Clean energy ETFs have done something in 2026 that most investors stopped expecting after the brutal 2022 to 2024 stretch: they have led the market. The Invesco Solar ETF (NYSEARCA:TAN) is up about 28% year to date, the Invesco WilderHill Clean Energy ETF (NYSEARCA:PBW) is up roughly 31%, and the iShares Global Clean Energy ETF (NASDAQ:ICLN) is up around 27%. Over the trailing year, the moves are larger still: PBW has roughly two-and-a-half-bagged and TAN has more than doubled.
What makes the rally worth taking seriously is the backdrop. ICLN, the bellwether, lost about 45% across 2022, 2023, and 2024. The fund came into 2026 priced for a category that policymakers had abandoned and capital had given up on. That setup, combined with three durable changes underneath the surface, is why this rally looks structurally different from prior policy-cycle rebounds.
What sets this cycle apart from the 2020 redux
Prior clean energy spikes ran on subsidy headlines and zero-rate enthusiasm. This one is being underwritten by something cruder: power demand the grid cannot meet without renewables. The International Energy Agency expects global data center electricity consumption to roughly double to around 945 TWh by 2030, with AI workloads the main driver. Utilities cannot permit and build enough gas turbines or new nuclear fast enough to fill that gap, so solar and storage are getting signed into power purchase agreements as the marginal supply.
Economics back up the demand story. BloombergNEF pegged the global benchmark levelized cost for a fixed-axis solar farm at about $39 per megawatt-hour in 2025, still the cheapest new bulk power source in most markets. That is what makes the call structurally different: solar is being bought primarily on cost economics.
The third factor is the domestic manufacturing buildout that finally took hold. First Solar alone expects to operate more than 14 gigawatts of US annual capacity in 2026 across Alabama, Louisiana, Ohio, and a new Louisiana site. Tariff walls around Chinese modules have stayed in place, and the resulting US production base is now a genuine moat.
The macro setup is also more accommodating than it was during the last leg down. The Fed has cut 75 basis points since May 2025 and held the funds rate at 3.75% since December. The cuts fall well short of zero-rate fuel, and the 10-year Treasury is still near 4.6%, yet a stable cut cycle changes the discount rate math on long-duration renewable cash flows in a way 2022 to 2024 never did. The VIX sitting near 17 tells you the rally is unfolding in a calm market environment.
TAN: the concentrated pure-play on solar
TAN is the sharpest tool on this list for one specific thesis: that solar manufacturing and US-flagged module producers capture most of the upside from AI power demand and the domestic manufacturing buildout. The fund holds around 32 names tracking the MAC Global Solar Energy Index, concentrated in module makers like First Solar, inverter manufacturers Enphase and SolarEdge, and downstream installers and developers. It runs a 0.7% expense ratio on roughly $1.9 billion in assets, which is the most institutional footprint in the pure-solar category.
The concentration cuts both ways. TAN’s five-year return is still negative at about -16% even after this year’s run, a reminder that when solar drawdowns hit, they go deep. Investors should also understand that meaningful weights in Chinese-listed names mean tariff policy and trade headlines move this fund harder than a diversified renewable basket.
ICLN: the boring core of clean energy exposure
ICLN is the fund most investors should default to if they want clean energy as a portfolio sleeve rather than a trade. It tracks the S&P Global Clean Energy Transition Index, spreads holdings across 22 countries including the US, Spain, Denmark, China, and Brazil, and charges 0.39% in fees, making it the cheapest of the three by a wide margin. Top holdings typically anchor on First Solar, Enphase, Iberdrola, Vestas, Orsted, and US utilities, which gives the fund a barbell of pure-play renewables and utility-scale developers.
That utility weighting is the entire reason ICLN has lagged TAN and PBW in 2026. Utilities dampen drawdowns and they also cap the upside when sentiment turns. The flip side is that the -44% slide from end-2021 through end-2024 happened in a fund holding regulated utilities and grid operators, which is as defensive as clean energy gets. The 2026 recovery, paired with that history, tells investors what to expect: smoother ride, slower upside, and FX risk from the global mandate.
PBW: the contrarian small-cap basket
PBW is the overlooked pick and the highest-beta way to play the theme. The WilderHill index uses an equal-weighted methodology that tilts toward smaller, earlier-stage US clean tech names across solar, wind, EVs, hydrogen, batteries, and grid hardware. It is a small fund at roughly $530 million in assets with a 0.6% expense ratio, which is part of why most allocators ignore it.
The case for PBW right now is leverage to the same structural drivers without the mega-cap weighting that capped ICLN’s 2026 move. The case against it is right there in the chart: PBW is still down about 39% over five years even after the YTD spike, and its small-cap basket has historically been the most violent name in the category in both directions. This fund deserves a slot only for investors who can sit through 30% drawdowns without flinching.
Which fund fits which investor
The decision is simpler than the category makes it look. ICLN is the right answer for investors who want clean energy as a long-term sleeve and care about expense ratio, drawdown control, and global diversification. TAN is the right answer for investors who specifically want concentrated exposure to solar manufacturing and the AI power demand thesis, and are willing to accept tariff and concentration risk in exchange. PBW is the right answer for a smaller satellite position when an investor wants the highest-beta small-cap basket and accepts the volatility that has historically come with it.
The risks worth naming before sizing any of these are the same risks that broke the category last cycle. Policy can move against renewables in a single election, the 10-year Treasury sitting near the 98th percentile of its 12-month range means rate pressure has not gone away, and the AI power demand tailwind is not evenly distributed. Hyperscalers signing large baseload contracts have leaned heavily on nuclear and gas alongside renewables, so the cleanest read of the thesis is that solar takes meaningful share of new generation, not all of it. The structural case for clean energy in 2026 is stronger than it has been in years. It comes with real conditions attached.
