Summary and Outlook
A puzzling question facing all of us in the steel industry is the reality of 30%-50% steel price increases in the past three months despite a reported industry operating rate last week of no more than 75%. To reconcile the seeming inconsistency we first need to conceptualize the notion of “available” capacity and “nominal” capacity. We believe that available capacity is most probably 20%-30% less than nominal capacity, so current operating rates vastly understate the real market dynamic.
Our view is that the price of steel is no longer the driver of the steel mill restart decision it once was. Many other variables are in play – availability and unpredictability of the costs of raw materials are other keys. Still another driver is duration of the cycle - we’ve seen unprecedented short cycles in the last 21 months - three mini-cycles lasting three to nine months each. So now the real question is not just “what price” but also “how long.”
Exacerbating today’s tightness are “peaking” problems - with “available” capacity now running flat-out for months, we’re seeing typical supply disruptions beginning to bubble up in the sector. We believe that HRC prices even as high as $900/ton may be insufficient to justify restarts – and the “clearing price per ton” to restart a blast furnace for only three months may be near infinity. Hence the true operating rate – as a percent of available capacity – may more realistically be closer to 90%.
Raw Material Cost Increases Drive Price Increases
While steel players are citing as rationale for their pricing moves the unrelenting march higher of raw material costs, in our experience, raw material cost pass-through only works where there is either a contractual arrangement or a relatively full order book. The mini-mill model differs mainly to the extent that the mini-mill is dominated by variable costs and has a fluid production model, so given vastly greater degrees of freedom, the mini-mill is far more likely to cut the operating rate the moment the spreads narrow to the point of negative returns. The business of making steel has become less and less fixed cost over time, so that unless pricing is of a sufficiently higher magnitude than costs – meaning margin exists – productive capacity won’t be deployed.
Huge Costs to Restart Idle Capacity
Powering on and off blast furnaces has always been an expensive process. As raw materials become a greater proportion of total overall costs, this variable piece has driven far earlier production cuts than we’d ever seen before – we call this concept “turning gold into straw.” But now there is a new variable – the durability of the business cycle. With steel pricing cycles now running at six months – on average – over the past 21 months, the decision to restart idle capacity is as much about how long the pricing cycle lasts as it is about what today’s price is. Restarting idle capacity involves huge fixed “lump” costs, such as bringing back laid-off workers (and restarting the clock on their layoff benefits) arranging volatile-priced raw material purchases, etc. The clearing price for the restart of idle blast furnace capacity for only a three-month period may be near infinity.
Domestic Operating Rate Lags and Is Understated
Weekly steel production so far in February is being reported at a 75.3% operating rate, up from a low of 66.8% in November 2010. There is a nuance in this data most observers are missing - half of the companies reporting production to the American Iron & Steel Institute (AISI) actually only report monthly data – and AISI uses the most recent full month as a “plug” on this weekly reporting. If we assume that companies not reporting “real time” are increasing production at the same pace as the ones that are reporting weekly, then the implied operating rate of the industry is closer to 77%.
Available Capacity Much Less Than Nominal Capacity
If we adjust for U.S. capacity that we know is currently idled - which includes ArcelorMittal USA’s (MT) three smallest blast furnaces as well as Severstal North America’s Sparrows Point, MD, and Steubenville, OH, steelmaking operations – as well as the monthly/weekly issue – then the domestic operating rate probably is closer to 83%. We would also point to the typical “peaking” outages we’ve seen in the past month at various mills with supply disruptions. We would argue here that these “disruptions” can be almost predicted as a normalized “peaking” phenomenon – we couldn’t possibly predict which integrated mill will have disruptions, but saying that a 5%-10% disruption “rate” is normal at full capacity is about one of the truest things we know.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.