US HRC imports jumped 64% in August — but the numbers say volatility, not a trend

What the data actually say

The US Department of Commerce's preliminary census figures put August 2026 hot rolled sheet imports at 137,058 mt, up 63.8% from July and 82.8% from August 2025. Value rose to $92.1 million, from $56.6 million in July and $50.8 million a year earlier.

The headline growth rates look dramatic. They are also the least informative part of the release, because they are being measured against a very low base. Working backwards from the reported percentages (my calculation), July 2026 volumes were roughly 83,700 mt and August 2025 roughly 75,000 mt. In other words, August's "surge" took imports back to a level that is still modest by historical US standards — it did not establish a new high.

The price signal: volume, not value

The most useful cross-check is unit value. Dividing value by volume gives a remarkably flat series:

*Author's calculation. Base-period volumes are derived from the reported growth rates and are reliable to roughly two significant figures; dollar values are rounded to $0.1m, so unit values carry error bars of about ±1%.

Value growth (+62.7% m/m, +81.3% y/y) tracks volume growth almost exactly. That tells us the August increase was entirely a tonnage event: no collapse in import prices, no distressed selling, no shift in product mix visible at this level of aggregation. Anyone constructing a narrative around foreign mills dumping cheap steel into the US market will not find support in these numbers.

Korea: the swing factor, and a warning about the data

The origin breakdown is where the story gets interesting — and where the volatility becomes obvious.

*Author's calculation.

South Korea remained the largest supplier, but its trajectory is the single most important detail in the release. Korean volume was 41,310 mt in August, down from 49,096 mt in July — a 15.9% decline — yet up from just 9,420 mt in August 2025, a 4.4x increase. A supplier whose monthly volume fluctuates by a factor of five year on year, and which falls 16% in the same month total imports rise 64%, is not delivering a steady commercial flow. It is delivering shipments that land when they land.

That pattern is consistent with vessel scheduling, order bunching and the working-through of quota or tariff arrangements — but the preliminary census release does not provide licence-level or exemption-category detail, so the reason for the Korean swing cannot be established from this data. Treat any explanation as hypothesis, not fact.

Why the headline number matters less than it appears

The 82.8% year-on-year increase will circulate widely, and it is technically accurate. But scale it properly. US apparent hot rolled coil consumption runs on the order of 8–9 million short tons a month (my estimate from general industry knowledge, not from this release); 137,058 mt is roughly 151,000 short tons, or somewhere around 1.7–1.9% of monthly consumption. That is an order-of-magnitude check only, but the conclusion is robust: import volumes of this size do not themselves reprice the US market.

What they can do is shape sentiment. The same trade press page carrying this data reports US flat steel prices rising alongside higher October scrap even as flat steel imports post big gains — an apparent contradiction that resolves once you accept that 1.8% of supply is a marginal, not a dominant, pricing force. Domestic mill pricing power in the current cycle rests on scrap costs and domestic utilisation, not on a single month of import tonnage.

The origin list is the more durable signal

Set aside the August spike and look at the composition. Three Asian origins (Korea, Japan, Vietnam) supplied a combined 87,252 mt, or roughly 64% of the total. Vietnam's 20,920 mt is the entry worth watching: Vietnamese hot rolled capacity from Formosa Ha Tinh and Hoa Phat has been expanding, and Vietnamese material arriving in the US at this scale will draw scrutiny of transshipment and origin certification regardless of whether the tonnage is genuinely Vietnamese. Turkey's 6,149 mt similarly needs to be read against whatever quota or tariff regime was in force that month — which the release does not disclose.

Canada's 17,328 mt is the counterweight: a structurally stable, tariff-advantaged supply line that will not fluctuate for the same reasons Asian cargoes do.

What to watch before drawing conclusions

Three cautions apply to this release specifically:

1. It is preliminary. Census steel import figures are routinely revised, sometimes materially, in subsequent months. A single month should not be treated as settled.
2. Two months make a trend; one does not. August's 63.8% gain followed a July that was already elevated on Korean volume. If September confirms imports holding above 120,000 mt, the story changes from shipment timing to genuine import pressure. Until then, the base effect dominates.
3. The policy regime is invisible in this data. Section 232 tariffs, country-specific quota arrangements and exclusion volumes determine which origins can move and by how much. Without the licence-category breakdown from the underlying Commerce dataset, the origin mix in this release cannot be properly interpreted.

The honest reading of August 2026 is of a market where imports remain a marginal supply source oscillating around a low base, with Korea as the swing supplier and Vietnam as the new variable to monitor. The growth percentages are arithmetic; the analysis is not yet supported by enough data to call a trend.