Overview of key methods of global iron ore price modeling
0 likes83 views
Long-term iron-ore prices will be 30-40% higher than average industry forecasts of $75/t, largely due to depletion of existing iron-ore deposits. Three forecasting methods were used: marginal costs, global marginal incentive price, and big three marginal incentive price. All concluded prices will be $102-105/t by 2020, 30-40% above consensus. Depletion means new projects must supply 310Mt in 10 years to sustain production, but consensus forecasts do not account for this. Higher long-term prices are needed to incentivize the large investments required for new projects.
Overview of key methods of global iron ore price modeling
- 1. 22 Mining Journal June 8, 2012 SPECIAL REPORT IRON ORE BY ALEXANDER MALANICHEV AND ALEXANDER PUSTOV A NALYSTS at steel producer OAO Severstal have used three methods to forecast long-term real prices of iron ore, which they say will be crucial for valuing investments in greenfield iron-ore projects over a period of more than five years. The forecast was obtained using a marginal-costs method and two approaches based on incentive prices: the Global Marginal Incentive Price (MIP) and the Big Three MIP. The analysts conclude that there has been a structural shift in the iron-ore market and long-term iron-ore prices in US dollar terms will be 30-40% higher than average industry forecasts.This is largely due to one key factor: the depletion of existing iron-ore deposits. In addition, escalating industry costs are expected to keep nominal iron-ore prices at high levels in the long-term. IRON-ORE MARKET There was a long period of declining prices in real US dollar terms from the mid-60s to the early 2000s, which led to underinvestment in new iron-ore projects. Generally there was no incentive for companies to invest in greenfields projects when demand prospects and investment returns were questionable. So 10 years ago when Chinas steel production boom was set to emerge, tight iron-ore supplies and a strong growth in demand allowed prices to take off. From 2003 to 2011 the price of 62% Fe concentrate product increased seven-fold to approximately US$150/t on an FOB Brazil basis. These high prices make the iron-ore market very attractive to newcomers hoping to add supply to meet demand and earn solid profits. But in order to justify capital investments in new projects, future price assumptions are required. Given that iron-ore projects have long lead times, with engineering and construction taking up to 10 years, investments need to be evaluated in the long-term so Severstal has used 2020. Usually, long-term prices used in financial models are converted into real dollar terms to negate inflation. So the question is, where is the price headed? Is it going to grow further on the wave of Chinese development or fall significantly as seen in the past? According to consensus forecasts by the leading investment banks, the iron-ore price will ease to US$75/t in the long-term from US$150/t in 2011. At first glance this looks logical considering that the current price is exceeding the previous high by three-times and iron-ore reserves are sufficient to meet 40 years of consumption at current rates. Historical analysis of commodity price cycles confirms that peaks were often not sustainable. Moreover, the concept of a long-term price itself assumes that the market would cool down after the boom period. China would decelerate its steel production and iron-ore consumption growth, India would not become a second Chinaas it does not need that much iron ore to be imported. Iron-ore supply tightness should ease with the entry of new projects where, in theory, total volume in could account for almost 3,000Mt by 2020, if projects are delivered within the announced timeframes. However, there are several reasons to question whether the consensus forecast for long-term price properly reflects the new market reality: It is not clear what assumptions stand behind the consensus forecast as there is no explicit relationship between the supply-demand balance and iron-ore prices in the available models; There are no transparent models certified by the scientific community. At least, it is difficult to refer to any detailed publications on the matter; Traditional forecasts do not explicitly take into account iron-ore deposit depletion as well as industry operating and capital costs inflation; and, Based on a 10-year retrospective analysis, it is evident that consensus forecasts are usually lower than actual iron-ore prices. OVERVIEW OF APPROACHES The marginal cost approach generally assumes that long-term price equals marginal operating costs, ie the highest production cost needed to bring the last piece of supply to the market. The two elements of data required in this approach are operating costs split by country and assumptions regarding future demand growth. The marginal cost approach gives the lowest threshold of prices at a given demand, because it assumes that new projects, being marginal, will never pay back the invested capital. The second and third approaches use the concept of incentive pricing.This concept is based on the notion that producers will only invest if they believe that prices will be high enough to cover all their operating and capital costs and provide a targeted return on capital over the life of the mine. Both marginal investment approaches assume that the long-term price equals the marginal incentive price.The key difference between Global MIP and the BigThree MIP is that the former considers incentive prices for all announced projects, while the latter is based solely on incentive prices for projects owned by the three biggest producers: RioTinto, BHP Billiton andVale SA. In the long-run a number of key assumptions should be used to determine long-term prices including: Demand growth in 2011-20 is taken as compound annual growth rate (CAGR) of 3.4%, down from 7% over the past decade due to an expected slowdown of steel consumption growth in China; Iron-ore consumption is not dependent on price as iron-ore constitutes about 30% of rolled-steel Long-term iron-ore prices will be 30-40% higher than average industry forecasts. This is largely due to a key factor the depletion of existing iron-ore deposits A model futureEncouraging project development 0 20 40 60 80 100 120 Consensus Sep-Dec 2011Big-3 MIPGlobal MIPMC US$/t(real2010terms,FOBBrazil) $105 $105 $102 $75 RGP 5Simandou BHP BilliRio Tinto AustralGuinea 150 $70 $24 $83 $56 $39 -$12 $49 -$3 Operating costs Capex charge Freight difference 75 0 -75 US$/t(real2010terms,FOBBrazil) 22_23MJ120608.indd 22 07/06/2012 09:36
- 2. June 8, 2012 Mining Journal 23www.mining-journal.comwww.mining-journal.com production costs and the latter contributes 5-20% of the final costs of products such as automobiles or buildings; Excluding China, iron-ore capacity would have grown at an average level of annual capacity additions over the last five years: 88Mt.This equates to 880Mt by 2020, including 90Mt required to displace Chinas inefficient capacity. However, only 590Mt of this new supply will come to the international market and participate in the price setting; Chinas government follows rational economic behaviour and stops domestic iron-ore production if it is less cost-efficient than imported ore. In our base scenario, one-third, or 90Mt, of production would be stopped; Iron-ore prices are determined by a marginal producer based on a supply curve of international trade and Chinas domestic supply; and, Depletion of existing deposits is assumed at circa 30Mt globally each year, according to McKinsey estimates, this means that new projects need to deliver 310Mt to the market in the next 10 years simply to sustain production at current levels. MARGINAL-COST-BASED FORECASTS Gauging the 1,760Mt on the supply curve, which is based on operating costs, the long-term price comes out at US$105/t, or two-thirds of the price in 2011. This marginal cost is due to some high production costs in China, and at that price the country will produce 200Mt/y versus around 290Mt/y currently. Operating costs for new projects are assumed to equal ex-China weighted average operating costs as of 2010 due to absence of more precise data. However, this number could be an underestimate given that new mines are less cost efficient than old ones, with operating costs for some new projects in Australia reaching US$70/t. If we were to assume that supply and demand grows at consensus expectations of 5% and 3%, respectively, long-term prices will fall to US$51/t, but demand could easily accelerate to 5%, and the iron-ore price would jump to US$105/t. Putting these scenarios on a probability matrix and excluding the unlikely cases of slow demand and rapid supply growth, the long-term price would range from US$91/t to US$158/t with a 94% probability considering the correlation of different scenarios. The marginal cost approach does not take into account the capital cost element, which is quite important for development of new projects. INCENTIVE PRICE CONCEPT The incentive price is the level that encourages the entry of a new project by covering not only operating costs but also discounted capital investments and a return on capital. It should be mentioned that a calculation of incentive prices for different iron-ore projects also includes the normalisation to a single base of iron content, 62% Fe in our case. The future incentive supply curve was built using incentive prices for new projects and operating costs for existing ones.This approach is based on the assumption that new projectsdelivery requires the forecasted long-term price to be equal or higher than the incentive price. Moving on to the specifics, in 2020 iron-ore demand is projected to account for 1,760Mt, therefore, the long-term iron-ore price would equal US$105/t. This is the same price as derived using the marginal cost approach, and in fact the marginal producer is one and the same Chinas high cost iron-ore producer.The highest incentive price for a new project here is US$72/t.Therefore, operating costs as well as invested capital are fully covered for all new capacity. Again excluding the unlikely cases of slow demand and rapid supply growth, the long-term price would range from US$91/t to US$158/t with a 71% probability (or 94% considering correlation between scenarios) which is the same as the marginal cost approach. So, while the global marginal incentive price approach is easily explainable and takes capital investments into accounts, it still has most of the shortcomings of the marginal cost method: It requires a long-term demand growth assumption, which is hard to predict correctly; It requires futureincentivesupply curve, which is built on fragmented data regarding project operational costs and capital investments; It is highly sensitive to small changes in demand as the 4th quartile of the supply curve is very steep; and, It assumes the iron-ore market to be in perfect competition, while it really is an oligopoly. The next approach, based on the BigThree MIP does consider the oligopolistic character of the iron-ore market and does not require demand and supply projections, as the BigThree are assumed implicitly to provide this information in making their investment decisions. THE BIGTHREE APPROACH This method assumes that the iron-ore market is an oligopoly, where RioTinto, BHPB andVale are the most informed and influential market players controlling 70% of international trade and are expected to hold this influence in the long term.Therefore, the incentive price ofThe BigThree marginal project could be used as a benchmark for the long-term price. It also assumes that it is in the interest of the Big Three to underestimate capital expenditure of new projects to increase entry barriers for other companies and get favourable funding.This means that the approach estimates the lower boundary of the long-term price. The long-term price is US$102/t, which is the incentive price forVales Carajas Serra Sul project. One can see that the marginal incentive price forecast is almost equal to the long-term price in the other methods. But it is clear thatVales Serra Sul is not on the futureincentivesupply curve.There are two possible reasons for this: Firstly,Vale could have more optimistic views on either demand growth or different views on the set of new projects coming on stream. For example, they could assume that the projects being developed by larger and experienced miners would be realized first, even if those projects are not the most efficient ones in terms of incentive price. Secondly,Vales marginal project could be an option that would be developed only if the market is favourable. If this is the case, the long-term price would be downgraded to the incentive price of BHPBs RGP-5 expansion. COMPARING LONG-TERM PRICE FORECASTS In conclusion, the long-term iron-ore price in real 2010 dollar terms would decline from the current level of circa US$150/t in any of the presented scenarios down to US$100/t. The consensus is more than 40% lower than our estimates, suggesting that forecasts included in the consensus do not include deposit depletion.This is very likely to be true because analysts typically do not explicitly mention this. While forecasts obtained by all of the approaches are almost identical, ranging from US$102/t to US$105/t, butThe BigThree approach seems to be the most straightforward method. However, the long-term price forecasts would not hold true if demand collapses, for example, if China experiences a hard landing. To make it more concise, if in the next 10 years China experiences a 10% reduction of steel production compared with 2010, the long- term price would sit on the low marginal cost level (roughly US$50/t). Considering different scenarios of supply and demand and their correlation, the long-term price would range from $US91/t to US$158/t with a 94% probability. Presented forecasts are in real 2010 dollar terms and so exclude dollar inflation and industry operating costs escalation (wage growth, local currencies appreciation, etc). Using a halved historical industry deflator of 6%, long-term prices escalate to $US185/t by 2020.This is a 15% growth rate compared with nominal 2011s US$160/t. So, the actual probable peak is in nominal iron-ore prices is yet to come. Alexander Malanichev is head of strategic marketing at Severstal; Alexander Pustov is senior analyst, strategic marketing, at Severstal From 2003 to 2011 the price of 62% Fe concentrate product has increased seven-fold to approximately US$150/t on an FOB Brazil basis Carajas Serra SulRGP 5 ValeBHP Billiton BrazilAustralia $83 $105 $56 $70 $35 $39 -$12 22_23MJ120608.indd 23 07/06/2012 09:36