GLG News by James May

Implications: Contract prices likely to soar in 2008 Rising supply and weaker demand may lead to weakness in 2009

Analysis: Spot Indian iron ore prices are currently around $175/tonne cif China. For those Chinese mills paying contract prices for Australian iron ore of $52/tonne fob, even a $30/tonne freight rate means they are getting a bargain. Everyone knows this and the only question is whether price increases for iron ore in the 08 contract are up 25% or higher.

I want to focus on 2009, because while the consensus is for further price gains then, I think the market will look very different this time next year.

Firstly, Chinese iron ore demand will slow in 2008. For the last five years, Chinese steel output has been able to grow faster than demand as Chinese steel initially replaced imports and then started to export. However, the imposition of export tariffs (likely to be extended) means that exports will be at best flat in 2008 compared to 2007 and could be lower. That means that steel output will rise in line with domestic demand - probably around 12-13% if you take the average of the last 2-3 years. This is significantly lower than the 20% plus steel output growth of the last few years. As a result, iron ore import growth will slow - our number crunching suggests a still large increase of up to 100m tonnes (raising the total to around 500m tonnes).

The problem is, we expect output growth to be more than 100m tonnes. In Australia, BHP Billiton is adding 20m tpy of capacity under Rapid Growth Plan 3, Rio Tinto is adding 40m tpy of capacity at Yandicoogina and others, while has a further 30m tpy at Hope Downs coming on late in the year or early 2009. Fortescue Metal Group has 55m tpy of capacity coming on at some point during the year, even if it is delayed. In Brazil, MMX will sell at least 15m tpy additional material as it starts its mines, while CVRD plans an additional 25m tpy of sales. CSN is also planning 15m tpy of extra sales. Kumba of South Africa will have an additional 10m tpy extra for sale, and a whole host of junior Australian, Brazilian, Chilean and North African miners will add around 20m tpy of capacity. A back-of-the-envelope calculation makes that around 200m tpy of extra sales in the market.

Can this additional material go elsewhere. Not really. European demand may be up 2-3% - possibly an additional 5m tonnes, and the same for Japan. Other Asia may take an additional 2-3m tonnes, while the Middle East could take up to 10m tpy. North America is essentially supplied internally.

Even if some of this capacity is delayed until later in the year, it will definitely be developed in 2009 and there are a whole host of other new mines coming on in at that point. The visibility of the market and the dynamics will have turned this time next year.

What about the consolidation of the iron ore industry you say. Well, 2008 will see the emergence of 2 new significant players. FMG and MMX will be in a position to sell more than 50m tpy each in 2009, and both intend to go to more than 100m tpy by 2011. CVRD wants to add another 100m tpy of capacity at least by 2012 from 2008. It may be willing to negotiate price in 2009 in order to ensure volume sales.

Implications: US steel industry gains direct benefit from market protection against the major sources of potential over-supply over the next five years. This won't insulate the industry entirely due to a spillover effect. ArcelorMittal illustrates its phenomenal lobbying power and this has to be a success for this company.

Analysis: The International Trade Commission (ITC) voted to keep for a further five years high anti-dumping (AD) tariffs on direct imports of Chinese and Indian (and Indonesia, Thai, Taiwanese, Ukrainain and Indonesian) HR coil. This means that steel from these countries cannot be sold directly to the USA (although Thai steel has only a low tariff). Given that the Chinese and Indian steel sectors are likely to be the most likely to cause global over-supply over the next five years, this has to be a positive for US steel pricing.
This is not to say that US steel prices may not be impacted by over-supply from these nations at some point in the future. They will export to other markets and displace material which will then be sent to the US market, but it does provide some protection to the industry due to the additional time lag and complexity.

The big winner in this is ArcelorMittal. It managed to persuade the ITC to lift duties against HR coil from Kazakhstan, Romania and South Africa. These are countries where it is the sole (or dominant in the case of South Africa) HR coil producer. The ITC appeared to buy its argument that the company would not therefore deliberately dump into the US market where ArcelorMittal is the dominant player. That may be true, but it does give ArcelorMittal the opportunity to bring in product duty-free and supply to the US market if required.

More importantly, it will be able to point to the ITC decision in other anti-dumping cases and argue that the ITC is a precedent and this may help them gain access to markets that they were not in before. ArcelorMittal is the world's biggest steel maker and is beginning to show its clout in its lobbying power. It, for example, is probably a significant driver in pushing Eurofer and the EU into complaints on China as well.

Implications: This "announcement" is not about substitution, it is about upcoming contract negotiations with its steel suppliers, and an attempt to influence the agenda to offset the supplier power of the steel industry. Substitution is occuring, and will continue to occur (even towards steel and aluminium) but it is not fundamentally price-driven and is not a short-term development. Magnesium has potential and is growing in terms of penetration, but price and performance mean that it has a limited role.

Analysis: Right now, US HR coil prices are $530/ton, whereas GM is paying around $560/ton or slightly higher on a base price. US steel mills are consolidating with Arcelor-Mittal, US Steel (having just acquired Stelco of Canada) and AK Steel probably supplying at least 80% of automotive steel to the US and Canadian industry.
The automotive and steel industry sit down about now to agree on prices for 2008 contracts. Despite spot prices being below contract prices, US steel mills are likely to push their supplier power to push up contract prices again in 2008, and there is little GM can do about it. They cannot afford to buy more spot material given the specific grades and volume requirements.
They have therefore resorted to bluster and trash talk with the threat of substitution. Remember, the US auto industry buys around 15m tons of steel per year, so a $10/ton difference can add to the bottom line. Will it work - probably not - but this is not an explicit threat, it is trying to set the agenda.
Significant substitution cannot take place in the short-run or the medium-term. The tooling set-up, design knowledge and the dominant force of inertia of steel as the prime material for the automotive sector is difficult to replace. Materials such as magnesium have been making some penetration into the sector for some years, but the technical development is slow and its price advantage is not there - having said this, Ford has been much more active in adopting magnesium than GM.

Implications: Pricing volatility has not been reduced. US HR coil prices have just dropped $150/ton (25%) in the last six months and could easily go up by the same amount in the next six. The difference is that the bottom of the current cycle is above most producers' cost level.

The theory is that consolidation will reduce price volatility as producers can reduce supply to meet demand, without the fear that others will simply supply in their place (i.e. no free-riding). That assumes a closed system. An open trading system is not closed by definition and even if the industry is consolidated on a continental level, it is not consolidated globally.

Consolidation has yet to be tested in a significant industry downturn. I would argue that consolidation in its current form will be useless in preventing prices going below operating costs in the event of a serious reduction in global demand, without an increase in protection.

Analysis:

While the major mature economies' steel industries (EU, North America and Japan) are consolidated into 3-5 major suppliers, there remain second-tier players that will continue to free-ride on the "responsible" producer cutbacks. We note AK Steel and Algoma and some Mexican suppliers were prepared to cut prices first and to lower levels than the larger players.

Moreover, 60% of world supply is in Asia, and here there has been little meaningful consolidation outside of Japan. There are around 100 steelmakers that produce more than 2m tpy of steel that are not in the global Top 10. State or family ownership here makes consolidation far less likely in these current good times.

Consolidation works best where there are barriers to entry either technical e.g. in automotive or packaging where there are quality restraints, or geographically where anti-dumping duties are in place.

Steel can travel globally and it can move to the markets where prices are highest. So even if responsible producers cut supply in one market in an effort to keep prices high, over-supply from other markets can come in and compete them away.

Steel demand growth has averaged over 8% in the last five years, and there is currently little spare supply capacity in the system. This is keeping prices high. However, that growth and profitability has spurred enormous capital expenditure on expansion.

Now imagine an Asian Crisis where global demand dropped by 10%. That would leave in excess of 120m tonnes of steel looking for a home. At that stage, prices would be forced down to below average costs, and the effects of consolidation would be really tested.

 

 

Implications:

Indian consumption was just over 30m tonnes of finished steel in 2005. To reach 200m tonnes by 2020 would require a compound growth rate in excess of 13% for a 15-year period. The previous 15 years has seen a compound growth rate of less than 5%.

Chinese consumption compound growth over the last fifteen years was 12%. Have Indian bureaucrats simply looked at the Chinese performance over the last fifteen years and said that we can do the same? If they have, then they could be setting themselves up for a fall.

 

Analysis:

For steel consumption to grow at 13% or so, GDP would have to grow by around 9-10% per annum, while fixed asset investment would have to grow by 20% plus. Historically, India has been nowhere near those numbers.

Taking a more reasonable long term forecast of 6.5% and steel consumption growth would be 9%, which is still strong. On a compound basis that would take demand in 2020 to around 110m tonnes - the previous forecast.

Why won't India reach the 13% growth level? Manufacturing in India only accounts for around 25% of GDP, compared to 60% in China. Services in India account for 50% and this is not steel-intensive.

Infrastructure investment will be the driving force for steel consumption. It has been rising in India over the last three years, but is now at around 10-15% year-on-year growth - not the 30% plus seen in China. With infrastructure investment subject to bureaucratic delays as well as the complexities of Indian federal versus state politics and the fact that the private sector is responsible for a significant proportion of the investment, the chance of reaching a sustained Chinese growth rate is unlikely, which was primarily state-directed. Could India push through a Three Gorges-type project in the same time? I think not.

I think the 13% compound growth target is overoptimistic. Steelmakers building to this run the risk of creating substantial overcapacity. Based on a very conservative view of planned capacity coming on-stream and steel consumption growth of 9%, India will build a structural excess of steel of more than 10m tonnes in the next five years that it will be seeking to export. That will be in competition with rising Chinese steel exports and continued protectionism in the target markets of the USA and Europe.

In the mid-1990s, the liberalization of the economy triggered economic growth and the entrance of new private-sector players resulted in a wave of debt-financed investments in the steel industry. A global slowdown in steel demand coupled with weakness in the domestic market over 1998-2001 resulted in multiple bankruptcies, some closures and debt restructuring. The industry did not become profitable again until global steel prices rose in 2004. Most of the planned capital investments of $75bn to build new capacity are to be financed with debt and, in my view, the industry is running the real risk of overextension. It is not being helped by overconfident forecasts and assertions from bureaucrats and politicians.

Implications: An examination of the recent major M&A activity in the steel market suggests that the majority of buyers have been companies with dominant shareholders.

These appear to have a higher appetite for risk than their publicly-traded brethren.

The result will be a fast dwindling band of listed steel companies with widely-dispersed shareholders, and those that are left will command a rising premium.

Analysis:

The thesis suggests that steel companies with a dominant shareholding group (family or individual) are taking on greater risks and are willing to pay more to become sizeable.

Looking at the recent major deals:

CSN and Tata are fighting over Corus. Tata has bought Millennium and Natsteel in the last two years.

Evraz has bought Oregon, Highveld, Palini e Bertoli and Stratcor in the last two years.

Mittal has bought Arcelor, ISG over the last two years as well as a number of other companies prior to that.

Novolipetsk (Lisin) has bought a 50% share of Duferco.

Tenaris bought Maverick and Hylsamex in the last two years.

Severstal bought Lucchini.

Even as Mittal-Arcelor was forced to divest in Europe, the buyers were single dominant shareholders in Duferco and Alfonso Gallardo.

Of course, there are exceptions that prove the rule, with Ipsco purchasing NS Group and of course Arcelor and TKS fighting over Dofasco, but I would suggest that that was an unusual case and an unparalleled opportunity for both.

However, we look at the relative inactivity of the other large global companies. US Steel, TKS, JFE, Nippon, Posco, Riva have not made a substantial deal between them for five years. The one thing that unites these companies is that they are relatively old and management is primarily by consensus and embedded in their corporate culture and memory is the fact that the steel industry for a long time was cyclical and hugely destructive of capital. When running the numbers, they tend to look at the downside cyclicals.

On the other hand, those companies with dominant shareholders recognize that they have only one chance to get to the top table and that is by acquisition.

With the number of steel companies with widely dispersed shareholder lists dwindling quickly, the chances of getting to the top table are shrinking fast...and consequently the competition will intensify.