We Tested Whether Waiting Pays: 253 Days of Scrap Prices

We tested every trading day of the past year against our own price series. Waiting an extra month before selling copper would have left you better off 71 percent of the time, with a median gain of 4.2 percent. That sounds like a case for holding, but it is really a description of a rising market, and sorting a load still adds 15 to 25 percent in any market at all.

Scrap Price Today market analysis, written from exchange settlement data, mill and refiner pricing practice and published policy documents. Figures are examples; today’s numbers are on the live scrap price board.

Most timing advice is assertion. This is a measurement, with its limits stated, and the conclusion is not quite what either side of the argument expects.

What we measured

For each of the last 253 trading sessions we asked a simple question: if you sold on that day, would waiting 21 more sessions, roughly a calendar month, have left you better or worse off? We ran it across four derived series.

On its face, patience won. Copper and brass rewarded waiting on roughly seven days out of ten, and the typical reward was a few percent.

Why that number is not advice

Here is the honest caveat, and it is the whole reason to publish the method alongside the result.

Over the window we measured, copper rose 52.6 percent. In a market that spends a year climbing, any test of waiting will find that waiting paid, because the trend is the answer rather than the timing skill. Run the same test across a falling year and it will invert just as cleanly.

What the numbers genuinely tell you is the shape of the risk. Aluminium had both the best single outcome and by far the worst, at minus 15.7 percent, because it peaked in late May and fell hard afterwards. Ferrous barely moved either way. Those distributions are stable properties of each market; the direction is not.

The comparison that actually matters

Sorting a mixed load into its grades typically returns 15 to 25 percent more, because each grade is paid at its own price instead of the lowest one present.

Set that against the median month of waiting on copper, at 4.2 percent, and the ranking is clear. Sorting is roughly four times the size of the typical timing gain, it works whether the market is rising or falling, and it takes about half an hour with a magnet and cutters.

Selling in tier quantities comes second. Most yards pay better above a few hundred pounds and better again above a thousand, and that gap is also larger than a typical month of price movement. The best scrap price guide covers both.

Ferrous is a different question entirely

Our ferrous series barely moved on this test: waiting paid on 53 percent of days with a median of 0.3 percent, and the whole range ran from minus 2.4 to plus 9.2 percent.

That is because ferrous is not traded, it is settled. Mills publish buying programs and hold them, usually resetting monthly, and the benchmark index behind our board is itself published monthly. Watching a ferrous board daily is wasted effort; knowing your mill’s settlement week is not, as our steel analysis sets out.

When holding is defensible

• You are close to a volume tier and one more collection run would clear it.
• A ferrous settlement is days away and programs have been moving up.
• You are outside the United States and your currency is weakening fast, which lifts the local value of a dollar-priced benchmark without the metal doing anything.
• Material is stored dry and secure, so holding costs you nothing but time.

When it is not

• Waiting for a round number you read in a forecast. Copper spent the year making levels that each looked like a peak at the time.
• Holding aluminium on a twelve-month chart when the three-month direction is down, which is exactly the situation now.
• Holding a converter for a rhodium move, which has swung thousands of dollars an ounce in both directions with no warning.
• Holding wet or outdoor-stored material, where moisture deductions and rust erode the grade faster than the price improves.

A routine that works in either market

• Sort as you collect rather than in one session before selling.
• Check the live board and the charts on the morning of the sale, not the week before.
• Use the calculator to convert your weights into an expected figure so a poor quote is obvious.
• Phone two yards with the same description of grades and weights.
• Outside the United States, check the exchange rate at the same time; it moves local boards on days when the metal does nothing.
• Sell when the number is fair for the grade, then start collecting again.

Key numbers

• Waiting a further month paid on 71 percent of days for copper and brass, 68 percent for aluminium and 53 percent for ferrous, over the last 253 sessions (Scrap Price Today series analysis).
• Median gain from waiting a month: copper 4.2 percent, brass 3.8 percent, aluminium 3.1 percent, ferrous 0.3 percent.
• Worst outcome from waiting: aluminium at minus 15.7 percent, reflecting its fall from a late-May peak.
• Copper rose 52.6 percent over the measured window, which is why the test favours waiting and why it should not be read as a rule.
• Sorting a mixed load typically returns 15 to 25 percent more than selling it as one grade (Recycled Materials Association grade definitions).
• Ferrous benchmarks are published monthly, so the market steps rather than trades (FRED producer price data).

What this metal pays around the world

Scrap is a global market priced in US dollars, but what reaches a seller depends on trade position, freight to the nearest consumer and how many buyers compete locally. The table below converts the same grade into each market’s own currency and units at the day’s exchange rate, and shows the factor we apply.

How we produced this, and what it cannot tell you

The test compares each session’s close with the close 21 sessions later across our derived daily series, counting outcomes rather than modelling them. It covers one twelve-month window, which was a strong one for copper and a weak one for aluminium after May.

It is backward looking by construction. It says nothing about what the next month will do, and anyone presenting a result like this as a strategy is misreading it. We publish it because the distribution is useful even when the direction is not: it shows how much a month of patience is typically worth against how much sorting is worth, and that comparison holds regardless of trend.

The full derivation of every series is on our editorial policy page.