|
Trade Drama and USMCA. Fortress
North America was a theme of the event, with AISI noting the importance of
confronting global overcapacity, while STLD highlighted transshipping
through Canada and Mexico (into the U.S.) and the need for more critical
border enforcement by our neighboring countries. 50% S-232 tariffs that
went into effect on June 4, 2025, have dissuaded imports, with 2025 steel
imports down 10-15% y/y (sheet/strip down ~30% y/y) and 2026 steel imports
down ~15% YTD (sheet/strip down >30% YTD). Steelmakers have also done
what the Trump Administration wanted, with 2026 domestic production up ~6%
YTD, 2026 utilization at ~79% (vs. 77% in 2025), and new investments. Many
speakers discussed the need for clear, predictable tariff and/or quota
guidelines, noting that melted & poured rules of origin are key to
further incentivize U.S. production and investment, not just from the
mills, but also their downstream partners. With the 2028 U.S. Presidential
Election looming, we remind investors that both S-232 and S-301 remained in
place under the Biden Administration. While steel/aluminum tariffs might
not be top of mind for the average voter, there are plenty of companies and
employees for which the metals trade landscape remains an important issue.
Given the success of the mills since 50% tariffs were implemented, we do
not envision tariff revocations under a new Democrat or Republican
Administration.
Last Wednesday, multiple media outlets
reported the U.S. and Canada had reached a tentative deal to lower tariffs
on certain Canadian steel and aluminum exports into the U.S. to 25% (from
50%). This was prior to a Friday conference call held by the White House
with a group of industry leaders. Although things seemed to be moving
toward a deal at the end of Friday’s business day, things fell apart
shortly thereafter. Canada has now announced counter-tariffs of 50% (from
25%) on U.S. steel and aluminum products. Zekelman Industries CEO Barry
Zekelman doesn’t see any way Canada wins this fight, advising Canada to
find a way to “walk away with a black eye and not get its head kicked in,”
but also recognizing the political popularity of opposing President Trump
in Canada right now. As we said earlier, clear rules of order are key to
these agreements, with Zekelman outlining his desire to invest ~$80M into a
fully automated warehouse, but noting he is not comfortable making an
investment like that until the trade relationship is resolved with surety
between the U.S. and Canada. This is just one of many industry growth
investments that hang in the balance of an uncertain trade backdrop.
Can Anyone Get Enough Steel? Given
the tariff constraints, most steel buyers have been struggling to get
sufficient levels of steel for months, with contract tons limited and spot
tons effectively unavailable in any real way. Service center inventories
are down meaningfully YTD, but shipments have increased, leading to much
tighter MOH figures. NUE echoed this sentiment, remarking that every
service center customer wants more steel than they are getting right now.
Given it was starting up new assets, STLD admitted it had a heavier contract
mix heading into 2026, and will be less aggressive pushing massive contract
tonnages in 2027. Further, buyer complaints about persistently higher
pricing in the face of frustrating mill performance (ever-extending quoted
lead times + subpar on-time delivery) were common in our discussions this
week. One analyst we heard from noted some of his clients are on allocation
and not getting tons at all, with some of his clients buying imports for
the first time ever. Service center consolidation is one solution to this
issue, with the combinations of Ryerson/Olympic and Worthington
Steel/Kloeckner making each entity a more powerful buyer from the mills. WS
noted that more acquisitions are part of its go-forward strategy. With
steel buyers under such pressure to get material and stay running, domestic
prices have increased enough that imports have ticked up in recent months,
albeit still at historically low levels. We would not be surprised to see
more imports land in the months ahead, coinciding with seasonally softer
4Q26 demand, which likely portends a pricing decline in the
September-October time frame. Seems like a pretty calm picture as we enter
next year’s contract negotiations, right?
CRU's HRC price hit $1,200/t this week for
the first time since April 2023 vs. levels in the low-$800/t range at this
time last year. While these levels are not expected to hold through 2027,
most people we spoke with this week expect 2027 average HRC pricing to
exceed $1,100/t and exhibit more seasonal monthly trends vs. 2026. As the
industry kicks off contract negotiation season, steel buyers want firm
volume commitments and better delivery performance from mills, while the
producers want higher fixed price contracts, lower CRU discounts, and
tighter min/max tonnage ranges. We see 2027 contracts exhibiting the
following features: thinner discounts for buyers vs. CRU (3-6% vs. 5-8% in
2026); rebates incentivizing buyer volume commitments; and rebates incentivizing
better mill lead times and delivery performance. With so much up in the
air, we expect this batch of contracts will take longer to resolve vs. more
normal years.
End Markets.
Outside of the two bright spots we’ve been highlighting – data centers and
the border wall – the construction
market remains challenged. In general, high interest rates continue to
stifle construction demand. With the next Fed meeting scheduled for
mid-September, the market awaits a decision to either hold or raise rates,
with next-to-no expectation of a rate cut. If and when a rate cut
eventually happens, it generally takes 12–18 months for that to show up in
non-residential demand vs. ~6 months to show up in residential demand. The
urgency of data center projects is driving solid demand across the
industry, with one speaker noting his expectation that this boom will
follow a similar pattern to the chip factory buildout a few years ago. He
expects annual data center capacity additions to likely peak in 2028
(quarterly peak in 4Q27) before falling in 2029-2031. Another speaker noted
that what was once considered a data center megaproject in 2019 is no
longer considered a data center megaproject today given companies have
dramatically increased their footprints. The average cost of a data center
build now sits at ~$1,200/sq. ft. (vs. $600-$800/sq. ft. one year ago). Automotive has seen
a portion of its steel reshored to the U.S., while there also have been
tangible examples of OEMs shifting to steel (from aluminum) given the
runaway MWTP this year. One speaker estimated this shift has represented
~400Ktpa, with more likely on the way. Industrial markets continue to expect
pent-up demand will eventually release and have seen positive manufacturing
trends in recent months. All told, we see overall real steel demand up 2-3%
y/y in 2026 and expect relatively similar trends in 2027.
|
