And Still Crush the S&P 500
Why Top Traders Average Only 2-5% Per Month — And Still Crush the S&P 500
Verified traders average 2-5% monthly returns, not the "$5,000 morning" numbers you see online. Compounded, that's roughly 27-80% annually — enough to beat the S&P 500's 10% by roughly 13x over a decade.
Build the Risk Plan Before the Entry Rules — How to Simulate a 7-Trade Losing Streak
Build the Risk Plan Before the Entry Rules — How to Simulate a 7-Trade Losing Streak
The most overlooked work in trading is not finding the perfect entry — it's designing what happens during the inevitable drawdown. A four-part risk plan covering per-trade loss caps, weekly stops, streak probability math, and max drawdown triggers.
Previous Posts
AI Memory Showdown: Why Micron Crushed 5 Rivals in a 6-Round Face-Off
AI Memory Showdown: Why Micron Crushed 5 Rivals in a 6-Round Face-Off
When I scored MU, AVGO, MRVL, WDC, STX, and NTAP across six identical metrics, Micron won 4 of 6 rounds with the best margin (41.5%), revenue growth (194.1%), and balance sheet (14.9% debt/equity) in the group.
Hidden Leverage in AI Memory Stocks: Seagate at 1,046% and NetApp at 236% Debt-to-Equity
Hidden Leverage in AI Memory Stocks: Seagate at 1,046% and NetApp at 236% Debt-to-Equity
Among 6 AI infrastructure stocks I scored, Seagate (1,046.6%) and NetApp (236.1%) carry debt-to-equity ratios that vastly exceed the 50% threshold for non-financial companies. The leverage hidden behind their headline growth stories deserves closer scrutiny.
Six AI Infrastructure Roles, One Data Flow: How MU, AVGO, MRVL, WDC, STX, NTAP Differ
Six AI Infrastructure Roles, One Data Flow: How MU, AVGO, MRVL, WDC, STX, NTAP Differ
In an AI datacenter, Micron is the fuel, Broadcom the highways, Marvell the bridges, Western Digital the engine, Seagate the warehouse, and NetApp the traffic control. Six stocks, six distinct positions in the same data stack.