$140 a Ton Hides a Split Basket: What Rising Recycled Commodity Prices Do to MRF Economics

$140 a ton. That's where Republic Services now pegs recycled commodity prices for the back half of 2026, up from about $136/ton in the second quarter, and the swing was big enough to lift the company's full-year guidance on 7 August (per Waste Dive's read of the Q2 call). For anyone who underwrites a materials recovery facility, that blended figure is where the trouble starts. It averages grades that spent the year moving in opposite directions, and it says almost nothing about who keeps the money.

So call it a rebound if you want. I'd call it a divergence. Old corrugated cardboard and natural HDPE have been carrying the basket while PET fell to pennies a pound, and the gap between those grades is doing more to MRF economics than the headline average ever will. The operators capturing this so-called recovery aren't the ones running the sorting floor. They're the ones who own the step right after it.

The blended price hides a basket pulling apart

Look at what the grades actually did through spring 2026. National baled prices, from Resource Recycling's May survey, don't rhyme.

Fiber and the polyolefins (HDPE and polypropylene) rebounded on tight supply and real demand for recycled resin. PET went the other way, as a wave of virgin capacity and soft bottle demand knocked the floor out from under it. PET is structurally thin to begin with; EPA's national data puts the plastics recycling rate near a tenth of what gets generated, against roughly a third for the stream overall, so it doesn't take much oversupply to push recovered PET toward worthless. A MRF's revenue is a weighted blend of these lines against whatever its inbound mix happens to be. But change the mix, or change which grade is hot, and the same facility prints a different top line with no change in how well it runs.

The sorting floor is a price-taker

The headline misses the mechanism. A MRF sells bales at spot, and it has almost no pricing power over a tonne of OCC or a bale of PET, because recovered fiber in Ohio is the same product every mill in Ohio can buy from anyone else. The one line on a MRF's income statement you can take to a lender is the processing fee it charges the hauler or the municipality. Everything downstream of that fee is merchant revenue wearing a budget line's clothing. So when people ask where rising recycling revenue actually lands, the answer is mostly not on the floor that did the sorting.

Which is why the sorting floor competes on yield and purity, not on a price it can set. A cleaner bale clears at a smaller discount to the index and gets rejected less often, so the marginal capital that earns its keep in a rising market is optical sorting, not another shift of manual pickers. A Tomra unit that lifts natural HDPE recovery a few points, or trims the outthrow that gets a fiber bale downgraded, is worth more when grades are firm than when they're on the floor. It's the ground where waste-intelligence software earns its return, and where I've watched capex cases over-credit the vision model and under-credit the mechanical pre-sort ahead of it (the network can't recover a fraction the pre-sort already dumped to residue).

In practice, the rebound reaches that floor only as a slightly better discount on a price the floor never set. Useful. Not what re-prices the business.

Who captures the rebound: the pellet, not the bale

Republic's answer is the polymer center, and structurally it's a spread trade. Buy baled HDPE and PP at the recovered-material price, wash and pelletize it, and sell recycled resin priced off virgin polyolefins. Management said its two operating centers are running more volume than modeled and capturing the higher recycled commodity prices directly, with a third under construction in Allentown, Pennsylvania, due early 2027 (from the same Q2 disclosure). The spread isn't guaranteed, since wash yield, energy, and pellet quality all eat into it, but in a firm resin market it's a far better place to sit than the bale. That's vertical integration doing what it always does. It moves the margin to whoever owns the conversion step closest to the buyer.

The demand pulling HDPE and PP up isn't a mystery. Recycled-content mandates put a floor under it. California's AB 793 requires plastic beverage bottles to carry 25% post-consumer resin in 2025, rising to 50% by 2030, and every brand chasing that target needs pellets, not bales. That policy bid lands on the processor who makes pellets. It skips the MRF that makes bales. Same tonne of plastic, two very different places to stand in the recycled plastic commodity market.

There's a premium layer on top of the mandate, too. Brands reporting recycled content on their ESG scorecards will pay up for certified, traceable pellets over anonymous bales, and that certification premium accrues to whoever runs the wash line and holds the chain-of-custody paperwork. A MRF can supply the feedstock for it. Capturing it is a different business entirely.

Geography decides how much of that margin survives, too. Recovered fiber and baled plastics are freight-sensitive, low-value-density cargo, and a MRF three hundred miles from the nearest mill or pelletizer ships a chunk of its gross away in trucking before it sees a cheque. In the export channel the same logic runs on ocean freight and FOB-versus-delivered terms. I spent a decade pricing that kind of spread, and the lesson repeats: the posted index is the easy part, and the basis to your actual buyer is where the money is won or lost.

What the split does to MRF contracts, and where it doesn't hold

This is why the contract structure matters more than the price deck. Most municipal MRF deals, the contracts that actually fund landfill diversion, either split commodity revenue with the city above a floor or run a flat processing fee with the commodity upside kept by the operator. When the basket rises, the revenue-share cities feel shortchanged. When it falls, the fee-only operators wear it. Neither structure fixes the real problem, which is that you can't debt-size against a line that can halve in a quarter. It's the same merchant-revenue trap I keep flagging in waste-to-biofuels credit economics: the number looks bankable right up until the market reminds you it isn't.

Where does this not hold? At the small end, mostly. A rural MRF running mixed paper and glass doesn't carry enough HDPE or PP in its stream to care what recycled resin does, and it can't underwrite a polymer center's capex on that volume. Those limits are structural, not a matter of this year's prices. High-contamination single-stream feedstock changes the calculus too, because the value sits in the purity and a dirty bale sells at whatever the mill will tolerate that week. And quality risk cuts deeper than price risk, which I relearned the expensive way. On an RDF export book I ran in 2024, one chloride-spec rejection at the discharge port turned a 40,000-tonne cargo into a demurrage bill, and we lost the better part of a quarter arguing over whose lab was wrong. The commodity price never moved. The margin evaporated anyway. That's the exposure a waste operator selling into zero-waste-to-landfill solutions has to price before the first bale leaves the building.

So watch the polymer centers, not the price deck. Republic will have three of them running in early 2027, and every competitor with a balance sheet is doing the same bale-to-pellet arithmetic right now. As that conversion capacity comes online, the blended commodity line on a MRF's income statement gets less informative each quarter, because more of the real margin is booked one step downstream, closer to the resin buyer. Recycled commodity prices will keep swinging. What changes is who's positioned to catch the upswing, and an independent sorting floor selling bales into a splitting basket isn't near the front of that queue. Its leverage was never the commodity. It was always the gate.

Sources & Notes

• Republic's Q2 2026 numbers, the $136/ton actual, the $140/ton full-year assumption, and the polymer-center commentary, come from Waste Dive's write-up of the 7 August earnings call.
• The grade-by-grade bale prices are pulled from Resource Recycling's May 2026 market report; I rounded them and dropped the regional spreads to keep the table readable.
• The recovery-rate context leans on EPA's national MSW figures, where the structural gap between fiber and plastics recovery is the backdrop for why PET economics stay thin.
• The RDF demurrage story is my own, from a 2024 export book; the counterparty and the port stay unnamed for the obvious reasons.