EMLP vs. ICLN: Same Power Bill, Two Different Bets

Same bell, same clock, one question: over the next twelve months, does the better energy return come from the pipes moving fossil fuel, or from the panels, turbines, and utilities chasing the switch away from them?

Two ETFs, one energy trade. First Trust's EMLP is a $4.18-billion fund of North American gas and oil pipelines, Canadian utilities, and midstream MLPs — the infrastructure that moves the barrel. iShares' ICLN is a roughly $2-billion global basket of solar, wind, nuclear, batteries, and the utilities that sell the power — the equipment that replaces the barrel. They sit on opposite sides of the climate debate. But in 2026, both are up, and both are being driven by the same invisible buyer: the AI data center, which needs a well-over-100-megawatt power bill running 24/7.

The question this card answers is not "fossil vs. clean." It's "contract vs. price." EMLP earns a fee per unit of product moved, set in advance. ICLN earns a price for electricity or equipment, set in the market.

The Card, Frozen Before the First Bell

Element	Rule
Contestants	EMLP (NYSE) vs. ICLN (NASDAQ)
Starting line	100 paper points each at the September 7, 2026 close — EMLP at $44.14, ICLN at $17.77
Finish	September 7, 2027 close, twelve months
Score	Total return, distributions reinvested
Costs	EMLP 0.95%, ICLN 0.38% — already in the price
Currency	Both USD; no FX adjustment
Opening odds	EMLP 55/45, on the five-year record and the distribution cushion
Update cadence	Monthly close, plus any corporate action that touches scoring
Exits	If either fund is delisted, merged, or halts, the last disclosed tradable close is the score. No substitutions.

No money is on the table. Paper points. The reader does not need to hold either fund to follow the card.

Why the Match Works — and Where It Doesn't

The two funds answer the same investor question — how do you earn on rising U.S. energy demand? — through genuinely different economic mechanisms. EMLP buys long-term contracts to move a barrel of oil or a unit of gas across a pipe. Revenue is set in advance. Margins are predictable. The cash comes back as a distribution. ICLN buys companies selling the future — the panel, the turbine, the battery, the grid connection — and hopes the electricity transition compounds faster than the discount rate.

The difference that matters: EMLP's income is contractual. ICLN's income is market-priced. That distinction drives everything else on the card.

Where the match is not perfectly fair: EMLP is actively managed; a fund manager picks the holdings and charges 0.95%. ICLN is passively managed, tracking an index at 0.38%. The active manager adds judgment. The index adds discipline. The reader should know which one they are buying.

The Scoreboard Before the Match

The last five calendar years tell the real story, and it is not the one the "clean energy is the future" headline implies.

EMLP, total return by year:2021 +23.2%, 2022 +10.4%, 2023 +8.0%, 2024 +33.4%, 2025 +9.7%. The fund has compounded in every one of those years. The 2024 spike came on a rotation back into midstream MLPs after the 2021-2022 rate shock. As of this writing, EMLP is up 16.5% in 2026.

ICLN, total return by year:2020 +141.8%, 2021 -24.2%, 2022 -5.4%, 2023 -20.4%, 2024 -25.7%. The 2020 spike is the pandemic-era policy bid. Then four down years, each eroding the 2020 gain. ICLN's 2026 YTD return is around +6%, after a mid-year push that briefly took it near +29% in June.

Read those two lines together and the "clean energy is the future" framing gets complicated. The future has been a four-year losing streak for the ETF that is supposed to capture it. The past — the pipe, the contract, the distribution — has been a five-year compounding machine.

That is why EMLP opens the card at 55/45. The edge is not a belief in oil. The edge is a distribution cushion that pays the holder through the ICLN drawdowns, and a track record that has not produced a down year in five.

What Each Fund Actually Owns

EMLP's top ten are 48.4% of the fund. The largest position is Enterprise Products Partners at 7.0%, which moves crude oil and natural gas liquids through the Permian Basin. Kinder Morgan is second at 3.7%, running one of the largest pipeline networks in the country. A spread of other pipeline, storage, and utility names fills out the rest. The portfolio is roughly 55% utilities and 45% industrials. The business is deliberately boring: charge a fee per unit of product moved, hold a long-term contract, pass the cash through to the shareholder.

ICLN's top ten are a different animal. First Solar leads at 8.2%, followed by China Yangtze Power (7.6%), a Chinese utility that runs, among other things, the Three Gorges dam. NextPower (6.9%) is a wind and storage operator. Bloom Energy (6.8%) makes fuel cells — the ones showing up in AI data centers this year. Enphase (4.7%) makes residential solar inverters. Vestas (4.5%) makes wind turbines. The fund is 40% utilities, 26% industrials, 24% energy. The business is not boring: selling hardware into a policy-driven market, or selling electricity into a price-discovery market.

The top 25 ICLN holdings are 71% of the fund. Three of the top five are U.S. solar or wind equipment makers, all exposed to Chinese competition and to the pace of new-build utility additions.

The Mechanism That Changed in 2026

Here is the fact that breaks the "fossil vs. clean" framing. The 2026 rally in both funds is driven by the same buyer.

A large U.S. data center is a 100-to-500-megawatt electricity consumer, running 24/7. The cheapest way to power it, in most U.S. markets today, is a gas turbine. That is the business of the midstream companies in EMLP's top holdings. EMLP's pipes deliver the gas to the turbine.

The same data center, when it cannot get gas fast enough, turns to the grid. That is where ICLN's utility holdings — NextPower, Equatorial, the Japanese and Brazilian utilities — pick up the load. The equipment side (First Solar, Enphase, Bloom) benefits when the data center operator wants a power purchase agreement, a co-location with a solar farm, or a backup fuel cell.

So the "fossil vs. clean" framing breaks. Both funds are on the same side of the trade: rising U.S. energy demand from AI. The path is different. EMLP's path is a contracted pipe fee. ICLN's path is a policy-supported electricity price. And the same macro fact — WTI crude spiking to $114 in April before settling back toward $100 — has pushed both funds higher through the entire energy complex. When oil goes up, the whole energy trade re-rates, and both funds inherit the move.

The ICLN rally was real. It was up to roughly 29% YTD in June. It has since corrected to around 6%. The EMLP rally has been more gradual and more consistent. Both are up. The question is which one keeps going.

The Two Scoreboards

The official score is total return, and that is what decides the match. But the reader who wants to know which thesis is working needs a second scoreboard — the mechanism board. Four variables:

Variable	EMLP	ICLN
Distribution yield (TTM)	~2.75%	~0.2%
Revenue source	Long-term take-or-pay contracts	Electricity prices, equipment sales, subsidies
Growth driver	Contract renewals, organic M&A	AI data center PPA demand, equipment backlog
Policy risk	Low — permits held, contracts in place	High — incentives, interconnection rules, tariffs

The distribution yield is the single most important difference for a retail investor. EMLP pays it every quarter — roughly $1.21 per share over the trailing twelve months. ICLN pays almost nothing. If the investor needs income from this position, the match is over before it starts. EMLP is the only one of the two that pays the holder to wait.

If the investor is holding for capital appreciation, the question is different. ICLN's 2026 rally was a capital-gains story, and it was the better performer through June. But the 2021-2024 record says the rally is the exception, not the norm.

The Countercase, Honestly

EMLP's case is weaker than the scoreboard suggests. The top holdings are exposed to a flat natural gas price. If gas demand stops growing — and if the AI data center builds more on nuclear and renewable than on gas — the contract renewals stop expanding. The 2024 spike of +33% came on a one-time rotation back into MLPs after a two-year rate shock. It will not repeat. And the 0.95% expense ratio is more than double ICLN's 0.38%, which is a real drag over a twelve-month horizon.

ICLN's case is weaker than the "AI power bill" narrative suggests. The fund is concentrated: the top 25 holdings are 71% of the fund. Three of the top five are U.S. solar or wind equipment makers, all exposed to Chinese competition. The policy that has supported the fund since 2020 is now in question in several markets. If the AI data center builds on gas and nuclear, ICLN's equipment makers lose to the EMLP holdings on the same power bill. And the 2021-2024 record — four down years — is a reminder that the rally can reverse.

The honest read: the 2026 rally is a macro trade, not a sector trade. It is oil up, gas up, and AI up, all at once. When one of those legs fades — and one of them will — the two funds will diverge again.

The Two Earnings Reports That Set the Next Score

The first scheduled update is the September 2026 close. Two events matter before the next update:

ONEOK's Q3 earnings. ONEOK, an EMLP holding, announced a $9 billion acquisition of Brazos Midstream's Permian assets in August, funded partly by a $9 billion minority equity investment from Apollo. The Q3 report is the first look at whether the deal is accretive to the contracted cash flow the fund depends on.

First Solar's Q3 earnings. ICLN's largest holding had a rough 2025. The Q3 report is the first look at whether the AI data center PPA demand has turned into booked revenue for the panel maker.

The reversal variable for the match is the gas-versus-electricity split in new U.S. data center construction. If gas keeps winning, EMLP extends its lead. If the grid wins, ICLN closes the spread.

The Lesson the Card Teaches

The "fossil vs. clean" framing is a 2020 debate. In 2026, the two funds are on the same side of a different question: does the investor earn the power bill through a contract or through a price? EMLP takes the contract. ICLN takes the price. The contract pays first. The price pays more, when it pays.

The 55/45 handicap stands. The investor who needs income does not need to watch the match — the distribution does the work. The investor who is holding for capital appreciation gets a real contest, and the next two earnings reports will show which leg of the AI power trade is actually loading.