The Cost-Price Squeeze | Why Steel Gets Cheaper Every Year

Introduction (00:00–02:05)

Welcome to Steelonthenet dot com Podcasts. I'm Dr Andrzej M Kotas.

Today I want to talk about a force in the steel industry that never makes the headlines, and yet shapes every long-term plan in the business.

Most of the things you buy cost more every year. Your groceries, your insurance, your car. That is simply what inflation does. Steel is different. Take a tonne of steel of a given specification, follow it across a few decades, and the underlying trend is that it gets cheaper to make. And over time, it gets cheaper to buy.

The industry has a name for this. It is called the cost-price squeeze, and it has been part of the vocabulary of steel strategic planning for forty or fifty years. And yet I still see business plans that ignore it entirely.

So let me say at the outset who this episode is for. It is for corporate planners and strategy departments in steel companies. It is for the people who write feasibility studies, and for the lenders' technical advisers who review them. And it is for investment analysts. In short, it is for anyone who has to build a fifteen- or twenty-year model of a steel plant, and decide what price to put in it.

Over the next quarter of an hour, I'll cover four things. What the squeeze is. Where it comes from. How big it is. And what it means for your plan. And I'll argue that a plan which ignores it is optimistic before it starts.

What the Squeeze Is (02:05–04:20)

Over time, steelmakers find ways of making steel more cheaply. They waste less. They use less energy. They need fewer people for every tonne. And they get more output from the same equipment. Because steel is a competitive business, those savings don't stay with the producer for long. They are passed on to customers in the form of lower prices. So, for the same product, the revenue a mill earns on each tonne drifts down over the years.

That is the whole idea. Costs fall, and prices follow them down. Falling costs squeeze prices. Hence the name.

Now, a little vocabulary. When I talk about the squeeze, I mean a long-term trend. It is not the price cycle. Steel prices can swing by tens of percent within a couple of years, as demand rises and falls and as raw material costs move. The squeeze ignores all of that. It sits underneath the cycle. If you picture the cycle as waves, the squeeze is the gentle slope of the seabed beneath them. You will never notice it from one quarter to the next. Over twenty years, however, you cannot miss it.

Not the industry as a whole. Quite the reverse. It is what keeps steel competitive against concrete, timber, aluminium and plastics. A material that became more expensive every year, relative to its rivals, would steadily lose its markets. The squeeze is, if you like, the price steel pays for remaining the material of choice.

Where it is bad news is for the individual producer that fails to keep up. I'll come back to that.

Where the Cost Reductions Come From (04:20–07:21)

Mostly technology, applied steadily over decades. Let me give you some examples, working down the plant.

Start in the melt shop. The open hearth furnace was, for a long time, the workhorse of steelmaking. It was very energy-inefficient. It was slow, with long tap-to-tap times. And its labour productivity was low. The move from open hearth to the basic oxygen furnace and the electric arc furnace cut steelmaking costs substantially, and brought better metallurgy and better environmental performance along the way.

Next, casting. For most of the twentieth century, liquid steel was poured into ingot moulds, and the ingots were later reheated and rolled. Continuous casting replaced all that, by casting slab, billet and bloom directly. Apart from big gains in productivity, the main benefit was a yield improvement of around seven and a half percent. In other words, far less steel was cut off and sent back for remelting. As I discussed in episode six, on yield losses, every tonne you waste carries with it all the labour and energy spent processing it. So a yield gain of that size is a cost saving of real consequence.

Then, consider the rolling mills. Hot charging, which saves reheating energy. Fewer breakouts and cobbles. Better process control. Each of these lowers the conversion cost of turning a slab or billet into a finished product.

Next, planning and scheduling. Computerised scheduling allows bigger batches and less set-up waste. Modern production planning and control means less inventory tied up in the plant.

Add to that what the industry calls capacity creep. Year by year, small bottlenecks are relieved, small investments are made, and operators get to know their plant better. Output from the same plant rises, by perhaps up to one percent a year. That spreads fixed cost over more tonnes, which is another quiet source of cost reduction. Episode seven, on capacity utilisation, has more on this.

And labour. A generation ago, a large integrated flat products plant might have needed seven and a half, perhaps even ten, man-hours for every tonne of steel. A modern plant of the same kind can manage with closer to two and a half. Some of that is technology. Some of it is organisation: fewer management layers, and more flexible, multi-skilled teams.

None of this is exotic. That is rather the point.

Evolution, Not Revolution (07:21–09:46)

And I think it is important for anyone forecasting this industry. Steel technology changes by evolution, not revolution.

In plain English: steelmaking rarely changes overnight. Most of the gains come from many small improvements, each one modest, adopted plant by plant. And even the genuine step changes take decades to spread across the world.

Look at the electric arc furnace over the last twenty years or so. Ultra-high-power transformers. Scrap preheating. Eccentric bottom tapping. Direct current furnaces. Not one of those, on its own, was a revolution. Each was an improvement that one plant adopted, then another, then another. Together, they have transformed the economics of the furnace.

Why does the industry move this way? There are four reasons. First, the equipment is enormously expensive. Second, it lasts a very long time; plants of forty or fifty years old are still in operation. Third, because of that, most new technology arrives as a retrofit into an existing plant, not as a clean-sheet rebuild. And fourth, the industry is understandably cautious. When a new process fails in a melt shop, it fails expensively, and sometimes dangerously.

Continuous casting is the classic example. It was a clear winner. Nobody seriously doubted its economics. And yet it still took the best part of a generation to become close to universal.

There is a consequence of all this for planners, and it is a helpful one. Because technology arrives gradually, the squeeze is steady rather than lumpy. And because it is steady, you can put it in a model.

There is also a word of caution. When you hear about a breakthrough that will transform steelmaking costs, treat it with care until it has run at commercial scale, for a sustained period. And even then, assume it will take many years to spread. In this industry, the revolution is usually announced long before it arrives.

How Big Is the Squeeze? (09:46–13:04)

Let me be straight with you. It is impossible to calculate precisely. The reason is that the product does not stay the same. Over twenty or thirty years, alloy content changes. Widths and gauges change. Coatings and finishes change. Customer specifications become tighter. A tonne of hot-rolled coil today is simply not the same product as a tonne of hot-rolled coil a generation ago. So there is no clean, like-for-like price series from which you could read the squeeze off directly.

What we use at Steelonthenet dot com is a rule of thumb. The cost-price squeeze is equivalent to a loss of roughly one percent a year from revenue per tonne. I want to be clear about what that is. It is a working planning assumption, based on long industry experience. It is not a measured statistic, and I wouldn't present it as one.

So how do we arrive at it? The most logical way is to build it up from the costs themselves. The squeeze is about operating cost per tonne falling. So take each main input, ask how fast the amount needed for each tonne has fallen, and weight that by its share of total cost.

Start with labour. As I said earlier, a large integrated plant has gone from between seven and a half and ten man-hours a tonne to about two and a half, over roughly thirty years. That is a fall of around four percent a year. Labour is typically ten to fifteen percent of cost. So labour alone contributes about half a percent a year to the squeeze.

Next, energy. Energy use per tonne has fallen steadily, by perhaps one to one and a half percent a year. Energy is around fifteen to twenty percent of cost. That adds about a quarter of a percent a year.

Then yield. Every gain in yield means less metal, less energy and less labour spent on steel that is never sold. Continuous casting was the big step, but the gains have continued since. That adds perhaps a tenth to a fifth of a percent a year.

And finally consumables: electrodes, refractories and the like, where consumption per tonne also keeps falling. Perhaps another tenth of a percent.

Add those together, and you get somewhere between nine tenths of one percent and just over one percent a year. That is where our figure comes from. One percent a year is our best estimate. The individual numbers will differ from plant to plant and from route to route, but the total keeps landing in about the same place.

Supporting Evidence from Plant Capital Costs (13:04–14:58)

There is, from the capital cost of steel plant. At Metals Consulting International we have analysed the capital cost of steel plant, in dollars per tonne of capacity, for around five hundred plants, covering twenty-seven different plant types, from electric arc furnaces to billet casters and bar mills, over the period from two thousand to twenty twenty-four.

Over those twenty-four years, the capital cost of steel plant rose by around fifty-three percent. That is about one point eight percent a year, compounded. But part of that is because the equipment itself has become better. The electric arc furnace of today, with all the features I listed a moment ago, is more equipment than the furnace of twenty years ago. Our estimate is that roughly half the increase reflects technical advancement, and the other half, about nought point nine percent a year, is price inflation proper.

Now, over the same period, general prices rose by around two and a half percent a year. So for the same equipment, steel plant has actually become cheaper in real terms, by about one and a half percent a year.

That does not measure the squeeze directly. The squeeze is about operating costs, and this is the cost of the plant. But it tells the same story. In real terms, the cost of making steel keeps falling, both the equipment and the operation. And that is what gives me confidence that one percent a year is the right order of magnitude.

Why One Percent Matters (14:58–15:47)

Because it compounds. After twenty years, revenue per tonne is about eighteen percent lower than it would be if you had simply held prices flat. On a product selling at six hundred dollars a tonne, that is a difference of about one hundred and ten dollars a tonne by year twenty. For many plants, that is the whole margin.

I should add that some practitioners use a considerably harsher squeeze, of up to two and a half percent a year. But that is for stress testing, in especially demanding circumstances, rather than as a central case.

What It Means for Planners and Strategists (15:47–18:25)

Let me give you four practical consequences.

The first is what I call the flat-price trap. A great many business plans hold the steel price constant, in real terms, for fifteen or twenty years. That looks cautious. It is not. If the squeeze is real, a flat-price plan overstates revenue from the very first year, and the error grows every year after that. I discussed in episode nine how feasibility studies go wrong. This is one of the quietest ways.

The second is that you have to run to stand still. If the market takes one percent a year off your revenue per tonne, then you need to take at least one percent a year off your costs just to hold your margin. That changes how you should look at a cost reduction programme. As I discussed in episode two, on profit improvement programmes, a plan that delivers one percent a year of savings is not a gain. It is defence.

The third follows from the second. Standing still means sliding. A plant that does not improve does not stay where it is on the cost curve. It moves steadily up it, as its competitors keep improving. Nothing dramatic happens in any single year. But after a decade, a mill that was comfortably mid-table can find itself among the marginal producers, and wonder how it got there.

The fourth is about who keeps the savings. Mostly, it is the customer. The industry as a whole passes its cost reductions on. The only producers that capture a lasting margin from technology are the ones that improve faster than the industry average. Being as good as everyone else simply keeps you in the game.

So here is the practical rule. Build the squeeze into your price forecast. Build an equivalent cost reduction path into your operating plan, and be specific about where it will come from. And then test the case where the savings don't arrive, because that is the case that tells you how exposed you really are.

Limits and Honest Caveats (18:25–19:40)

Yes, and I want to be honest about the limits.

First, raw material and energy shocks can swamp the squeeze for years at a time. When iron ore, coking coal, scrap or power prices move sharply, steel prices move with them, and a one percent trend disappears from view.

Second, in periods of high general inflation, the squeeze can be hidden inside rising prices. Steel prices can go up in money terms while the squeeze is still at work underneath.

Third, if a company moves into higher-value products, its average revenue per tonne may rise, even though the squeeze is still operating, product by product. Don't confuse a better product mix with an absence of the squeeze.

And fourth, the squeeze is a trend, not a forecast for any single year. It will not tell you what steel will cost next year. It tells you which way the floor slopes.

Conclusion (19:40–20:35)

Steel gets cheaper because steelmakers keep getting better. Not through sudden breakthroughs, but through a steady accumulation of small improvements, year after year, in plants all over the world. And the industry, being competitive, gives most of those savings to its customers.

That is good for steel. It is not necessarily good for your plant.

So if your business plan holds steel prices flat for twenty years, it is not conservative. It is optimistic. It assumes that the rest of the industry will stop improving.

They won't.

Thank you for listening. This is Dr Andrzej Kotas for Steelonthenet dot com.