If you ask an executive at your local mill in the United States which way prices will go in 2020, they will probably tell you nowhere but up.
That would not be incorrect. Between late October and the end of last year, US steel mills rolled out four price increases totaling at least $150 per ton ($7.50 per hundredweight). And prices have since risen steadily.
Fastmarkets’ daily steel hot-rolled coil index, fob mill US, for example, averaged $491.40 per ton in October - a three-year low - when the first price increase was announced.
US HRC prices jumped by 6.9% in November to a monthly average of $525.20 per ton and continued to rise into December, in large part due to higher scrap prices that month and in anticipation of additional gains in 2020.
Domestic mills were seeking HRC prices of at least $600 per ton heading into 2020 - a level that, if achieved, would represent a 22.1% gain from the October low.
The big questions are what demand and prices will look like further into this year, and will prices have indeed bottomed out.
One big warning sign is that US steel demand growth slowed in the fourth quarter of 2019 compared with the same quarter the previous year in all major end markets, including the construction, automotive and energy sectors, UBS analyst Andreas Bokkenheuser wrote in a December research note.
“We remain cautious on the US steel sector, given the risk that steel prices [will] drop again after the current price hike/restocking process,” he wrote.
Fastmarkets research analysts, meanwhile, are less concerned about the end-user outlook. According to the steel weighted industrial production (SWIP) index they created based on the American Iron and Steel Institute's end-user breakdown of steel shipments and Oxford Economics' sector-by-sector industrial forecasts, underlying demand should recover by 1.3% after slipping 0.7% last year. Given their latest estimates that apparent steel consumption fell by a similar speed to their SWIP last year (-0.9%), they also don't believe the recent stock build will necessarily thwart prices in the short term, so long as "real" demand revives as expected.
Mind the HRC-scrap gap
With uncertainty about demand and the direction of US steel prices heading into 2020, some market participants questioned whether US HRC prices reached a floor and rebounded in the fourth quarter - a theory supported by mills' HRC lead times stretching into February - or whether that rebound, like another over the summer, might prove short-lived.
One way of testing whether the market has hit a floor is by examining the spread between the price of US HRC and domestic ferrous scrap over time. The metric is particularly useful because an increasing amount of flat-rolled steel in the US is from electric-arc furnace (EAF) steelmakers, whose primary raw material is ferrous scrap rather than the iron ore used by integrated mills.
Fastmarkets compared average monthly prices for its steel HRC index with the average for the steel scrap No1 busheling, consumer buying price, delivered mill Chicago. Fastmarkets typically lists scrap prices in gross tons but converted these into short tons in this report to make direct comparisons possible.
At a glance, the data suggests that, looking purely at input costs, US prices are not near a floor.
US HRC prices averaged $491.40 per ton in October, down by 11.7% from $556.60 per ton in September and the lowest since they averaged $482.20 per ton in October 2016.
Scrap prices trended in the same direction. They averaged $202.64 per ton in October, down by 15.3% from a September average of $239.29 per ton and the lowest since they reached $188.13 per ton in October 2016.
The result is that the price for HRC exceeded that for No1 busheling by $288.76 per ton in October, not substantially different from the $294.07-per-ton spread observed in October 2016. Those figures suggest that, in recent years, HRC has bottomed when its spread versus scrap narrowed to around $290 per ton.
There is a logic to that figure. The cost to convert scrap to HRC is around $100 per ton in a strong market when capacity utilization is high and $150 per ton in a poor market when capacity utilization is low, industry sources said. Taking the midpoint of those figures - $125 per ton - would leave a respectable metal margin of $165 per ton for flat-rolled EAF mills even during pricing troughs.
But just because it is unlikely that US mini-mills have lost money making steel over the past three years, it does not mean they have not in the more distant past.
The lessons of busts (and booms) past
US HRC hit a more-than-10-year low in December 2015 - averaging $359 per ton - in a month in which scrap prices averaged $142.86 per ton. HRC that month exceeded busheling by $216.14 per ton. In other words, mills - even EAF steelmakers, which are often perceived to be lower-cost producers than their integrated competitors - recorded significantly narrower metal margins in the wake of the 2014-15 oil price collapse than they did during more recent steel market downturns.
Oil prices are strongly correlated with steel and ferrous scrap prices. And oil prices fell to less than $50 per barrel for much of the first two months of 2015 from more than $100 per barrel in July 2014, according to data from the US Energy Information Administration. One might argue that a collapse of that speed and magnitude is rare. Perhaps, but an even steeper decline in steel prices - perhaps to below the cost of production - is not unheard of, according to a review of Fastmarkets' HRC and ferrous scrap pricing data.
US HRC prices fell in November 2001 to what has to date proven to be the lowest point of the 21st century following the dot-com bust and the September 11, 2001, attacks on the World Trade Center in New York and the Pentagon in Washington.
The domestic HRC price averaged $210 per ton that month and the busheling scrap price averaged $73.21 per ton. The result was that HRC exceeded busheling by $136.79 per ton, meaning that mills’ metal margins in late 2001 were squeezed close to, or perhaps below, conversion costs.
It is probably no coincidence that October 2001 saw former industry giant Bethlehem Steel file for Chapter 11 bankruptcy protection. But it is also worth noting that such squeezes do not only happen when demand is poor. They also occur when the market overheats.
Steel prices were near their highest point of this century in August 2008, the month before investment bank Lehman Brothers collapsed, triggering the Great Recession.
US HRC prices averaged $1,068.60 per ton that month and scrap prices averaged $765.73 ton - meaning that finished product, despite its elevated price, exceeded the raw material by $302.87 per ton. That metal margin - in the hottest steel market of this century - was not substantially above those seen in recent downcycles.
The reason for the squeeze at the time might have been the speed at which both HRC and scrap prices rose in 2008. To put that period of steep price escalation into perspective, consider that in January 2008 the HRC price averaged $625.20 per ton, meaning that prime material could have been bought, put into inventory and scrapped for profit a few months later. And indeed some savvy customers did just that.
So what should an informed steel buyer be on the lookout for in 2020?
Mind the spread between HRC and ferrous scrap. It’s much wider now than in past downturns - meaning that prices, if demand does not improve, could potentially have significant downside.