Supply Disruptions Raise Our Near-Term Iron Ore Price Forecast

Near-term tightness in the iron ore market has persisted and intensified, with several developments in Brazil further restricting Vale’s VALE supply and Cyclone Veronica off Australia interrupting Pilbara shipments. We’ve factored in a reduction of another 20 million tonnes in Vale’s output in 2019 and 10 million tonnes in 2020. We now expect Vale to produce 350 million tonnes in 2019 and 370 million tonnes in 2020, down from an estimated 390 million tonnes in 2018. For Rio Tinto RIO, BHP BHP, and Fortescue, we’ve lowered our forecasts by 10 million tonnes in total for 2019 due to the cyclone. The estimated 30 million tonnes of lost supply from Vale and the Pilbara in 2019 is a more than 1% reduction to the seaborne iron ore market.

Disruptions mean that higher-cost iron ore is needed to balance the market, such as from domestic mines in China. The iron ore price has averaged $83 per tonne year to date, well ahead of our prior $65 per tonne forecast for 2019. Accordingly, we are raising our near-term iron ore forecasts to $73 in 2019, $60 in 2020, and $50 per tonne in 2021. Our prior forecasts were $65 in 2019, $55 in 2020, and $40 per tonne in 2021. Our unchanged $40 per tonne long-term forecast now starts a year later, in 2022.

All major iron ore miners we cover benefit from the higher price forecasts, including Vale. However, for Vale, there’s uncertainty around the cost to rectify the Feijao dam failure and compensate the victims as well as legal action that may affect the operation of other mines. Fortescue benefits most because it’s an iron ore pure play and has lower margins than BHP or Rio Tinto, which brings greater leverage to the price.

We’ve not changed our $40 per tonne long-term forecast, given the relative flatness of the iron ore cost curve inside the steep tail of smaller-scale and marginal producers, most which we eventually expect to exit. Disruptions to Vale’s supply should resolve within the next few years. In terms of iron ore supply additions, the lost output from Vale, including Samarco, should come back in the medium term. The S11D project should also expand to reach capacity over the next few years. BHP and Rio Tinto should grow modestly as those companies reach their installed capacities. Anglo American’s NGLOY Minas Rio mine in Brazil should add more than 20 million tonnes per year after being shut to rectify slurry pipeline leaks. Most of the additional output from Anglo will come in 2019. From a disrupted 2019 base of about 350 million tonnes, we expect Vale’s output to grow to around 425 million tonnes a year from 2023.

Beyond a normalization of existing supply, the key driver of our call for long-term prices well below current rates and consensus is the outlook for demand from China. Recent demand for steel and iron ore in China has been stronger than we expected. We underestimated China’s willingness to prop up steel consumption through poor-returning projects funded by ever-increasing debt. Continued growth in residential construction has also been a surprising contributor but seems unlikely to continue, given that China’s working-age population is shrinking, as are the requirements for new housing.

Stronger-than-expected steel demand from China in the past few years does not represent a new higher base from which future demand will grow. Rather, it just brings the economy’s total stock of steel closer to the Western world average of around 12 tonnes per person. At this point, the stock of steel tends to stabilize, with demand largely satisfied by scrap generated from recycling. As steel-containing things such as cars, buildings, and infrastructure reach the end of their useful lives, that scrap steel is recycled, lessening the requirement for production of virgin steel using iron ore and coking coal. We expect China’s demand for steel to shrink from elevated 2018 levels by about 3% a year to 2025. The growing proportion of demand being met through scrap rather than virgin steel means that we expect virgin steel production to fall at a slightly faster rate. If China continues to add virgin steel to its stock of steel at the rate it did in 2018, it will have reached Western world levels within a decade. For this reason, it’s logical to expect the production of virgin steel to contract.

This outlook for recovering supply and lower demand is why we expect iron ore prices to fall meaningfully in the future. It’s also why we still see the companies exposed to iron ore mining as more than 30% overvalued on average.

We see some risk that additional longer-term supply may come into the market. Would-be entrants may be encouraged by prevailing high prices. Mineral Resources is an obvious candidate, given its strong financial position. The company wants to add 30 million-50 million tonnes of new production and has the financial firepower to push ahead. We forecast Mineral Resources to have nearly AUD 500 million net cash by June, supported by the sale of 50% of Wodgina to Albemarle for $1.15 billion. There’s also a chance the large, high-grade greenfield Simandou deposit in Africa could be developed with funding from China. Privately owned Hancock Prospecting is reportedly looking at a small 5 million-tonne expansion of Roy Hill. The company’s longer-term designs are unknown, but the financial position should be robust and rapidly strengthening with current prices. Growth in supply from any of these sources would expand the relatively flat part of the cost curve at the expense of the steeply rising tail of high-cost producers.

The additional forecast production losses for Vale of 20 million tonnes in 2019 and 10 million tonnes in 2020 account for the potential for mines that are not directly affected by Brazil’s move to ban upstream tailings dams to nonetheless be indirectly affected by legal action and other regulatory restrictions. Dams are being inspected and tested for stability, so operations may be interrupted. We’ve already seen this with just over 50 million tonnes of capacity temporarily halted. This is beyond the original 40 million tonnes per year of disruption that Vale anticipated from operations directly affected by the upstream ban.

Vale mines temporarily disrupted include Brucutu, Timbopeba, and Alegria. Brucutu is the largest of the three, producing about 30 million tonnes a year. Brucutu was forced to close after its provisional operational authorization was suspended in early February. The mine was set to restart this week, but legal action has again stopped Vale from discharging Brucutu tailings into the Sul dam. We estimate nearly two months of output, or about 5 million tonnes of iron ore supply, has been lost so far from Brucutu. The Timbopeba mine produces around 13 million tonnes a year and was halted in mid-March due to concerns about the Doutor tailings dam. However, it seems likely the delay should be short, given the dam was inspected by the national mining agency in Brazil, which verified the dam does not have any condition that compromises its safety. The outage at the 10 million-tonne Alegria mine could be more protracted. Initial analysis of the structures could not guarantee their stability under stressed conditions. Further analysis and potentially some rectification work are required before the mine can restart.