Analysis of the Proposed Amendment to the LME Warehouse Rules
October 21, 2013
By Alcoa

  1. Executive Summary
    1. This paper is made available by Alcoa Inc. ("Alcoa"). It outlines the background to the LME's proposal to include a linked load-in / load-out requirement in the LME Policy Regarding the Approval of Warehouses (the "LME Warehouse Rules") (the "Proposal"). The paper explains Alcoa's serious concerns in relation to the Proposal.
    2. Alcoa holds itself out to investors as providing potential shareholders exposure to the aluminium market – Alcoa does not engage in any significant hedging activities around price movements in aluminium. Thus Alcoa’s business depends on achieving fair prices determined by supply and demand fundamentals. This in turn requires a market place that is fair, transparent and well-functioning. Alcoa believes that the fairness and proper functioning of the aluminium market is seriously jeopardised by the Proposal.
    3. Alcoa believes that it is fundamentally wrong for an exchange (which should act as a neutral platform for buyers and sellers to interact) to alter the rules of the market with the express aim of moving prices (in this case regional premia). Regardless of whether the LME succeeds in moving prices, Alcoa believes that all market users and the economy as a whole will ultimately suffer if prices are to be influenced by LME policy rather than being determined by supply and demand fundamentals. Furthermore, the Proposal will not result in any countervailing benefits that could justify the damage to the operation of the market. The Proposal is not an appropriate or effective means of achieving the LME’s own stated goal of reducing warehouse queues.
    4. Large stocks of metal have built up in warehouses throughout the world, including in LME-approved warehouses, as a result of the 2008 financial crisis and its lingering effects. These stocks are largely owned by financiers who bought the metal to hold as part of financial transactions rather than for physical consumption. As financiers have requested the delivery of large volumes of metal out of LME warehouses in order to move it to non-LME storage, queues have developed at some warehouses.
    5. The LME has proposed the change to its rules in an effort to reduce the queues and to reduce the regional premia (the difference between LME on-exchange prices and the actual price of physical delivery of metal). In addition to the implications of the LME’s conduct for overall market integrity outlined above, the way in which the LME has developed the Proposal, and the substance of the Proposal, raise serious concerns.
    6. First, there are serious questions as to whether the LME has adopted a fair and transparent process in formulating and publishing the Proposal. Concerns include: (i) an apparent lack of equal treatment of market participants; (ii) lack of a genuine public consultation; (iii) lack of safeguards against conflicts of interest; and (iv) lack of adequate consideration of alternatives to the Proposal.
    7. Second, as outlined above, the LME has acted outside its proper role and purpose by proposing a rule change with the intention of influencing the level of prices (regional premia) that would otherwise be determined by market factors. This sets a dangerous precedent as regards the proper conduct of investment exchanges.
    8. Third, the LME does not appear to have given adequate consideration to the harmful effects which might arise from the Proposal. These are likely to include reduced transparency, increased volatility and the possibility of market distortions, such as artificial backwardations.
    9. Fourth, the Proposal fails to address the underlying issues confronting the market. For example, the Proposal fails to address the market’s need for proper hedging instruments to help purchasers of physical metal to manage premia (instead, it simply seeks to reduce premia artificially).
    10. The paper concludes by outlining the measures that are necessary to ensure the industry can have confidence that any future rule changes have been subjected to an open consultation process and thorough evidence-based assessment.
  2. Alcoa’s Interest in the Proposal
    1. Alcoa is a producer and a consumer of metal. Its business model is based on exposure to prevailing aluminium prices and it does not, for example, engage in significant hedging of aluminium price movements. Alcoa’s revenue streams, investment decisions and overall business model are fundamentally dependent on metal prices (comprised of the LME price and regional premia).
    2. Accordingly, Alcoa has a vital interest in:
      1. achieving a fair market price that reflects supply and demand – without the risk of that price being distorted by outside influence from the LME; and
      2. being able to plan its business rationally, including making forward-looking investment decisions based on forecasted supply and demand – without having to factor in outside influence from the LME or the uncertainties of potentially large inventories of aluminium in opaque, non-LME locations.
    3. Fair market-based prices can only be achieved in a well-functioning market place where prices are determined by supply and demand fundamentals. However, in formulating the Proposal, the LME has sought to exercise its own influence over a component of the aluminium price (the regional premium). In discussing the Proposal, the Chairman of the LME, Charles Li, reveals the following motivation:  "I view a persistently high premium as an unhealthy physical condition, such as swelling, that is not life threatening but is highly irritating and has the potential to cause long term harm. This can’t be solved with either a bazooka or a surgical operation. Instead, a modest approach, such as ‘Chinese Medicine,’ might be the cure. That is what this consultation is all about."1
    4. Alcoa is one of many market users who have been concerned by the LME taking it upon itself to seek to influence market prices. Regardless of whether the Proposal has any actual impact on prices, the LME’s action has caused a great deal of turmoil and confusion that is inconsistent with a well-functioning market. Implementing the Proposal would only compound the damage and would set a precedent whereby market users must expect that metal prices will be determined not only by supply and demand but also by LME policy.
    5. Furthermore, the Proposal will not result in any countervailing benefits that could justify the damage to the operation of the market. Even taking the LME’s stated objective of reducing warehouse queues, the Proposal is a misguided and inappropriate measure that is unlikely to result in long-term benefits for market users.
  3. Background on the Aluminium Market and LME Warehouse System
    1. The factual background on aluminium market pricing and the history of the development of queues at LME warehouses helps to demonstrate that the Proposal will neither address the underlying cause of queues nor result in genuine benefits to metal users.

      Aluminium pricing
    2. Almost all aluminium that is sold to consumers who will use it in a manufacturing process (known as the physical market) is sold at a price comprised of:
      1. a price set on the LME exchange (usually the daily spot price); and
      2. a regional premium.
    3. The regional premium reflects costs associated with a number of factors which are not relevant to the LME price, including:
      1. the costs of transportation (physical contracts provide for delivery to the customer, whereas LME contracts are for collection of the metal at an LME warehouse chosen by the seller), insurance, and handling; and
      2. regional supply and demand conditions. The three main regions are North America (where Platts Midwest Premium is the most common reference), Europe (Metal Bulletin Rotterdam duty-free and duty-paid are most common), and Asia (Platts Main Japanese Ports). China is not integrated into the LME system.
    4. The LME has a near monopoly on exchange-traded derivatives based on aluminium. But the LME’s influence extends beyond the futures markets. As outlined above, aluminium prices charged by producers and paid by purchasers are almost all set with a reference to an LME price and an applicable regional premium. Thus any actions the LME takes that affect the LME price or regional premia are likely to have an immediate and tangible impact on the physical market.

      LME warehouse system
    5. The LME warehouse system is intended to:
      1. act as a supplier of last resort for the physical market (as noted below, the LME warehouse system is ill-designed to serve consumers regularly); and
      2. provide a means for arbitrage between LME prices and physical prices.
    6. Each LME trade notionally involves a contract for physical delivery. If either the buyer or seller chooses to settle their LME contract physically rather than financially, then the seller has to present the buyer with a "warrant" that represents an interest in a specific lot of aluminium in an LME warehouse. Only LME-approved warehouses can be used to hold warranted metal.
    7. Consumers of aluminium do not often use the LME system to source metal, as the system is not well-suited to those with specific purchasing requirements. This is because:
      1. the LME rules dictate that the seller can present a warrant for metal at any warehouse in the LME system, such that a buyer will pay for the aluminium before the buyer knows where the metal is stored. The LME system consists of 37 locations (each location is a city that usually contains multiple warehouses) in 14 countries. Thus a buyer of aluminium on the LME with a rolling mill in the UK might well end up with a warrant for metal in a Detroit warehouse in the United States; and
      2. the current LME cash-in-advance structure does not coincide with consumers’ financing demands and the reality of extended payment terms.
    8. Certain warehouse owners have been paying depositors incentives to place metal in their warehouses, thus increasing the volume of metal upon which the warehouse can earn revenue. A warehouse’s revenue derives from (i) rent; and (ii) a free on truck ("FOT") charge (i.e. the charge to have metal removed from the warehouse). Depositors pay daily rent until they transfer the warrant to a purchaser via an LME trade. The purchaser will pay rent until the metal leaves the warehouse (and will pay the FOT charge to have the metal removed). It is reported that the LME receives around 1% of the LME warehouse revenues.
    9. The LME regulates the minimum amount of metal that a warehouse location must load-out each day. From pre-2008 until 1 April 2012, the minimum load-out rate was 1,500 tonnes/day for the largest warehouses. On 1 April 2012 (following a formal process which began in July 2011), the minimum load-out rate for the largest warehouses was doubled to 3,000 tonnes/day.
  4. Background on Warehouse Queues
    1. It is generally agreed, and the LME has repeatedly stated,2 that broader macro-economic conditions created by the financial crisis caused a large build up of aluminium inventory as a result of the following factors:
      1. demand in the physical market dropped;
      2. decrease in supply was delayed because it is expensive to shut down a smelter and there were continued substantial capacity increases in China and the Middle East after 2008; and
      3. historically low interest rates and prolonged market contango (a situation where prices for selling to a future delivery date are higher than the cost of purchasing metal at the current spot price) made it profitable for financial players to purchase the resulting excess supply of spot metal, sell it forward at a profit, and hold it in a warehouse.
    2. Another important factor was that, as outlined above, certain LME warehouses were paying large incentives to attract metal away from competing warehouses and to induce depositors to bring metal into their locations. This led to a large accumulation of aluminium in a few warehouses (in Detroit and Vlissingen in particular).
    3. It has been reported that stocks of metal at LME warehouses rose from 1.5m tonnes at the start of 2008 to 6.2m tonnes by the end of 2009, with stocks in July of this year at around 7.7m tonnes.3 However, aluminium in LME storage represents only a small proportion of the total global market (which is reportedly in the region of 45m tonnes).
    4. Alcoa understands that inventories at non-LME warehouses have also increased substantially post-financial crisis. As with the LME warehouse inventories, this increase in inventory is a result of demand for metal from financial investors who hold metal in storage for significant periods (as opposed to using it in a manufacturing process).
    5. It is against this background that long queues have formed at a few LME locations (five of the 37 locations, with Detroit and Vlissingen having by far the longest queues for aluminium). In particular, when the financial investors who purchased LME metal have sought to move large amounts of metal out of LME warehouses into other storage, the waiting times for delivery of metal out of some LME warehouses has increased. Queues of more than 100 days have been reported at some warehouses.
  5. Aluminium Prices Post-Financial Crisis
    1. Contrary to some market commentators’ views, it is not the case that long queues have resulted in higher overall metal prices. In fact, the total price for delivered aluminium continues to be well below the pre-financial crisis levels.
    2. First, as outlined above, queues have not resulted in any material restriction on the supply of metal. Large amounts of metal are available for purchase outside of the LME system. Second, there is evidence to suggest that aluminium prices are not correlated with LME warehouse queues. From 2008 onwards, warehouse queues have generally been increasing, whereas aluminium prices have moved up and down:
      1. in 2008, pre-financial crisis, the price of aluminium was over $3,000/tonne in the physical market (LME spot price plus the regional premium);
      2. the physical market reached a low of about $1,400/tonne (approximately a 60% drop) in February 2009. The price recovered to about $2,900/tonne in April 2011; and
      3. the price then dropped to around $2,000/tonne in July 2013.4
    3. The LME has stated that its warehouse system, including the queues, is not causing any overall price increases in the aluminium market.5
    4. The relative weight of the components of the physical price has changed substantially since 2008. Pre-crisis, the regional premium was less than 3% of the overall physical price. In 2013, the Midwest regional premium reached a height of about 12% of the overall price, before dropping to below 10% more recently.6
    5. The LME portion of the physical price can be hedged easily by purchasing financial instruments on the LME exchange. However, the regional premia, which have become larger portions of the overall physical prices, are expensive and difficult to hedge because this has to be done in opaque, illiquid OTC trades.
    6. The LME derives no direct or indirect revenues from trading in products based on the regional premia, but it does derive substantial revenue from trading in LME-denominated products.
  6. The LME Proposal
    1. The LME proposes to tackle long queues by requiring warehouses with a queue of more than 100 days (calculated initially over a Preliminary Calculation Period from 1 July 2013) to load-out 50% more metal than they load-in until their queue is less than 100 days.7
  7. Concerns Regarding the LME’s Procedure
    1. The procedure by which the LME has formulated the Proposal is unsatisfactory. One industry participant has noted that the Proposal represents an "unprecedented intervention" in the market by the LME.8 Yet the LME has not conducted an open and transparent consultation in formulating the Proposal.

      Lack of equal treatment
    2. There are question marks over whether the LME has acted fairly and transparently in formulating the Proposal.
    3. Alcoa understands that the LME entered into non-public consultation with a limited group of LME users before announcing the Proposal. It is not clear:
      1. on what basis those users were selected for private consultation; or
      2. whether those users had a financial interest in the LME adopting the Proposal

        Lack of a genuine public consultation
    4. The formal public consultation is not sufficient to remedy the potential defects earlier in the process. The nature of the proposed rule change means that the Proposal itself had an immediate impact on market behaviour, so as to pre-empt a proper period for consultation.9 Indeed, the mere fact of the LME’s announcement that it intended to use the Proposal to reduce premia caused confusion and altered market perceptions, resulting in significant movements in premia since July.

      Lack of safeguards against conflicts of interest
    5. The Proposal appears to be in the LME’s own interest. In particular, the Proposal seeks to shrink artificially the regional premia (on which the LME receives no revenue) in order to maintain the prominence of the LME component of pricing (from which the LME receives revenue in the form of trading fees). The LME has taken this approach rather than creating exchange-traded regional contracts that would allow market participants to hedge premium risk effectively (see further section 10 below).
    6. It is also worth noting that the LME has announced that it expects higher rents (of which the LME receives a portion) to be announced in the near future. The LME’s consultation document notes that "Future rent increases for the 2014 rent cycle are a reality if such a policy change is implemented, thus increasing the disconnect between LME storage and other forms of storage costs."10 If these rent increases were to off-set the expected reduction in LME warehouse inventories, this would provide a further benefit to the LME from adopting the Proposal.
    7. However, the LME does not appear to have put in place any safeguards to ensure that its own interests do not conflict with maintaining a set of rules that is fair and even-handed to all users.

      Lack of adequate consideration of alternatives
    8. The LME has published the Proposal without producing any evidence to indicate that it has carefully studied its likely impact or assessed the relative merits of the Proposal when compared to alternative approaches.
    9. The Proposal seeks to redeploy a similar tactic to that which was used in April 2012, when load-out rates were doubled in order to try to address warehouse queue lengths. The previous rule change did not resolve the issues around warehouse queues – in fact, queues continued to lengthen. Notwithstanding this background, the LME has not proposed a significantly different course of action in its latest proposal.
    10. Furthermore, there is no indication that the LME has considered properly any alternative approaches to reducing queue length. For example, the LME has not made reference to the approaches outlined in an economic study by Europe Economics commissioned by the LME in 2010. Indeed, it is noteworthy that the LME has not published the full Europe Economics study, instead choosing to publish an Executive Summary only. Consequently, it is not possible to review the substantive details of the other policy options proposed by Europe Economics.11
  8. LME Exceeding its Proper Role and Purpose
    1. The LME’s motivation in formulating the Proposal raises serious concerns that it is exceeding the proper role of an exchange. If implemented, the Proposal would set a dangerous precedent as regards the proper conduct of investment exchanges.
    2. The LME’s owner has publicly acknowledged that the current queues have not stopped consumers getting access to metal and that overall aluminium prices (including premia) reflect supply and demand.12 However, as outlined above, the LME considers that premia are excessive and it is proposing the rule change in order to reduce the premia.
    3. The announcement of the Proposal has caused turmoil and confusion in the market in recent months. Regardless of what impact the Proposal ultimately has on prices, as a for-profit company, the LME is ill-positioned to play the role of a price regulator. The economic role of an exchange such as the LME is to act as a neutral platform for buyers and sellers and thus to enable price discovery (and associated activities such as hedging) based on economic fundamentals related to supply and demand.
    4. While it is the role of an exchange to develop rules that enable this process to occur efficiently and without artificial distortion, it is not the role of an exchange to form a view on the ‘proper’ level of price components and seek to use its rules to shift the prevailing price components toward its own view of the ‘proper’ balance. 10
    5. This is even more inappropriate when (i) the LME’s own economic interest is affected by the relative importance of the price components and (ii) the regional premia in question are not themselves formed on the exchange but are the product of several unrelated factors.
    6. In this case, the LME has formed a view that regional premia are ‘swollen’ (i.e. too high) and has formulated the Proposal as a means of reducing these premia. It has taken this course of action notwithstanding that:
      1. regional premia are a product of a range of legitimate market factors (supply/demand, transport costs etc., as outlined above) entirely unrelated to the operation of the LME; and
      2. the LME itself recognises that premia reflect genuine market conditions.13
    7. In taking this course of action the LME appoints itself as de facto price regulator for premia. This has damaging effects:
      1. it reduces the efficiency of price discovery for regional premia. The market operates less efficiently as prices are distorted by LME intervention as opposed to being determined by supply and demand fundamentals. For example, in times of over-supply, buyers would (rightly) be aggrieved if the LME sought to change its rules in an effort to prop-up prices and deny purchasers the low prices that market fundamentals were dictating; and
      2. it reduces confidence in the LME more generally as it is perceived to be open to making rule changes as a result of lobbying from vested interests. This may ultimately undermine confidence in its on-exchange products.
  9. Potential for Harmful Collateral Effects Arising from the Proposal
    1. A range of harmful collateral effects could arise from the Proposal. A responsible exchange would conduct a proper risk assessment and identify measures to mitigate such risks. However, the LME does not appear to have identified all of the relevant risks or proposed adequate safeguards to deal with these risks.
    2. The Proposal is likely to result in more metal being held off-warrant, such that LME inventory data is a less meaningful indicator of true supply and demand. The risk of 11 reduced market transparency is explained in Alcoa’s submission to the LME dated 10 September 2013.14 This reduced transparency could result in:
      1. greater price volatility: as the artificially depleted LME inventory levels become more heavily influenced by large amounts of metal being moved on/off-warrant at short notice; and
      2. greater potential for market abuse: as those with privileged information on the genuine inventory levels (i.e. including off-warrant metal) and upcoming metal movements are able to take advantage of this information in on-exchange trading.
    3. The LME itself has recognised that the Proposal might lead to distortions in the LME price formation process. For example, by linking load-in and load-out requirements, the Proposal will make it more difficult for warehouses with long queues to accept metal on-warrant, which could, for example, result in artificial backwardations (when the market is structured such that the nearby price is higher than the future price).15 However, the LME has not proposed any measures to mitigate the risk of such distortions occurring.
  10. LME’s Failure to Address the Underlying Problems
    1. The LME has sought to justify the Proposal on the basis that it is useful to reduce regional premia in order to maintain the effectiveness of the LME as a hedge for aluminium prices. However, this represents a short-sighted and misguided approach.
    2. Alcoa believes that market participants need to be able to hedge effectively the full cost of their metal transactions – including premia. However, simply using the LME rules to seek to depress premia does not remove the underlying need for premia hedging. As the LME recognises, premia are an inevitable market feature and are driven by market fundamentals. The industry therefore requires sustainable and market-based hedging tools, such as exchange-traded contracts based on the regional premia. These tools would allow premia to continue to be determined by market fundamentals whilst providing market users with an effective hedge. This approach is explained further in Alcoa’s submission to the LME dated 10 September 2013. 16
    3. However, the LME has not indicated that it will look to develop such contracts. Instead, the LME has simply sought to address the symptom by trying to depress regional 12 premia artificially. This does not constitute a responsible approach to developing robust price discovery mechanisms and hedging instruments that will serve the metal industry over the long term.
    4. Similarly, the LME has not taken any action to address ongoing concerns in relation to market transparency. Alcoa’s submission to the LME dated 10 September 2013 suggested that the LME should adopt an approach similar to the CFTC’s Commitment of Traders ("COT") reports. Publishing COT reports would help to clarify for market participants the impact and relative influence financial investors have on the price discovery process, thus increasing transparency on important market dynamics.
  11. Next Steps
    1. Alcoa considers that the current Proposal and the process by which it has been formulated are irreparably flawed. In order to address the concerns it is essential that the LME:
      1. suspends any imminent plans to implement the Proposal;
      2. publishes any economic analysis it has conducted or commissioned on the cause of warehouse queues and likely effects of a linked load-in / load-out requirement. If no such analysis has been conducted, it is essential that such an analysis is conducted and published;
      3. publishes the full version of the Europe Economics report relating to warehouse queues that was commissioned by the LME in 2010;
      4. undertakes and publishes a cost / benefit analysis of the type that the Financial Conduct Authority would normally undertake before making changes to its own rules. This should include a thorough assessment of the harmful collateral effects that might arise from any rule change;
      5. addresses the underlying problems confronting the market, including the lack of an effective hedging instrument for regional premia and the need for greater market transparency; and
      6. undertakes a new consultation process only after completing the above steps. The new consultation should (a) be fair and transparent, offering all market participants an equal right to be heard, (b) safeguard against conflicts of interest and (c) give full consideration to the range of alternative approaches.


18 October 2013



Analysis of the Proposed Amendment to the LME Warehouse Rules





Related News

October 21, 2013 (Reuters)

Alcoa Complains to UK, U.S. regulators about metals warehousing




Related Press Releases

September 11, 2013

Alcoa Urges LME to Improve Transparency for Clear Aluminum Price-Setting and to Enable Premium Hedging for Price Risk Management




Related Content 

Alcoa’s Response to Proposed London Metal Exchange (LME) Warehouse Policy  

By Alcoa 

September 11, 2013 



Alcoa Calls for Increased LME Transparency in Statement to U.S. Senate Committee  

By Alcoa 

July 23, 2013 


1. See Charles Li’s blog


2. See, for example, the section of Charles Li’s blog entitled "What macro-economic factors lead to warehouse queues?"

3. Jack Farchy, "Metals pricing under threat from warehousing rule change", Financial Times, 4 July 2013.


4. Based on the LME cash settlement price and Midwest premium.

5. For example, Charles Li states in his blog: “the metal price has nothing to do with the queues, since it is the availability of metal on the open market that determines price and there is plenty of metal available there. I had talked last year about a surgical operation if the price of metal was being distorted, but we’ve determined that this is not a problem”.


6. See Figure 3 in Alcoa’s submission to the LME dated 10 September 2013.


7. See LME Notice 13/208: A201:W076.


8. See statement by RUSAL’s CEO, Oleg Deripaska, 25 September 2013.


9. See Maria Kolesnikova and Agnieszka Troszkiewicz, “Aluminum Costs Seen Dropping as LME Unclogs Depots: Commodities”, Bloomberg, 7 October 2013: “The plans may already be altering the flow of metal to depots, with average daily arrivals dropping to 5,849 tons last month, from about 19,000 tons a year ago.”

10. Paragraph 18 of LME Notice 13/208: A201:W076.

11. This is in contrast to the approach adopted by the LME to previous proposed rule changes. For example, in 2007 the LME published the full economics report commissioned in respect of a proposed change to FOT charges.

12. In relation to queues and prices, Charles Li’s blog states: “it's very important to note that warehouse queues do not stop real-world metals users from getting access to metal…the metal price has nothing to do with the queues, since it is the availability of metal on the open market that determines price and there is plenty of metal available there…The size of the premium depends on supply and demand”.

13. Charles Li’s blog states: “This premium, which is usually a small percentage on the LME price, is charged to those who want immediate access to metal in a particular location. The size of the premium depends on supply and demand in that particular location. Premiums have always been a feature of the market”.

14. Alcoa's Response to LME 

15. Paragraph 15 of LME Notice 13/208: A201:W076.

16. Alcoa's Response to LME