2015 GS commodity outlook 中英文版 v

基础，大宗货物和黄金

矿业?新常态? – 中国经济放缓，成本降低

中国的?新常态?和成本通货紧缩是2015年的重要主题。

中国继续转变，成本快速通货紧缩。

我们在中短期的金属和矿业商品定价中看到两个主要的逆风情况： 中国的?新常态?；和成本通货紧缩。在发达世界中需求增长的缓慢加速，不足以抵消美元价格发展的影响。 成本通货紧缩第1部分：能源/其他大宗商品价格的降低 成本通货紧缩源于能源价格（石油、煤炭和天然气）的降低，以及钢铁、铜、水泥和其他大宗商品价格的降低。

成本通货紧缩第2部分：美元刚刚开始走强 鉴于大宗商品以美元计价，美元走强也是成本通货紧缩的原因。 降低大宗商品以美元计算的生产成本，美元比大宗商品生产国的本国货币更受青睐。

更倾向于钯金、镍和锌；看淡铜和黄金 鉴于我们对矿业商品的相对偏好性没有发生太大变化，我们降低了大部分中短期价格预期，以对现实和预期价格通货紧缩进行反应。

此外，尽管铜价与去年相比降低了20%，我们仍然看淡铜价，我们不仅将最近的变动视为基本价值推动（成本快速降低，需求增长变弱），我们预期在未来的12个月内基本价值进一步降低。 实际上，我们新的12-月铜价预期为$5，200/t（当前价格-c.9%），我们从从该预期看到了下跌的风险。

整天上将，在我们的新预期中，从12个月视图来看，矿业商品中最具下跌风险的商品为铜(-c.9%)，金（-c.8%）和铁矿石(-c.6%)。 .相反地，从我们12个月的视图来看，最为看涨的商品为钯金 (+24%)，镍(+22%) 和锌 (+16%)。 与我们观点相一致的是，因为钢铁由铁矿石和焦炭制成，钢铁最有可能降低能源资本支出（钢铁需求的c.10%）。 镍则正相反 (通过不锈钢市场) ，在石油和天然气价格下跌的背景下，生活消费品需求最有可能加速（30%）。 - 内容

矿业商品：因为实际及预期的成本通货紧缩，需求增长疲软，我们降低了我们的预期 2015年的3个关键主题：成本通过紧缩、供应差异化和中国的?新常态?6 铜运营成本通货紧缩9

铜：为下个范围设置的场景 11 铜：2015年将仍是艰苦的一年 12

铝：我们的短期预期是中性的，我们的中期展望是更加乐观的 14 镍下跌很有限，明显上升 16

锌：接近市场紧缩，相对乐观 2015/16 17 金：欧洲和日本应增加看空黄金的前景 18

PGM：鉴于需求令人失望，重组仍是关键；应该注意白金，钯金仍是我们最爱的品种 24 铁矿石:一场痛苦的消耗等待战争 32

冶金用煤:均值回归在起作用: 热能煤:达到退役期 36 碳酸钾:绕道回归正常 38 矿砂:困于流沙之中 41 披露附录

矿业商品：因为实际及预期的成本通货紧缩，需求增长疲软，我们降低了我们的预期

我们为矿业商品更新我们的价格预期(表1，变动为渐变的紫色，之前的预测在括弧之内)，这构成了我们分析师对其覆盖矿业公司进行预测的基础.在我们的预测期间基础、大宗货物和黄金总体上降低c.10%- 20%。

我们预期变动的主要原因为成本通货紧缩 - 受实际和预期美元走强、能源和其他投入成本降价的综合影响，我们预期矿业活动会有所改善。 我们对前一个周期的分析表明，生长率增长可以维持较长的时间，因此在最近牛市中的效率损失，可以在即将到来的10年中反转（参见表2和3）。结果就是，恒业成本曲线将会向下并随之走平。 此外，中国和欧洲需求弱于预期的情况，通过增加边际生产能力以及进一步推动未来需要新的能力，同样促成了通货紧缩的环境。

以前也发生过这样的市场动态（在20世纪80年代和20世纪90年代），但是2014年通货紧缩的步伐较为引人注目，同时在石油价格和FX也有剧烈波动。 但是，我们相信一些局外人有可能逆转价格下跌的趋势。 大宗商品在最后的牛市中已经不受青睐，因此，例如铁矿石和铜等大宗商品将异常的短缺资金，其结果也就是在我们的预测期将会看到更好的基本价值。

我们以更加详细的岩石 & 矿石列出我们降低大宗商品价格预期的基本原因。在铁矿石行业将面临一场持久的消耗战，热流传感器：热能煤达到了退役期，以及岩石&矿石： 冶金用煤平均值回归（今天都有公布） 。

在基础金属中，我们已经看空铜超过一年，在这期间，铜价下跌了20%，我们最新的成本分析表明，在现行价格的基础上， 2015年将会有进一步下跌；这是我们的新基础情况。 与此同时，因为供货周期的差异化，我们预期镍、锌和钯金在2015年的表现将大幅超越铜。 因此，我们重申我们的卖空铜/持有镍的贸易建议，我们在2014年11月20日(+6%)公布的该消息。 除了镍这一例外情况以外，我们新详细成本分析支撑我们前期的$17，000/t LT预期，我们对基础金属LT的价格进行10%的下调。 基础金属短期预测 从3个月来看，我们看多镍（我们的新预测在当前点位看多c.8%），谨慎看空铜(-c.4%)。 从12个月来看，我们看多镍（+c.22%）和锌(+c.16%)，谨慎看多铝(+c.2%)，看空铜(-c.9%)。

就基础金属评估我们2015年的偏好时，在其他事物的基础上，我们考虑以下因素： 供需平衡的前景，

能源和其他投入成本通货紧缩的相对风险。 美元成本通货紧缩到的相对风险， 消费者可自由支配开支的相对风险，和 能源行业资本支出的相对风险。

2).铜费用在大部分这些措施中，都处于最差的情况，因此我们下调我们的3个，6个，和12个月预期至$5，500/t (从 $6，600/t)，$5，500/t ($6，200/t)，$5，200/t ($6，000/t)（表 2）。

我们预期相对于需求，供应会出现冗余，以及根据疲软中国资产市场和乏力欧洲需求前景高风险，设定的20年1次的供应周期。 实际上， 我们预期中国的需求会持续疲软，15年第1季度SHFE和LME的现货库存会有所增加，在近期非常可能为价格造成压力。 在15年第2季度，我们认为铜价反弹的程度，将取决于中国政府在15年第1季度的采购数量，以及需求增长季节性上升/下降的强度。

更广泛来讲，鉴于大部分的边际铜生产商都是中国和美国以外的厂商，位于那些我们已经看到严重货币贬值的国家（从2013年开始，成本大于>$2.50/lb的铜生产商，平均折旧已经达到17%），铜价高度面临美元升值的风险。我们预期在未来的12-24个月内，会进一步贬值。

铜价也同样相对高度的暴露于消耗品和柴油成本通货紧缩的风险之中。 参见章节“铜成本通货紧缩，一种现象的研究 - 铜”。 因此，我们看到铜边际成本支撑，位于铜成本曲线的95%和90%，在2015年和2016年分别下降至$5，600/t和$5，000/t。 这些估计的风险处于我们观点中的下行风险。

最终，因石油价格下降，而引起的更强消费品潜能而言，仅有7%的铜作为生活消费品被消耗，与之对比的是不锈钢（镍）的c.30%。

与此同时，我们卖空铜/持有镍的建议（发布于2014年11月20日），目前上调c.6%，我们的目标在于20%的收益。 事实上，镍排在大部分这些措施的榜首，能源行业资本支出减弱的风险除外（在能源行业消耗了c3%的不锈钢）。 我们预计镍在2015年会大致持平或有小额赤字。 镍的边际成本主要是基于美元或人民币，包括中国电网的关税（我们相信NPI生产商电网关税在2015年可能在很大程度上保持不变）和基于美元的镍矿石价格。

此外，印度尼西亚矿石禁令表明鉴于2015年矿石储量的下降，镍矿石成本将持续上升，对较高的价格进行支撑。 -

对于铝，我们已经转变我们的观点至2015年大致持平（尽管我们持续期望强劲的中国以外生产商边际，跟随中国以外市场进入2014年的赤字）。 我们预期锌的价格在2015年相对走强，原因包括预期澳大利亚的世纪矿会关闭，下降的金、银和铜产量，造成了副产品的下降，以及中国对冶炼厂产能的限制。

2015的关键主题：成本通货紧缩， 供应差异化，和 中国的?新常态?

我们相信石油、铁矿石和铜，以及众多的其他矿业商品，已经从投资阶段进入了开采阶段，在该阶段在价格降低的情况下，通常需要提高生产率，这样可以降低成本，以顺应周期需要。 与此同时，我们认为我们正在见证美元的牛市，这将得到正在进行的页岩气革命和能源价格降低的支撑。 美元的持续走强和全球能源价格的下降，同样是成本通货紧缩的强大力量，这与21世纪最初十年矿业产品牛市的驱动因素相反，我们将此视为在未来12个月，矿业商品价格走低的背景。 这是因为处于平衡或盈余状态的市场，倾向于在生产边际成本周围或之下交易。

表3显示了这些动态是相关的，在某种程度上是自我延续的。

特别是，相对较强的美国经济，国内驱动和能源价格降低的支撑，意味着强势的美元，更宽松的美国货币政策，淡化的EM需求增长，贬值的商品生产国货币和矿业商品生产成本下降（成本通货紧缩），以及商品生产国可能出现的更多的供应量（与其他情况相比）。 这种背景下的成本下降和EM需求增长降低，是美国利率升高，促使资本从EM流回美国的结果，中国正在改革其经济增长引擎，从商品密集型FAI驱动增长转变至更少的商品密集的

消费和服务驱动增长。

从2009年开始，这种转变已经是中国GDP增长的强劲因素（特别是公司部门）。实际上，从2013年开始的打击腐败和放缓信贷（根据社会融资总额计算）已经放缓了中国经济增长和对矿业商品的需求。截止目前受影响最大的为钢铁，我们相信铜将在2015年和2016年受到影响，随着铜密集建设周期的完成，增长将在这一时期急剧下降。

如同2014年的情况一样，在本文中，我们持续将供应作为区分商品表现的重要因素。

铜运行成本通货紧缩

生产成本，特别是边际生产成本，对于商品市场的平衡或小幅盈余至关重要，这些市场倾向于在生产的边际成本附近产生交易。实际上，随着铜从过去十年的供不应求市场，到有适度盈余的市场，价格已经下跌，在铜成本曲线的顶部进行交易（表 13）.

我们预测边际成本，根据同成本曲线代表的95%，被设定至少降低t c.20%，从$7，000/t降低至2015/16的$5，600/t（表13）。我们预期有90%降低至大约$5，000/t。在适度盈余的时期，价格倾向于在90-95%交易，对我们在该范围内对铜进行的新预测，进行支撑。

这些成本预测将定高成本（2013年总现金成本>$2.50/lb）的铜生产商国货币产生20%的贬值（去年已经发生了17%，表14），柴油成本下降50%，电力、消耗品、服务和劳工成本不会发生下降。如果其他成本发生下降，或者如果剩余超出我们的预期，在未来12-18个月内，铜价格将有可能在$5，000/t以下。

我们可以将成本通货紧缩讨论分为两个部分，与美元走强相关的成本通货紧缩，以及直接铜矿成本通货紧缩（能源，劳工和服务等）：

1)生产国货币贬值对比美元（以本地货币计算的运营成本c.80%（表15））。表14显示了对于在2013年成本超过$2.50/lb矿山（成本超过矿产品20%），近期和预期未来货币波动。 从2013年开始，成本高企的铜生产商国家货币已经下降了c.17%，我们预期在未来的12-24个月内有进一步的4%的下降。鉴于没有收到货币救济，美国和中国的高成本生产商在成本曲线上显著上升（总共的高成本输出仅为180kt）。 实际上，我们找到了一个强有力的经验证据，证明边际铜生产成本和边际生产商国货币对比美元之间的反向关系（图16）。关于铜成本的更多细节，请参见我们2014年11月17日的报告，强势美元，弱势石油，疲软铜。

2)铜矿开采成本通货紧缩（在表17显示了在不同的成本元素中，N.B开矿成本的变化，倾向于如同表18所示，强烈正相关）： a)能源价格降低（直接成本的8%）。石油价格在过去的6个月中下不止下降了一半，历史上，柴油价格与石油价格几乎是1:1的关系。 b)消耗品价格降低（铜成本下降了32%）。消耗品包括钢球和炸药（石油衍生品），这些产品到的价格将随着钢铁和油价价格的下跌进一步下降。此外，炸药和钢球的生产商可能面临利润流失。

c)服务价格下降（直接成本的25%）。采矿行业的服务提供商有可能看到其利润降低，这是矿主服务成本的有效降低，在价格降低的现实面前，可能消减杂项开支。 d)电力价格有可能降低（直接成本的10%）。尽管鉴于其粘性价格，可能性不大。 e)通过增加生产效率，可能降低劳工成本（成本的25%）。在价格降低的情况下，进行重组

以降低成本和杂项开支。

f)在短期内将之抵消至一个微小的程度，会轻微的降低铜矿石等级（平均），增加融资成本。

铜：为下一次繁荣设置的场景。

我们在运营成本和新项目激励价格之间发现了一项重要并不断增加的楔形差距，表明在铜矿我们已经进入了投资不足的节段，因为从数据中很容易看到资本支出的减速，2014年，智利已经发生了资本投资的放缓。

更具体一点，项目激励价格和运营成本之间的楔形差距，是由相对粘性的资本成本驱动的，部分是由于项目质量恶化造成的（更加困难的地点，位于地缘政治风险较高国家的项目增加 – 也即‘可轻松实现收益的项目’已经被开采）。资本成本是矿主决定投资铜矿的主要考虑，‘激励性定价’的主要元素，激励性价格设定用于增加运营成本和价格（刺激性价格应降低10%，运营成本和价格可以下降20%-30%）。事实上，我们的新LT铜价预测为$2.70/lb，现行铜价为c.$2.56/lb，我们12个月铜价预期为$2.36/lb。 在这种环境下，随着矿主和投资者关注于铜的现货价格和扁平远期曲线，很有可能搁置铜矿项目的下一阶段投资。 从长期来看，我们相信随着行业矿石品位的下降，将会导致供应增长的下降，在这种情况下，将在2017/18看到需求增长趋势的收紧。

话虽如此，我们认为在现行铜价格基础上，转为牛市还为时尚早，考虑到中国和欧洲需求的风险，以及成本曲线的风险，我们认为仍倾向于下跌。

铜：2015年将仍是艰难的一年

鉴于在过去的4年中铜价已经下跌了45%，它试图在市场上反转。但是我们的研究表明，最近的10%-ytd下降和广泛的价格疲软都是根本因素驱动的（需求疲软，供应充裕，现货库存增加，以及成本前景的急剧下降），这些悲观的基本因素将在2015年，让我们见证铜价到的第5年下降。我们预测铜价将在未来的3-6个月内进一步下降5%，下降至$5，500/t，在未来12个月内，下降10%，至$5，200/t。话虽如此，2016年看起来更为平衡，2017年和2018年将看到铜供应紧缩，铜价反转的具体时间和水平尚未明确，下跌的风险主要源自全球需求下降。

我们接下来将探讨主要的驱动因素，并探讨围绕这一观点的要点： 我们的研究表明，铜价深受中国房地产的影响（表22-23），我们预测随着2015年和2016年，铜密集建设周期末端的完成，增长将显著放缓，紧随两年疲软的新启动数据。事实上，明显的钢铁需求下滑，对于后期铜需求来说，并不是好兆头。

与繁荣期间不同，由于已经较高的FAI，作为GDP的一个百分数和高杠杆率的中国企业的资产负债表，中国需求增长上行风险受限。

在过去的6个月中，欧洲现货铜溢价已经降低至2008/2009危机的低位，反映出比预期的欧洲需求疲软和/或中国对于欧洲出口产品的需求的下行（表21）。

中国国家电网的支出不能满足2015年和2016年的缺口。我们现在预期中国在2015年的需求增长为4%，国家电网铜消耗的增长为8%。

铜处于20年一遇供应周期的一半（从12年的后半年开始，直至2016年年底），紧接着是行业内的高资本支出投资，在过去的10年内，供应增长趋势从c.2%上升至c.4-5%。 我们的研究使用了100年的废弃金属数据，表明尽管价格上下降很多，废铜的可用性不会降低，而且有可能在2015年和2016年实现增长。这是因为实际价格仍然高企，在过去的十年

中已经建设好了成本较低的废弃金属收集基础设施，因此收集运营成本较低，废弃金属熔炼持续增长，全球需求增长指向新的废料。

铜生产的边际成本急剧下降，在强势美元对生产国货币、能源和其他商品价格通货紧缩，服务行业利润遭到侵蚀，在价格降低情况下，劳动生产率有可能增加，以及广泛强势美元环境的情况下（参见上述章节‘铜运营成本通货紧缩’）我们预期在未来2-3年中将继续下降。在我们的观点中，因为弱化的市场基本因素，该成本下降将从生产商转移至消费者。在未来1-2年中，成本结算将非常难以确定，但是，在接下来的月份中，这将成为大量争论的根源。 我们的‘倒置’供应和需求平衡持续指向2015年的c.500kt的过剩市场，市场将在2016年大致平衡。在2014年的500kt的购买之后，我们持续假设2015年的200kt 的SRB购买。 铜的长期前景是非常乐观的。我们不是结构性的铜熊市，我们同意长远时期的铜牛市。因此，尽管生产成本下降，我们可以看见2017和2018年的价格上升。

铝：短期而言，我们保持中性评级，我们的中期前景是较为乐观的

在经历了7年的熊市之后，中国以外的逐步消减铝产量，终于导致了2014年市场上出现了不足。中国以外的价格，在供应最受到约束的地方，市场价格急剧上升（与其2014年峰值相比上升>30%），利润扩展至2007年以来的最高水平。但是，近期需求增长预期被缓和，中国供应增长令人吃惊的上涨，导致在过去两个月中，中国和中国以外的价格上涨缓解。中国价格下降至前期低点（尽管一直在扩张，中国处于过剩状态），中国以外的价格仍然在生产成本溢价的状态交易（中国以外处于不足状态，尽管较少）。

对于2015年，我们继续认为中国以外的市场有丰厚的利润，中国市场较弱。在2014年中国铝冶炼FAI有25%的减少（源于价格低和信贷紧缩），促进了中国市场供需的平衡，紧随输出增长的降低，半成品出口增长的降低，以及价格的升高（源于中国成本曲线），因此在2016年之前，我们不会看到铝价格的全面上升。

当我们看到在2015年，全球市场将出现广泛的平衡乃至小幅度的盈余，在整个2014年，资金一直都是非常紧张的。最值得注意的是北美市场的内部消化能力很强，需求增长相对较为强劲。

所有中国以外的市场，至少在2015年价格应保持在中国价格之上，以便：a）吸引源自中国的半成品产品和b）刺激中国以外的生产重新启动。现在出口半成品产品的利润率为c.25%，我们已经符合标准a），但是我们仍需在中国市场以外达到c. $2,250/t的价格水平，这才能中国以外有充分的产能启动。该观点主要的下行风险在于，俄罗斯或巴西因为其货币贬值而重新启动其产能。进一步的下行风险在于中国供应持续让人吃惊，以及政府支持力度比预期要大（通过减税，冶炼厂获得资金建设电网，电力产生产能等）。

对于溢价，据报道（彭博，1月21日）LME为排队的终端用户审核租赁规则。我们预测假定2015年的溢价持平，但是预测期间的廉价出售，让溢价降低了大约$50/t，在2013年10月出版的“仓库交换经济”的观点与我们的工作相呼应。

镍下行非常有限，显著上升

我们认为现在的镍价代表一种强劲的买入机会，相对于铜尤其如此。尽管我们已经在全球需求预期疲软的基础上，降低了近期预期，我们的3个月和6个月预期，已经相应从$17,000/t和$18,000/t降低至$16,000/t和$17,000/t，我们的12个月预期保持$18,000/t不变，我们从现货价格看到了明显的上行趋势，为遭遇强烈价格支撑设定有限的下行风险（参见表40-41）。 镍矿石市场持续逐渐收紧，利润不断受到挤压的中国镍生铁生产商（NPI）开始在中国进行有意义的减产。展望未来，我们认为在四月中旬菲律宾矿石供应恢复之前，镍矿石市场将持

续收紧，而且即使菲律宾出口恢复以后，在年底也会吸纳库存。镍矿石市场的紧缩基本因素将在未来3-4个月内推动镍矿石现货价格的高企（已经增加了c.40% yoy），增加含镍生铁的输出成本。进一步，LME镍溢价将体现我们近期价格预测的上行。

从6-12个月来看，持续执行的印度尼西亚镍矿石出口禁令将使镍矿石市场持续紧张，即是菲律宾从季风季节恢复，该情况也不会缓解，因此我们认为1）中国的镍矿石库存已经下降至非常低的水平和2）不可能期待菲律宾镍矿石像2014年那样猛烈的增加，以弥补印度尼西亚造成的缺口。最终，我们保持长期镍价预期不变，我们已经与我们的股票研究同事联合进行了大量工作，以增加可信度。请参见其GS镍40：消费周期和供应链中断将推动镍价继续升高。

锌：市场紧缩，2015/16相对乐观

整体上，我们没有针对锌前景作出重大改变。一方面来讲，需求前景变弱，但从另一方面将，我们注意到了世纪矿山的关闭（大于全球供应量的2%），因为矿石品位的下降，在15年第3季度将受到责难，因此，整体上对于锌，我们保持适度乐观。事实上，对于锌需求的全球稳步增长，以及在2015和2016矿石供应增长的显著降低的组合因素，可以预见在未来12-24个月中锌将转入短缺。

虽然我们看不到任何短期反弹的催化剂，我们从强劲的锌成本支撑水平看到了锌价格的有限的下行风险。该成本支撑观点，主要取决于两个因素： i)高成本矿石供应位于中国，和

ii)锌矿收益的70%源自其副产品，包括铅、银、铜和金（最近所有这些产品都有显著的价格下降，或者我们预期在未来12个月内会有大幅价格下降）。 当锌收紧时，中国矿主和冶炼厂的反应将成为初始实例的关键因素。我们认为因为缺乏对新产能的投资以及执行越来越严格的环境控制，有可能出现锌冶炼的瓶颈。我们对锌的12个月目标仍保持在$2，500/t。

金：欧洲和日本仍然看空黄金前景

自从黄金价格从2013年中期降低以来，已经远远低于我们的预期。近期组合的支撑因素：（1）低于预期的美国经济数据；（2）欧洲宣布运行QE和（3）SNB令人意外的宣布移除CHFEUR原则，价格上升至$1，300/toz。

但我们仍预期黄金价格将在2015-16继续下降，尽管如此因为这些催化剂，我们仍将近期预期修订至一个较高的值，我们认为这些因素将很快被消化掉。随后，我们预期黄金价格将从15年第3季度继续下跌，与美国汇率升值周期的启动相同步。重要的是，虽然我们认为短期内黄金价格已经下跌，我们有信心认为鉴于在即将到来的几年中，由于低膨胀预期，以及由能源价格降低带来的采矿生产边际成本的通货紧缩影响，强势的美元和商品供应链转入开采阶段，黄金将会继续下跌。

当我们预期2014年黄金价格下跌，我们的经济学家预测美国经济增长加速，加速发生在欧洲、日本和全球经济增长放缓的情况下。这使得黄金价格大部分都是窄幅震荡，符合美国名义利率和实际利率都在窄幅震荡的价格行为，在整个2014年，两个市场仍表现出强烈的负相关性。这使得美国利率相对于其他G10利率升高，我们预期更高的利率水平。重要的是，汇率大致反映出真实利率的差异，我们的黄金价格模型展示了美元主导的黄金价格对利率水平的反应（参见2009年3月25日的，预测黄金作为一种商品）。其结果就是，当美元明显升值时，美元主导的黄金价格在窄幅震荡。

尽管以美元主导的黄金价格缺乏更多物质下降，我们持续预测黄金价格将在2015-16下降，考虑到经济学家预期上述的美国增长趋势将持续，随着宽松的金融和贷款条件以及降低的油价，帮助其余DM按顺序提升。特别的，当美国增长的加速至今已经在很大程度上转化为，美国实际汇率与其他G10国家汇率相比的相对增加，我们的利率策略师预期，美国利率开始收紧将最终提升绝对条款的美国实际利率。

在我们黄金价格框架中，这将转化为美元黄金价格的下降，尽管欧洲的QE和我们美国经济学季预期联邦基金利率将随后上升的组合，将有可能限制短期黄金价格的下降。

我们认为近期促进黄金价格升高的催化剂，将在现有水平对黄金价格进行支撑，因此，我们将我们3个月的黄金价格预期提升至$1,295/toz.

美国增长：与预期相比，近期的经济数据是令人失望的，在ISM制造业和非制造业，零售和一些住房相关的数据有所下降。重要的是，该连续放缓是发生在我们美国经济学家预测真实GDP增长，将从14年第4季度的4%以上降低至3%的基础上。

最后，他们预计增长仍坚定高于趋势，尽管由于油价下跌造成的有益净影响，以及强劲美元的逆风而带来的上行风险。他们的基线预期仍然为在2015年9月份基金利率提升，由于后来原因造成的风险，将持续造成令人失望的通货膨胀。

ECB QE： ECB今天宣布将在其资产购买项目中增加外债，该项目与前期预期相比相对较为积极，购买的规模和速度方面皆是如此（参见欧洲观点：2015年1月22日的ECB宣布取决于通货膨胀的收购项目）。

在我们的黄金框架下，欧洲的QE项目支持欧元主宰的黄金价格，主要是因为他施加欧元汇率的下行压力，同样引导EUR贬值。美元黄金价格的影响现在更为有限，但是，这取决于美国利率被欧洲利率拖低的程度。尽管金价在宣布时有长期上升，以及昨天的泄露对预期价格造成了巨大的影响，金价仍大涨。这与我们工作显示在美国，伴随QE的公布，对黄金价格造成了影响相一致（或如同美国QE3之前，广泛的推测相一致）。重要的是，我们的工作显示，EQ公布对黄金，而非实际现金流产生了重大影响（参见2012年12月5日的促进美国复苏的黄金周期）。

而这将暗示，已经大幅发生了针对黄金价格的反应，如果得到通货膨胀的保证，ECB的公告将进一步扩大该计划。结果就是，进一步增加项目的规模，如果通胀和通胀预期未能上升，将为黄金价格提供额外的支撑。重要的是，可能需要花费时间去实现该潜在支撑，而现行的黄金价格可能已经反映了现在的ECB QE项目。

SNB: SNB冲击对黄金价格的影响尚不明确：当CHF银行账号的负存款利率刺激将国内储蓄转变为其他避险资产时，黄金的历史性高价位波动（直至本周），与CHF相关度较低，自有的消极成本将有可能限制转化成黄金。此外，随着EUR CHF限制的移除，瑞士法郎再次成为外国人的避险资产，可以与黄金的避险资金流入相媲美，尽管随着本周的流动，CHF现在过滤与我们经济学家GSDEER公允价值估计相对过高的成分。我们的偏见在于，尽管我们的信念不是很确定，但从CHF转换为黄金仍受限制。

除去美国和全球复苏因素，尽管最近的欧洲和瑞士货币转换，我们仅预期黄金价格在一定程度上行，因为已经有很大程度的上升。尽管如此，我们同样看到了黄金价格的近期的有限催化剂，以追溯其近期收益。结果，我们提高3个月黄金价格预期至$1，290/toz，符合现行远期曲线。

当我们在向2015前行时，鉴于我们经济学家的预期，美国将持续在趋势增长以上运行，我们认为美元黄金价格将恢复其下跌趋势，随着宽松的金融和贷款条件以及较低的石油价格帮

助其余的DM实现后续增长。该预测中的关键观点在于，相对于2014，随着美国利率开始在15年第3季度收紧，作为催化剂，美国利率的增加实际发生于水平层面。我们分别将6个月和12个月以美元计价的黄金价格预测更新为$1,270/toz和$1,175/toz，我们2015年年底的预期为$1,190/toz ,2016年年底的预期为 $1,000/toz。 除了该利率观点，我们的预测反应我们的预期：（1）一旦美国利率上升，ETF将继续减少持有黄金（在过去三年中，保持该持有关系）；和（2）尽管近期观测到更加温和的不凡，央行（中国以外）将持续持有黄金。俄罗斯强劲黄金购买的减速将对我们的黄金预期产生下行风险。

2014年的重复将进一步放缓欧洲，日本并保持持续的低通货膨胀，对我们更新的黄金价格预期创造上行风险。尽管如此，我们继续认为全球经济增长的实质性放缓，将促使黄金价格持续高于现行水平。 重要的是，当我们对短期黄金价格下跌的信心下降时，我们对于长期黄金价格下跌的信心有所增加，因为除了在即将到来年份的低通货膨胀预期，以及能源价格降低对黄金采矿生产边际成本的通货紧缩影响之外，还有强势的美元和商品供应周期转入开采阶段的因素。结果就是，我们将2016年的美元主导的黄金价格预期从之前的$1,200/toz下调至$1,050/toz。

FX 和能源成本通货紧缩在释放成本压力 类似于其他矿物商品，黄金在过去6个月内帮助驱动成本通货紧缩的关键因素在于，在生产国（美国以外）降低能源价格和FX贬值。

能源输入直接占到矿物现金成本的10-20%，而原油的价格下降了约50%，纽卡斯尔煤炭在2014年下降超过25%。

由于充分的储存容量，以及需要将退出对页岩的投资，我们预期在2015年的大部分时间，石油价格将保持在当前水平。我们随后将预期价格上升，我们的修订降低了对长期价格的预期。对应的，能源输入成本降低的主要部分将被证明是持久的。

工资和其他本地货币主宰的成本将占据矿物现金成本的30-40%。我们预估，平均来讲，在2014年生产国货币兑美元的汇率已经下降了14%。基于GS预测，我们预期在2015年将会产生7%的贬值，在2016年，将会进一步产生12%的贬值，与美元相比，利率差异将会进一步扩大。

将这些FX和能源价格动态与我们的矿产品成本模型相结合，使用每个成本曲线的90%作为生产边际成本的代理，我们计算边际现金成本将从2013年的$1,000/toz下降至$850/toz（15%的下降）。我们首选的“全押”边际成本将从2013年的$1,200/toz下降至$1,050/toz（12.5%的下降）。

鉴于正处于商品周期的‘开采’阶段，可能还会有进一步的成本消减

正如我们之前的概述（2014年4月24日 当生产力回归时投资者将得到回报），通常商品供应周期‘开采’阶段的特点包括，成本通货紧缩的额外来源，大部分仅在黄金上开始生效：

已经开始有以本地货币结算的国内工资结算降低的迹象。

这种情况最清楚的例子发生在南非，在2014年，采矿业的薪酬几乎保持持平，而21世纪前十年，几乎都维持在两位数字的增长率。

从2013年开始，就业增长开始合理化发展，产出率逐渐走高，现在在增加劳动生产率方面有暂缓迹象。但是为了达到20世纪90年代普遍的劳动生产率，还有一段路要走。 我们期望看到资本生产率的升高，对新的资本开支进行大幅抑制。

鉴于在可预见的未来，预期矿石品位将持续下降，我们不会忽视持续技术进步的累计效应。实施上，矿石品位已经持续下滑超过1个世纪（平均-1.4%），但这并没有阻碍全球黄金生产的持续上升（平均+1.6%），远远超过实物黄金价格的增长（现在平均+0.6%，但是在历史上大部分都是0或负数）。

PGM：重组仍是令人失望需求的关键；对白金，钯金保持谨慎，仍然是我们赞成的观点

尽管为时5个月的罢工减少了全球矿物产量的c.40%，对于白金，2014仍然是令人失望的一年。2014年美元白金价格下降11%的原因包括：

1)库存释放和地面储量：3大生产商（Anglo Platinum，Impala和 Lonmin）根据其对行业行为延展时期的预期库存储备，也即，他们对服务客户进行清算。这与终端客户库存相结合。加上终端用户的库存，地面储量（包括回收者的库存）在罢工和启动期间，供应充足。 2)需求降低：2014年开始对欧洲汽车市场的预期较为乐观。但是，令人失望的汽车销售记录，再次打击了欧洲缓慢的宏观经济增长。近期的油价下降，可能会刺激消费者对汽车的需求，从而使白金受益。

3) Rand看跌：ZAR是Pt价格的关键决定因素。考虑到80%的白金矿石都源自南非，生产的边际成本在ZAR设定。

因此，弱势到的ZAR意味着弱势的美元价格，反之亦然。在2014年ZAR下跌了12%，对美元Pt价格有显著影响。

4)黄金价格降低：白金价格与黄金价格高度相关。在2014年强劲的宏观数据表明了美国的经济的态势，增加了美联储更早升息的可能性，这导致了整年间黄金价格保持静默，反过来，影响了白金价格。

5)没有重组：在5个月的罢工期间，3大生产商承诺采取重要步骤提升边际生产。但是，罢工结束后，关注点转移至迅速增加，而不是进行重组。

目前白金在生产边际成本附近交易 – 我们估算为￥1,250/oz（现货ZAR11.7 vs 美元）。但是，我们强调这可能随着ZAR的流动快速转变。同时，随着美元持续走强，ZAR的进一步下行可能为白金的美元价格带来更多的下行压力。

对于欧洲汽车需求，我们的基本需求低于趋势增长。我们的汽车股票分析师认为，2015年西欧汽车数量的增长，可能仅有略微的+1.5%。除此之外，他们预测会出现温和复苏，虽然只是位于一个仍旧低迷的水平 – 他们预测2015-18欧洲汽车销售增长率，将平均在2007年峰值水平的18%以下。ECB的QE和下行的油价可能成为汽车上行的催化剂，相关地，可以增加欧洲的白金需求。

供给方的重组更像是持续低价格的结果

从2011年开始，主要矿主的收益低于资本成本。Anglo Platinum，Impala 和 Lonmin都具备一些重组选择，但是都因不愿第一个大规模减少生产能力而采取行动。

如果需求情况没有明显改善，我们预期在2015年年底，一个大生产商将宣布关闭，这将成为价格的积极催化剂。Impala有望在2月份进行15年上半年的战略审查。

白金：尽管汽车增长和欧元出现积极态势，我们仍保持谨慎 6

短期内，如果在现行水平，价格低于生产边际成本，我们可以看到一些白金价格温和上涨的潜力。长期来看，我们仍保持对白金的谨慎观点 – 我们持该观点的原因有：

1.矿产品供应的反弹和二级供应的增加：我们预期主要和次要供应都会出现增加。开矿是一个有弹性的部分，公司控制成本的唯一途径是通过生产更多的盎司（将固定成本分摊至更多的畅销材料）。就其本身而言，我们希望二级供应出现反弹。同时，我们也可到了新生产力

的增加（在南非）同样看到二级供应（源于回收）增加，因为从2002-06更换了很多车辆（具备高Pt负载）。

2.欧洲汽车展望：考虑到欧洲疲软的宏观数据，主要汽车公司已经对其区域展望进行了平衡（戴姆勒近期将其2015销售预期从4%-5%调整至3%-4%）。其他主要生产商的说明 – 大众和福特 – 尚未令人安心。我们的汽车团队预期修汽车销售在2015年将会有温和的1.5%的增长。因此，我们预期凄恻需求仍然低迷，我们没有看到明显的上行迹象。

3.欧6 – 一种积极的增量，而不是变革者：我们一致认为，2015年随着欧6的实施，价格处于强劲的上行趋势。我们认为这将是一个积极增量，但不代表需求的显著上行。

4.投资需求减弱：即是在2013年，当白金市场位于一个c.1Moz缺口时，70koz投资前活动仅有一个小的不足。鉴于价格仍然低迷，我们预期投资需求仍保持稳定。萧条的欧洲汽车市场从欧6采取积极措施；近期的石油暴跌可能是一个潜在的催化剂。

汽车需求占年度白金总需求的c.40%，钯金的超过60%。因为其要求用于柴油发动机，而欧洲是一个c.50%的柴油市场，白金在欧洲面临更多风险。钯金更易受到汽油市场的影响，例如美国、中国和日本。

进入2015年，我们的汽车股票分析师预计西欧销售将持续保持萧条，增长率为c.1.5%，白金最终的市场，德国，英国和西班牙，预期为0增长。我们预期2015欧洲汽车需求增加c.2%至2040万辆。至于Pt负载，我们预期有一个小的激增（历史上，当规则改变时，通常会出现的情况）至每辆车2.25g，这转换为额外需求，将近100koz – 总需求仅增加了1.2%。 但是，石油价格近期的低迷，将刺激汽车需求的增长。在过去6个月中，石油价格下降了50%。石油价格的降低将刺激车辆需求，较多的车辆将需要较多的Pt负载。尽管我们预期短期汽车需求将被轻易满足，这将是一个正面需求 – 利好白金。

随着矿主试产扩量以及新的矿山开始投产，矿物供应将出现反弹；在罢工期间，回收也出现了增加，广泛预测将出现重组，关闭边际/亏损的矿山。但是，当罢工结束时，焦点转移至快速试产扩量，鉴于公司希望避免与工会进一步对抗，重组的讨论消退。我们希望在2015年矿山产量能得到反弹，完成试产扩量。同时新的矿山开始投产，包括PTM?s WBJV， Northam?sBooysendal 和 RBPlats? Styldrift，到2017年将增加c.400koz，应该看到供应反弹至2011年的水平。长期来看，我们看到因为新项目的投产，白金供应显著增加，例如PTM?s Waterberg和Ivanhoe Mines? Flatreef项目（现在都处于PFS阶段），达到了商业生产的条件。 进一步，我们看到了回收的贡献。在过去10年中，回收增加了超过70%，因为2003-07生产的车（每个车的峰值PGM期间），达到其使用寿命并被回收，我们预期增加额外的500koz（总需求的c.9%）。

白金处于不足状态 – 尽管是由投资引起的；投资前活动仍处于盈余状态；ETF外流价格的下行风险

除去投资需求，白金市场已经处于盈余，鉴于投资需求平稳，我们预期一个更小的盈余。2013年，因为新的南非ETF，投资需求明显上升。但是2014年是令人失望的一年，当罢工结束时，价格没有上扬。因此，2014年ETF增加仅仅超过了100koz。

ETF的一些关键点：1）鉴于价格上升时收益的唯一来源，ETF没有收益2）与黄金不同，其表象更像是货币，与金融需求相似最大的需求类别（c.70%），白金的价格由供应/需求决定，投资需求仅为总需求量的5%-10%；和3）全球ETF中有超过3.0mnoz白金和钯金，实际上超过地面冶炼库存，可以在任何时间重回市场，但是通常会在价格上升时这样操作。 潜在影响有两重：1）如果价格保持弱势，不会有显著资金流入全球ETF，白金将保持盈余状态；和2）如果价格保持低迷，白金市场可能看到类似于黄金的明显外流。 ZAR疲软是疲软美元Pt价格的关键原因；以下持续影响价格

世界上的白金大约有80%都源自南非。因此，白金生产的边际成本在ZAR设定，因此疲软

的ZAR造成了更加疲软的美元价格。从2010年后期造成疲软美元白金价格的原因在于，兰特贬值，从7.5:1贬值到11.7:1。

进入2015年，我们的经济学家预期ZAR将在强势美元的情况下，继续面临逆风，我们认为这将导致美元白金价格的下跌。

客户库存 – 价格的下行风险；珠宝一个价格上限机制

我们认为鉴于过去几年的价格波动，客户已经准备了库存，大部分的供应仍处于危险之中。供应的风险（例如罢工，电力中断）都已不是新鲜的事情，客户例如催化剂制造商和汽车生产商，都已经建立了库存。深陷美元价格为客户建立库存以及预防由天气原因造成的中断，创造了机会。这也意味着高于以往的地面存储将形成一个与较高白金和钯金价格对抗的缓冲区。尽管估计地面以上库存的范围在1- 2.5Moz，在包括回收商库存的情况下，很难对其进行估算。

从2010年开始珠宝需求增加了c.600koz，我们预期将仍保持强劲的势头。 但是，因为其对价格高度敏感，我们没有将其视为浮动因子。 因此，我们将其看做一个价格规管机制而不是一个价格设定机制（我们看到价格下跌时购买量的增加，构成一个地板，当价格上升时购买量的减少，构成一个天花板）。

白金：萧条的汽车增长抵消了欧6的积极影响；疲软的ZAR限制了潜在的上行

我们的汽车团队预期，2015年西欧汽车销售将有温和的增长，为1.5%。但是他们预期在2015-18增长趋势将由温和的复苏，欧洲汽车销售将平均为2007年峰值的18%以下。

欧6是白金的积极增量，但是实施成本很高(每辆车c.US$400-600) ，可以看到汽车公司在小排量汽油车辆上投入更多资源，这有利于白金。

鉴于未来几年，在2004-07年制造的高白金含量的车将被回收，在2014-18年，回收将为总供给增加c.500k oz (c.9%)。

我们看到市场在2015年由小规模不足，从2016年开始有结构性盈余。投资需求从2014年显著下降，从2013年的c.850koz下降至2014年的仅超过100koz。考虑到2015年低迷的价格环境，我们预期有有线的流入。随着市场处于进本冗余和贬值的兰特，我们预期白金的美元价格将在2015年下降至￥1,250/oz。因此，我们预期疲软的ZAR弱化名义通货膨胀的影响；因此，我们预期在2016/17，平均白金价格相应为$1,350/oz和1,425/oz。 可能因为行业活动或电力中断产生的生产停滞引起价格的上行风险。最近的油价暴跌，同样可能有利于白金，主要因为消费者购买新车，打车可能引起车辆需求的增加。

钯金：持续处于短缺，主要受强劲汽油需求；原油价格下跌和俄罗斯供应上行风险的影响 钯金是2014年仅有的几种没有大幅叠加的贵金属之一。2014年钯金美元价格升值c.13% 。这是因为两个因素的组合引起的：（1）南非的罢工；和（2）乌克兰的地缘政治紧张（俄罗斯供应了世界钯金供应的40%以上，鉴于涉及乌克兰情况，周围的供应安全受到关注）。 结构性的，我们认为钯金将持续处于短缺状态，因此商品价格将继续保持上涨，因为： 鉴于最近石油价格的下跌，美国汽车销售有所上升 – 这也是主要的汽油市场。根据十一月份的汽车销售数据，五种销售最佳商品中的四种都是皮卡/SUV – 对这些车辆的需求为9.3%， vs 汽车销量增加了1.3%。这些车车Pd载量较高，因此近期对于该金属属于正面支撑。 新车辆的最大增长将来自中国和印度，我们的全球汽车团队预期到2018年将增加1060万量车。这些市场主要为汽油市场，钯金较受青睐。我们特别强调，鉴于油价的降低可能会刺激需求，价格有上行的风险。

在新兴市场对排放标准的要求收紧，应追赶欧洲和美国的标准，应随着时间推移，查看每辆车中钯金用量的增加。

俄罗斯/乌克兰紧张的地缘政治都通增加了钯金价格的上行压力。应注意诺里尔斯克供应了

世界上c.45%的钯金，对该区域的任何潜在威胁都会对钯金价格造成显著的上行风险。此外，媒体报告（2014年5月16日路透社）已经声明俄罗斯的贵金属和宝石仓库，俄罗斯国库，正在计划从俄罗斯生产商购买钯金， 据报道Norilsk Nickel最有可能是卖家。

在过去的十年中，俄罗斯的库存销售有实质性下降。同时源自其它生产商Stillwater 和Norilsk的供应，已经被限制，因此，我们预期将由此产生小规模的上行风险。在柴油车辆中使用钯金替代白金的技术一直在研发当中。

在欧洲针对重型柴油车（HDD）立法的变更，很可能是对高钯金载量的支持，因为在美国2007年已经存在该法案。

考虑到汽油市场增加的需求以及基础短缺，我们认为钯金市场处于上升阶段，我们预期价格会有持续的上升。因此，我们将钯金价格从2015年的$863/oz上调至2016年和2017年的$988/oz 和$1,050/oz。

铁矿石: 一场痛苦的消耗战

回到2004年，铁矿石的售价为$35/t CFR中国（2015$条款），澳大利亚和巴西仅靠几个知名的生产商供应海运市场。

随着接下来数十年的高价格和重大投资，现在生产商数量超过20个。

2014年，是一个转型时期，铁价格下跌50%降至$68/t，但是市场平衡调整远未结束，市场将面临持续的过度供应。对需求低迷，供应增长强劲的市场进行平衡，将在成本曲线的顶端导致持久的战争，持续时间应超过2016.海运市场供应商的数量将在该时期萎缩。 成本通货紧缩 + 过剩产能 = 进一步下跌

中国成本曲线早于预期的下跌和成本通货紧缩日益增加的证据，使得我们将2015-18年的价格预期调整至$66/61/60/60。在中短期，相对于成本价格将下持续下行，以在2015-17年刺激矿关闭以及项目延迟投产。这种情况将在因为商品生产国货币贬值、输入成本下降和生产率提高引起边际生产下降时显现。根据之前对过去30多年间，不同商品供应曲线的分析，我们认为在一个持续期间，现货生产边际成本将从其现在的$75/t水平下降。但是，截止2017，价格必须保持在边际成本以下，以刺激冗余生产能力的关闭。 我们已经下调了我们的长期价格预期至$60/t（现货）。截止目前显著的过度投资将确保市场得到充裕的供应，而且中国钢铁行业的需求也日趋成熟，我们认为有必要刺激至少十年以外的下一轮大型绿地项目。

在2015-17期间，价格必须下滑

在每年海运供应的大约6%，在2015-17期间，每年需要进行调整的规模，需要得到上一年度过度投资规模和中国需求突然减速的验证。.配置边际铁矿石容量的过程将由两个基本市场动态确定。首先，相对于边际成本价格必须下行，以刺激矿的关闭。管理和维护一个亏损的矿，包括任何管理团队都试图避免的重大成本，范围包括供应商合同违约和雇员的辞退，以及矿自身的期权价值损失。为了推行减产，每吨的运行损失必须与公司的资产负债表的要求相平衡，和/或消弱信心，最终达到收支相抵。煤炭市场最近显示了这种动态：在2014年价格相对于我们的边际成本预估大致下降了20%，通常在几个连续的运营损失季度之后，矿山出现关闭。

第二，海运和中国铁矿石市场分离。直至最近，中国较高成本的矿石作为边际供应商，海运价格向中国国内价格靠拢。但是，这种关系已经被破坏。

中国铁矿石矿提供的缓冲，都是高成本和价格较为敏感的，对价格缺乏支撑，形成于2014年第2季度，市场出现供过于求时，现在该缓冲被大幅抵消。从我们的观点，未来中国在平

衡铁矿石市场的角色应被消弱，因此海运价格将由边际海运生产商之间的竞争确定。

冶金用煤：均值回归在起作用

商品周期平均持续25至30年，随着时间的推移，市场动态变化将返回至其初始状态，考虑到相对于人类记忆的长时间框架，一些变化将被认为是不可逆转的，直至其最终被否认。我们主张在冶金用煤均值回归的三个实例在发生作用：1）在一个通货紧缩的环境中，价格回归至边际成本，2）钢需求在中国回归至持续水平和3）美国出口商在价值曲线的顶部的市场份额恢复。

价格将最终在较低的边际成本水平集合 随着采矿生产力开始提高，生产成本开始下降。同时边际容量的逐渐位移（从2013年开始，关闭的矿等同于现行海运供应的10%），将导致行业成本曲线的逐步扁平化；在未来扩展这些趋势，我们预计在我们的预计期间，优质硬焦煤的边际生产成本将下降至$130/t。与此同时，相对于边际成本，煤炭价格已经下行。在2015-16中国国内市场持续供应过剩的情况将仍然存在，中国和其他供应区域的许多生产商都处于亏损运营的状态；价格的下降在成本通货紧缩之前。但是，及时成本在下降，价格也应向边际成本靠拢。

在此基础上，我们将优质硬焦煤2015/16/17的预期下调至$116/125/130/t FOB澳大利亚。我们同样下调了我们的长期预期至$130/t（下降18%），以对持续通货紧缩和诱导定价的延迟回归进行反应。

印度新兴市场是抵消中国成熟市场的一个亮点 如同铁矿石一样，冶金用煤容易受到日渐成熟中国钢铁市场的影响。中国经济在过去十年中，通过增加财富以及每年从农村向城市迁移2000万人口，已经发生了转变。从2004年开始人均创造的GDP已经翻番，城镇化率已经达到了罗马尼西亚的水平。在我们看来，中国钢铁消耗已经增加至一个不可持续的水平，并将最终下滑。中国因其作为边际消费者的角色，在海运价格中发挥了重要影响进口量是高度价格敏感的。但是，冶金用煤也受到印度日益增加要求的影响。印度在人口数量和钢存储方面，从其他新兴市场脱颖而出。在人均基础上，钢储量在印度经济（人均0.7吨）中远远落后于中国（5.6吨）和美国等发达国家（13吨）。换言之，就钢消费量而言，印度落后中国大约25年，钢原材料的长期需求前景是积极的。 美国煤炭将发挥一个更为温和的角色

就供应而言，生产成本在所有主要供应区域都在快速回落，美国除外，在这里生产商受强劲货币及更多生产率提高潜在因素的制约。此外，因为运费降低的原因海运市场的竞争在加剧。就我们的观点，低价格和其他供应区域竞争的加剧将持续边缘化美国生产商；从2013年开始，美国有44%的生产能力被宣布关闭，在2014年远高于其在海运市场19%到的股份。 随着时间的推移，其市场份额将从2011年的峰值28%下降至其牛市之前的水平11%。因此，美国的冶金用煤前景受到挑战。

从我们的观点，煤矿将继续关闭；在一些情况中，公司重组后，一旦其成本基础和资产负债表被重新设定，将促成一些亏损的矿恢复生产。

热能煤：达到退役周期

如同一名工人在庆祝其65岁生日之后，可以过上一种更为稳重的生活方式，回顾她过去取得的成就，我们认为热能煤已经达到了其退役周期。煤炭是电厂单一的最大的能源来源，而且它仍然是很多新兴市场及长期运行电厂燃料的选择，因此在燃料混合重的变更尽可能逐渐发生。虽然如此，热能煤需求和价格的黄金岁月显然已经消逝；我们认为也不可能再恢复。

管理和技术更新破坏了电厂对煤炭的需求，在2020年后达到煤矿消费峰值的前景是真实的。在之前商品周期的经验告诉我们，热能煤行业的备用容量将最终被耗尽，为另外一种商品周期的迭代和另外一个牛市让步，但是相反我们认为，如果能更有效的利用现有资产和低成本扩张，足以满足需求，直到峰值，则市场不需要新建矿和基础设施带来投资的激增。因此，在可预见的未来，在一个成本通货紧缩达到高潮的时期，热能煤价格将追随边际生产成本，我们相应降低我们的价格预期。

随着管理风险增加，下个十年将会不同 就天气变化的国际谈判进展缓慢，但是新煤电厂的管理风险一直在上升。电力公司在考虑如何在能源组合方面分配他们的资本支出，因此面临着一个难题：在当前市场条件下，煤电厂通常有利可图，但是前景更严格的法规会增加煤炭的风险。我们相信，至少，监管风险是因为许多国家鼓励多元化的混合燃料。与此同时，经过多年的经济的能源密集型行业的过度投资，中国的需求增长急剧放缓，我们相信对热能煤的需求将继续增长，至少到2020年，但增长继续温和的步伐，直到全球及海运需求达到最终峰值，除非火力发电和碳捕获超过预期，存在竞争成本的优势。

过去的投资足以满足未来的需求

在供应方面,我们相信,在最近一段时期的投资过剩生产能力的背景下，热煤产量将继续扩大。如同在80年代和90年代，尽管价格下跌，产出一直在增长，主要是因为前期投资的结果。未来的供应增长将受高运行效率和沿供应链潜在能力的支撑。例如，澳大利亚煤矿的采矿生产力存在一个失去的十年，期间，每个员工的产出下降了46%，每单位产出的资本存量下降了67%。根据我们的估算，在2014-14总生产率提高了6%。这些趋势可能在其他供应地区被复制，过去的经验表明，生产率的增长可以持续很长时间。结果，尽管就业总人数的减少和资本支出大幅放缓，生产量可以增加。低价格不太可能妨碍生产。事实上，商品生产商有额外的增加供应的动力。例如，印度尼西亚已经放弃了限制煤炭出口的政策，作为支撑价格的一种方式，且近期对国内消费的前景进行了消减。与此同时，Glencore在2014年停止了澳大利亚运行点三周的生产，以减轻市场上供过于求的情况；虽然如此，纽卡斯尔指数仍然下跌。

碳酸钾：在恢复常态的道路上绕行

在2014年，碳酸钾的价格超出了我们的预期。生产能力过剩的商品市场通常在边际成本或边际成本之下交易，但是在Canpotex/BPC双头垄断市场被打破两年后，钾肥仍不是正常的大宗商品。其他具有类似市场基本情况的矿业商品发生明显的价格下跌；反而，钾肥价格下跌结束过早和盈利能力在行业仍然有吸引力。虽然如此，鉴于钾肥行业变的更加各自为政，大型制造商的定价能力进一步受到侵蚀，价格在中短期将会疲软。

无可否认的，一些生产商开始以更加竞争力的方式运转。例如，Belaruskali首次希望进入美国市场，与Uralkali一起，采用铁矿石等低成本生产商采用的策略满负荷生产。但是，Canpotex成员采取了相对稳定的产能利用率，之前双头垄断时享有的一些定价能力仍然存在，根据显示，价格回升至2014年后半年的水平。生产商的这一良性结果等到中国，印度和巴西等市场需求显著上升的进一步支持。随着Agrium?s Vanscoy矿的暂时关闭，以及随后Uralkali?sSolikamsk 2矿的意外洪水和潜在随时，市场基本情况进一步改善。 随着生产能力超过需求，定价能力进一步被侵蚀

在我们看来，钾肥价格最近的走强是不可持续的。进口数量对该时期至2014年中期钾肥价格的下跌做出了回应，但是这种对于需求的暂时繁荣将很快被耗尽。此外，因为2014/15在美国和巴西等主要产区，农作物到的前景是积极的，存销比会进一步升高，因此，农民经济

不会支持价格高于废料需求趋势。因此，我们预期2015年的进口量会在2014年的合同数量上增加1Mt。从供应方来讲，Vanscoy重新开始生产，Uralkali在其投资组合中，通过增加产量补偿Solikamsk的损失。与此同时，在加拿大和其他区域的新产能调试，将在2018年增加11Mt全球产能，远远超过预计的增长率，降低了平均利用率。 我们的基础观点保持不变：鉴于独立生产商持有的全球产能的比例增加，旧的双头垄断的定价能力将受到进一步侵蚀，钾肥将越来越像是一个正常的商品。从双头垄断到完全竞争市场是渐进的，会绕一些弯路，但是我们相信最终的目的地市毫无疑问的。因此，我们将2015年的预期留给市场（升值11%），但是保留我们2016年的预测不变。我们同样下调我们的长期估值9%，达到$325/t CFR以反映以下内容：

在采矿行业，成本通货紧缩是根深蒂固的。运行成本将随着货币贬值（特别是俄罗斯和白俄罗斯卢布），运费的下降和提高采矿生产力的长期前景，加上其他采矿商品而下降。新项目的资本成本对应于现行的c. US$1，000/tpa将被降低。

尽管我们的观点是基于长期需求的建设性观点，通过高价格刺激新产能的需求可能早已过时。及时在加拿大的Jansen等大型项目停止运转的情况下，至少在2020年前，全球产能利用率也只能达到80%多一点。

矿砂：困于流沙之中

有人不幸落入流沙会感到不舒服，但是，与电影中通常描述的流沙相反，他不会下沉或被淹没。受害人的头部能够维持于水面之上，但是移动变的非常困难；需要耐心和毅力才能逃脱。在我们看来，这个比喻对矿砂市场进行了恰当的描述。 需求下降和供应下降保持平衡

2015年矿山市场的主要驱动因素与去年保持不变。对于瓷砖和油漆等成品的需求，受到消费者努力降低使用密度和中国建筑行业减弱的影响，生产商利润维持在压力以下，但是通过减少闲置的过剩产能，可以放置价格的进一步下滑。

矿山行业的利润已经走完了完整的周期；在2010年低合同价格下甚至不能保证无损经营的矿，在2012年EBIT利润上升至66%，今年仅回落了5%。 我们对中短期价格预期进行下调

这些市场动态，持续的需求下降被最大生产商的供应管理所抵消，可以在可预见的未来，在对定价产生有限影响的继续，但前提在于大生产商持续在量的基础上对价格进行支撑的策略。在我们看来，如果需求弱于预期或者源自第2梯队的竞争者打破了平衡，该策略就会出现问题。如果不出现这样的干扰，我们假定价格已经找到了平衡水准，以及更广泛采矿业的通货紧缩在可预见的未来将影响矿砂。在该基础上我们下调我们对于2015-17期间的价格预期。 我们同样更新了对锆和钛原料的长期价格预期，如同在其他矿石商品采用的方法一致。特备是，我们假设：

对于大部分产品，现行价格接近于边际生产成本的水平。

成本通货紧缩将通过货币贬值、输入成本下降（承包商利率，能源等）以及改善的生产率影响矿砂行业。

考虑到持续的需求下降和现存的生产能力过剩，需要通过高价格引导绿地项目 我们修订后的长期价格预期反映了最初一段时间的平均边际成本定价，紧随之后的是诱因定价的节段。在缺乏充分的数据绘制诱因成本曲线时，我们使用了最近项目的平均资本密度，以评估未来的开发成本。 在相对的基础上，锆石和氯级原料的前景比硫酸级原料更有吸引力。部分原因是更高层次的产业集中在这些细分市场，以及绿地项目的供应增长降低。但是这些产品的供应链是互相连

接的，一个细分市场的价格下降可能轻易的，通过直接竞争（例如，氯化物vs硫酸盐）或通过打扰龙头企业的供应减少，扰乱存在的平衡状态。这种可能产生的情况，会促使闲置产能重新复工，替代高成本竞争对手的价格，意味着我们价格预期的下跌。

Base, Bulks and Bullion

Mining?s ?new normal? – China slower, costs lower

China?s ?new normal? and cost deflation are key themes for 2015

China?s transition continues, costs are deflating rapidly

We see two major headwinds to metals and mining commodities pricing over the short to medium term: China?s ?new normal?; and cost deflation. A moderate acceleration in developed-world demand growth is not expected to offset the impact of these developments on US dollar pricing. Cost deflation Part 1: Lower energy/other commodity prices Cost deflation is resulting from lower energy prices (oil, coal, gas), as well as lower steel, copper, cement, and other commodity prices. Cost deflation Part 2: US dollar strength is just beginning Cost deflation is also the result of US dollar strength, as commodities are priced in dollars. The dollar is appreciating against the currencies of most commodity-producing countries, reducing commodity production costs in dollar terms.

Prefer palladium, nickel, and zinc; bearish copper and gold While our relative preferences across mining commodities are largely unchanged, we downgrade most of our short-term and long-term price forecasts to reflect both realized and future anticipated cost deflation.

In particular, we remain bearish on copper for 2015 despite the price declining more than 20% over the past year, as we not only view the latest move lower as fundamentally driven (sharply lower costs, weaker demand growth), we expect the fundamentals to weaken further over the coming 12 months. Indeed, our new 12-month copper forecast is $5,200/t (-c.9% from current prices), and we see risks to this forecast as skewed to the downside.

Overall, on our new forecasts, the mining commodities with the greatest downside on a 12-month view are copper (-c.9%), gold (-c.8%) and iron ore (-c.6%). Conversely, our 12-month views are most bullish for palladium (+24%), nickel (+22%) and zinc (+16%).

Consistent with our views, because steel is made from iron ore and coking coal, it has the highest exposure to reduced energy capital expenditure (c.10% of steel demand). Nickel by contrast (via the stainless steel market) has the highest exposure to accelerating demand for consumer goods (30%) on the back of the decline in oil and gasoline prices.

Contents

Mining commodities: We lower our forecasts due to actual and anticipated cost deflation, softer demand growth 3 Key themes for 2015: Cost deflation, supply differentiation, and China?s ?new normal? 6

Copper operating cost deflation 9

Copper: The scene is set for the next boom 11 Copper: 2015 to be another tough year 12

Aluminium: We are neutral over the short-term, our medium-term outlook is more bullish 14 Nickel: Very limited downside, significant upside 16

Zinc: Closing in on market tightening, relatively bullish 2015/16 17 Gold: Europe and Japan stall increasing bearish gold outlook 18

PGMs: Restructuring remains key as demand disappoints; Cautious on platinum, palladium remains our favoured exposure 24

Iron ore: A painful war of attrition awaits 32 Metallurgical coal: Mean reversion at play 34 Thermal coal: Reaching retirement age 36 Potash: A detour on the way to normality 38 Mineral sands: Trapped in quicksand 41 Disclosure Appendix

Mining commodities: We lower our forecasts due to actual and anticipated cost deflation, softer demand growth

We update our price forecasts for the mining commodities (Exhibit 1, changes are shaded purple, prior forecasts are in brackets), which form the basis of our equity analysts? forecasts for the mining companies they cover. Our forecasts generally decrease by c.10%- 20% across our forecast horizon for the base, bulks and bullion commoditie

The primary reason for the changes to our forecasts is cost deflation – driven by a combination of actual and anticipated US dollar strength, cheaper energy and other input costs, and our expectation of an improvement in mining productivity. Our analysis of previous cycles indicated that productivity growth can be sustained over a long period of time, thus the efficiency loss during the recent bull market is likely to reverse over the coming decade (see Exhibits 2 and 3). As a result, industry cost curves will shift downward and become flatter. In addition, weaker-than-expected demand from China and Europe has also contributed to the deflationary environment by increasing the amount of marginal production capacity and pushing further into the future the need for new capacity.

These market dynamics have occurred before (in the 1980s and 1990s), but the pace of deflation in 2014 has been notable, with simultaneous sharp moves in oil prices and FX. However, we believe a few outliers are likely to buck the trend of falling prices.

Commodities that were out of favor in the last bull market and which, consequently, were starved of capital will now be out of phase with major commodities such as iron ore and copper, and as a result may see better fundamentals over our forecast period.

The rationale for our lower bulk commodities price forecasts is laid out in more detail Rock & Ores: A long war of attrition awaits the iron ore industry, Heat Sensor: Thermal coal reaches retirement age, and Rocks & Ores: Mean reversion at play in metallurgical coal (both published today).

Among the base metals, while we have been bearish on copper for well over a year and prices

have declined by more than 20% over the period, our latest cost analysis suggests further downside from current prices in 2015; which is our new base case. Meanwhile, owing to supply-cycle differentiation, we expect nickel, zinc and palladium to substantially outperform copper in 2015. As such, we reiterate our short copper / long nickel trade recommendation, which we initiated November 20, 2014 (+6%). With the exception of nickel, for which our new detailed cost analysis supports our prior $17,000/t LT forecast, we downgrade our base metals LT prices by 10%.

Base metals short-term forecasts

On a 3-month view, we are bullish on nickel (our new forecast is c.8% above current spot) and modestly bearish on copper (-c.4%). On a 12-month view, we are bullish on nickel (+c.22%) and zinc (+c.16%), modestly bullish on aluminium (+c.2%), and bearish on copper (-c.9%).

When evaluating our preferences for 2015 across the base metals, among other things we take into consideration the following:

prospects for supply and demand balances,

relative exposure to energy and other input cost deflation, relative exposure to US dollar cost deflation,

relative exposures to consumer discretionary spending, and relative exposure to energy sector capital expenditures.

Copper fares worst on most of these measures and, as a result, we decrease our 3-, 6- and 12-month forecasts to $5,500/t (from $6,600/t), $5,500/t ($6,200/t), $5,200/t ($6,000/t) (Exhibit 2). We forecast a surplus of supply relative to demand, with a once-in-20-year supply cycle set against heavy exposure to the soft China property market and a weak European demand outlook. Indeed, we expect continued demand weakness in China and an anticipated rise in visible inventories on the SHFE and LME in 1Q15 to continue to weigh on prices over the very near term. Into 2Q15 the extent to which copper might rebound we believe will be determined by the amount of buying by the Chinese government in 1Q15 and the strength of the seasonal upswing/underlying demand growth.

More broadly, copper is heavily exposed to US dollar appreciation as most of the marginal copper producers are outside of China and the US, in countries where we have seen significant currency depreciation (median depreciation of 17% for copper producers with costs>$2.50/lb since 2013). We expect further depreciation over the next 12-24 months.

Copper is also relatively highly exposed to consumables and diesel cost deflation. See the section titled “Copper cost deflation, a case study – copper”. As such, we see copper marginal cost support, at the 95th and 90th percentiles of the copper cost curve, falling to $5,600/t and $5,000/t, respectively, in 2015/16. Risks to these estimates are to the downside in our view.

Finally, in terms of the potential for stronger consumer discretionary demand as a result of the decline oil prices, only 7% of copper is consumed by consumer-discretionary goods vs c.30% for stainless steel (nickel).

Meanwhile, our short copper / long nickel recommendation (initiated November 20, 2014) is currently up by c.6% and we target a 20% return. Indeed, nickel comes out on top on most of these measures, with the exception of exposure to weak energy sector capex (c.3% of stainless steel is consumed in the energy sector). We forecast nickel to be broadly in balance to a small deficit in 2015. Nickel?s marginal costs are predominantly US dollar- or RMB-based, being comprised of

Chinese power grid tariffs (we believe NPI producer power grid tariffs are likely to be largely unchanged in 2015) and US dollar-based nickel ore prices.

Further, the Indonesian ore ban suggests that nickel ore costs will continue to rise as ore stocks are depleted in 2015, supporting higher prices.

For aluminium, we have shifted our view to broadly neutral in 2015 (although we continue to expect strong ex-China producer margins following the ex-China markets? move into deficit in 2014). We expect zinc prices to be relatively strong in 2015 on the back of the anticipated closure of the Century mine in Australia, falling gold, silver and copper byproduct credits, and smelter capacity constraints in China.

Key themes for 2015: Cost deflation, supply differentiation, and China?s ?new normal?

We believe oil, iron ore, and copper, as well as a number of other mining commodities have moved from the investment phase to the exploitation phase during which there is typically a push to raise productivity in the face of weaker prices, which reduces costs and tends to be pro-cyclical. At the same time, we believe we are seeing the start of a US dollar bull market, which we believe will be helped by the ongoing shale revolution and lower energy prices. The continuing strength of the dollar and lower energy prices globally are also powerful forces for cost deflation, which are the inverse of the drivers of the mining commodities bull market in the 2000s, and we see this as a bearish backdrop to mining commodity pricing over the next 12 months. This is because markets in balance or in surplus tend to trade around or below the marginal cost of production.

Exhibit 3 illustrates that these dynamics are not unrelated, and to some degree are self perpetuating.

Specifically, a relatively strong US economy, domestically driven and supported by lower energy prices, means a stronger dollar, less accommodative US monetary policy, weaker EM demand growth, lower commodity producer currencies and mining commodity production costs (cost deflation), and potentially higher supply (than otherwise would be the case) from the commodity producing countries.

Against this backdrop of falling costs and weaker EM demand growth as a result of the export of higher US interest rates and a movement of capital from EM back to the US, we have China reforming its economic growth engine, from commodity intensive FAI driven growth to less commodity intensive consumption and services driven growth.

Necessitating this shift has been China?s strong growth in credit to GDP (particularly in the corporate sector) since 2009. Indeed, the corruption crackdown and slowing in credit (as measured by total social financing) since 2013 has resulted in a slowing in Chinese economic growth and mining commodity demand. Most impacted so far has been steel, and we believe copper is to come in 2015/16, with late cycle copper intensive construction completions growth to slow sharply over the period.

Within this context we continue to see supply as key to differentiating commodity performance, as was the case in 2014.

Copper operating cost deflation

Costs of production, and in particular marginal production costs, are critically important for commodity markets in balance or in small surplus, as these markets tend to see their prices trade with or around the marginal cost of production. Indeed, with copper transitioning from a tight market over the past decade to a period of moderate surplus, prices have already fallen and trade in to the top end of the copper cost curve (Exhibit 13).

We estimate that marginal costs, as proxied by the 95th percentile of the copper cost curve, are set to decline by at least c.20%, from $7,000/t to $5,600/t in 2015/16 (Exhibit 13). We expect the 90th percentile to decline to around $5,000/t. In periods of moderate surplus, prices tend to trade in to the 90th-95th percentile, supporting our new forecasts for copper in that range.

These cost forecasts assume a 20% depreciation of high cost (>$2.50/lb total cash costs in 2013) copper producers currencies (of which 17% has happened over the past year, Exhibit 14), an 80% pass through, and a 50% decline in diesel costs, and no declines in power, consumables, services, or labour costs. If other costs did decline, or if the surplus is larger than we anticipate, copper will likely trade below $5,000/t in the next 12-18 months.

We can split the cost deflation discussion in to two parts, US dollar strength related cost deflation, and direct copper mining cost deflation (energy, labour, services etc.):

1) Producer currency depreciation vs. the US dollar (c.80% of operating costs are in local currency terms (Exhibit 15)). Exhibit 14 illustrates recent and forecast future currency movements for mines which had costs above $2.50/lb in 2013 (the high cost top 20% of mines).

High cost copper producer currencies are down c.17% since 2013, and we expect a further 4% depreciation over the next 12-24 months. US and Chinese high cost producers are rising sharply up the cost curve since they have not received currency relief (total high cost output only 180kt).

Indeed, we find strong empirical evidence for a negative relationship between marginal copper costs of production and marginal producer currencies vs. US$ (Exhibit 16). For more details on copper costs please see our November 17, 2014 report, Stronger dollar, weaker oil, softer copper.

2) Copper mining cost deflation (N.B. mining cost changes across the various cost components shown in Exhibit 17, tend to be strongly positively correlated, as in Exhibit 18):

a) Lower energy prices (8% of direct costs). Oil prices have more than halved over the past 6 months, and historically diesel costs have an almost 1:1 relationship with oil prices.

b) Lower consumables prices (32% of copper costs). Consumables include steel balls and explosives (an oil derivative), the costs of which are likely to decline with falling steel and oil prices. Further, producers of explosives and steel balls are likely to be facing margin erosion.

c) Lower services prices (25% of direct costs). Service providers to the mining industry are likely to see their margins decline, which is an effective reduction in service costs to miners, and overheads may be cut in the face of weaker prices.

d) Potentially lower power prices (10% of direct costs). Power prices may decline, although this is less likely as they tend to be sticky.

e) Potentially lower labor costs via an increase in productivity (25% of costs). In the face of weaker prices restructuring to reduce costs and overheads

f) Offsetting this to a minor degree in the short term are slightly declining copper ore grades (on

average), and higher financing costs.

Copper: The scene is set for the next boom

We note an important and increasing wedge between operating costs and new project incentive prices, which indicates that we have already entered a period of underinvestment in copper mines. Indeed, while it is too early to see the capex slowdown in the data, anecdotally there was already a major slowdown in capital investment in Chile in 2014.

More specifically, the wedge between project incentive prices and operating costs is driven by capital costs being relative sticky, owing in part to deteriorating project quality (more difficult location, projects are increasing in countries with higher geopolitical risks – i.e. the the ?low hanging fruit? has been mined). With capital costs a major consideration in miners decisions to invest in copper assets, and a major component of ?incentive pricing?, incentive prices are set to rise relative to operating costs and prices (incentive prices could fall 10%, while opex and prices could fall 20%-30%). Indeed, our new LT copper price forecast is $2.70/lb, which compares to the current copper price of c.$2.56/lb and our 12-month copper price forecast of$2.36/lb.

In this environment, with miners and investors focused on spot pricing and the flat forward curve in copper, investment in the next phase of copper projects is likely to be put on hold.

Over the longer term it is our belief that this, together with declining industry ore grades, will result in a major slowdown in supply growth, which along with trend demand growth should see a major tightening in our base case by 2017/18.

Having said this, we believe it is too early to turn bullish on copper from current prices, even taking a medium-term view, given the risks to Chinese and European demand, and given the risks to the cost curve are, in our view, skewed to the downside.

Copper: 2015 to be another tough year

With copper prices declining by more than 45% over the past four years it remains tempting to call a turn in the market. However our research indicates that both the recent 10%-ytd decline and the broader price weakness has been fundamentally driven (demand weakness, supply strength, rising visible inventories, and sharply declining costs prospects), and that these bearish fundamentals will see copper prices decline for the 5th consecutive year through 2015. We forecast a further 5% decline in copper prices over the next 3-6 months, to $5,500/t, and a 10% decline in copper prices over the next 12 months, to $5,200/t. Having said this, 2016 looks more balanced, and 2017 and 2018 should see copper tightening, although the exact timing of and level from which copper prices turn is uncertain, with risks skewed to the downside owing to global demand downside risks.

We discuss the primary drivers and talking points surrounding this view, below:

Our research suggests that copper is heavily exposed to Chinese property (Exhibits 22-23), and we anticipate a major slowing in late cycle copper intensive construction completions growth in 2015 and 2016, following 2 years of weak new starts data and some payback from ?over? completion relative to prior new starts in 2014. Indeed, falling steel apparent demand does not bode well for later cycle copper demand.

Unlike during the boom period, there are limited upside risks to Chinese demand growth owing

to already high FAI as a % of GDP and highly leveraged Chinese corporate balance sheets.

European physical copper premiums have weakened to 2008/09 crisis lows over the past six months, reflecting weaker than anticipated demand in Europe and/or softer than expected Chinese demand for European exports (Exhibit 21).

China state grid expenditure cannot fill the gap in 2015 and 2016. We now forecast 4% China demand growth in 2015, and 8% growth in state grid copper consumption.

Copper is half way through a once-in-20-year supply cycle (which started in 2H12 and is set to last through to the end of 2016), following a decade of high capex investment in the industry, raising trend supply growth to c.4-5% from c.2% over the past decade.

Our research – using 100 years of scrap data – suggests that despite a major decline in prices, copper scrap availability will not decline and may in fact grow over the course of 2015 and 2016. This is because prices remain high in real terms, infrastructure to collect scrap has been built out with costs sunk over the past decade so operating collection costs are low, the old scrap pool continues to grow, and global demand growth points to growth in new (runaround) scrap.

Marginal costs of copper production are falling sharply, and we expected them to continue to decline over the next 2-3 years, on a strengthening dollar against the producer currencies, energy, and other commodity price deflation, as well as service industry margin erosion and potentially as an increase in labor productivity in the face of weaker prices and a broader strong dollar environment (see section ?Copper operating cost deflation? above). The transfer of this cost decline from producers to consumers will occur in our view, due to weak market fundamentals. Where costs settle over the next 1-2 years is very difficult to pin down, however, and this will be a source of considerable debate in the coming months.

Our “bottom up” supply and demand balance work continues to point to a c.500kt surplus market in 2015, and a market broadly in balance in 2016. We continue to assume 200kt of SRB buying in 2015, after 500kt of purchases in 2014.

The longer term outlook for copper is very bullish. We are not structural copper bears, and we agree with the longer term bull case for copper. As such, despite declining costs of production we see prices rising in 2017 and 2018.

Aluminium: We are neutral over the short-term, our medium-term outlook is more bullish

After a 7 year bear market, gradual aluminium production cuts outside of China eventually resulted in the market moving in to deficit in 2014. Ex-China prices, where supply was most disciplined, markets saw prices rise sharply (>30% to their 2014 peak) and margins expand to their highest levels since 2007. However, demand growth expectations have recently moderated, and Chinese supply growth has surprised to the upside, resulting in both China and ex-China prices moderating over the past two months. China prices have fallen to around their prior lows (China in surplus, albeit expanding), while ex-China prices are still trading at a premium to costs of production (ex-China still in deficit, albeit it less so).

For 2015 we continue to see strong margins ex-China, and a weak China market. We do not see all in pricing in aluminium improve until 2016, when a tightening in the China balance following from a 25% reduction in FAI in aluminium smelting in 2014 (owing to low prices and tight credit) should flow through to a slowing in output growth, a slowing in semifabricated export growth, and

a rise in prices (out of the cost curve in China).

While we see the global market broadly in balance to small surplus in 2015, there are pockets of tightness that have been very persistent throughout 2014. The most notable has been the North American market where closures of capacity were significant and where demand is growing relatively strongly.

Overall ex-China prices should need to remain above China prices for at least 2015 in order to: a) attract metal from China in the form of semi-fabricated products and b) incentivize production restarts outside of China. With the profitability of exporting semi-fabricated products now c.25%, we have met criterion a), but at c. $2,250/t all in ex-China we have yet to reach the price level that would result in sufficient capacity restarts ex-China in our base case. The main downside risk to this view would be major Russian or Brazilian capacity restarts following their recent currency depreciation. A further downside risk is that China supply continues to surprise on the upside, following from greater than anticipated government support (via lowering tariffs, smelters getting finance to build off grid power generating capacity, etc.).

With respect to premiums, the LME is reportedly (Bloomberg, January 21) reviewing rent rules for end users in queues. Our forecasts assume flat premiums in 2015, but a decline in premiums of around $50/t per annum over the remainder of the forecast horizon, in line with our work in “The economics of a warehouse exchange”, published in October 2013.

Nickel: Very limited downside, significant upside

We believe that current nickel prices represent a strong buying opportunity, particularly relative to copper. Although we have lowered our forecasts in the near term on a softening in global demand expectations - our 3- and 6-month forecasts are now $16,000/t and $17,000/t, from $17,000/t and $18,000/t, respectively, our 12-mo target of $18,000/t remains unchanged, and we see significant upside from current spot prices, set against limited downside risk emanating from strong cost support (see Exhibits 40-41).

The nickel ore market continues to gradually tighten, increasingly squeezing China nickel pig iron producers (NPI) margins and beginning to force meaningful production cuts in China. Looking forward, we believe the nickel ore market will only get tighter before Philippine ore supply is resumed around mid-April, and even post the resumption of Philippine exports ore stocks are set to draw in to year end. The tight fundamentals in the nickel ore market in the next 3-4 months are set to push nickel ore spot prices higher (they are already up by c.40% yoy), raising nickel pig iron output costs. Furthermore, a return to an LME nickel premium to costs presents an upside risk to our near term price forecasts.

On a 6-12 month view, the continuous implementation of Indonesia?s nickel ore export ban should keep the nickel ore market tight, even after Philippine is back from the monsoon season, since we believe 1) the nickel ore inventory in China has been falling to a very low level and 2) it is very unlikely to expect Philippine nickel ore supply to increase as strongly as in 2014 to fill the gap left by Indonesia. Finally, while we leave our long term nickel price forecast unchanged, we have increased conviction in it following a substantial amount of work done in conjunction with our equities research colleagues. Please see their GS Nickel 40: The consumer age and supply chain disruption to drive nickel price higher, for details.

Zinc: Closing in on market tightening, relatively bullish 2015/16

Overall we have not made significant changes to our zinc outlook. On the one hand demand prospects have weakened, but on the other hand we draw closer to the closure of the century mine (>2% global supply), slated for 3Q15 due to ore grade depletion, so overall we remain moderately bullish on zinc. Indeed, the combination of solid global growth in zinc demand and a significant slowing in mine supply growth in 2015 and 2016 should see zinc move into deficit over the next 12-24 months.

While we do not see any near term catalysts for a rally, we see limited downside risks to zinc prices from current levels owing to strong zinc cost support. This cost support view owes to two factors:

i) high cost mine supply is located in China, and

ii) 70% of zinc mine revenue comes from by products including lead, silver, copper and gold (all of which have either seen major price declines recently, or we expect major price declines over the next 12 months).

If and when zinc does tighten up, the response of Chinese mines and smelters will be key in the initial instance. We believe there is potential for a zinc smelting bottleneck owing to a lack of investment in new capacity and strict enforcement of increasingly stringent environmental controls. Our 12-month forecast target for zinc remains $2,500/t.

Gold: Europe and Japan stall increasing bearish gold outlook

While gold prices have trended lower since mid-2013, the decline has been well short of our expectations. Recently, the combined support of: (1) weaker-than-expected US economic data; (2) the run-up to the announcement of QE in Europe; and (3) the surprise SNB decision to remove the CHFEUR cap, have seen prices rise to $1,300/toz.

While we continue to expect that gold prices will decline in 2015-16, we are nonetheless revising our near-term forecast path higher due to these recent catalysts although we believe that these are likely mostly priced in. Subsequently, we expect the decline in gold prices to resume from 3Q15, in line with the start of the US rate hiking cycle. Importantly, while our near-term conviction for lower gold prices has declined, our confidence in lower gold prices for longer has increased on the back of ower expected inflation in the coming years and the deflationary impact on gold?s mining marginal cost of production of lower energy prices, a stronger USD and the shift to the Exploitation phase of the commodity supply cycle.

While we had expected lower gold prices throughout 2014 on our economists? forecast for accelerating US economic growth, this acceleration occurred in the face of slowing European, Japanese and global growth. This left gold prices mostly range bound, consistent with the range-bound price action of US nominal and real rates with both markets still exhibiting a strong inverse correlation throughout 2014. This left the rise in US rates as relative vs. other G10 rates, against our and consensus expectations for higher rate levels. Importantly, while exchange rates broadly reflect real rate differentials, ourmodeling of gold prices shows that USD-denominated gold prices respond to the level of rates instead (see Forecasting gold as a commodity, March 25, 2009). As a result, USDdenominated gold prices remained mostly range bound while the USD strengthened significantly.

Despite the lack of a more material decline in USD-denominated gold prices, we continue to expect that gold prices will decline in 2015-16, given our economists? expectation that the shift to above-trend growth in the US will continue, with easing financial and lending conditions and lower oil prices helping growth in the rest of DM sequentially improve. In particular, while the acceleration in US growth has so far mostly translated into only a relative increase in US real rates vs. the rest of G10 rates, our rates strategists expect that the beginning of the US rate tightening cycle to finally lift US real rates in absolute terms.

Under our framework for gold prices, this will translate into declining USD gold prices, although the combination of QE in Europe and our US economists? expectations for a later rise in the Fed funds rate than the consensus will likely limit the near-term downside to gold prices.

We believe the recent catalysts for higher gold prices are likely to support gold prices near current levels and, as a result, we raise our 3-month gold price forecast to $1,295/toz.

US growth: Recent economic data has disappointed relative to expectations, with declines in ISM manufacturing and non-manufacturing, retail sales and some housingrelated data. Importantly, this sequential slowdown is occurring as our US economists? forecast real GDP growth to slow to 3% relative to the above-4% rate achieved in 4Q14.

Ultimately, they expect growth to remain solidly above trend, with risks tilted to the upside owing to the beneficial net impact of lower oil prices and the headwind from a stronger USD notwithstanding. Their baseline expectation remains for a September 2015 lift-off in the funds rate, with risks tilted to the later side should inflation continue to disappoint.

ECB QE: The ECB announced today an expansion of its asset purchase program to include sovereign debt, with the program seen as aggressive relative to prior expectations, in terms of both the size and pace of purchases (see European Views:

ECB announces inflation-path-dependent purchase programme, January 22, 2015).

Under our gold framework, Europe?s QE program supports EUR-denominated gold prices as it exerts downward pressure on European rates and also leads to EUR depreciation. The impact on USD gold prices is more limited however, as this depends on the extent to which US rates are dragged lower by European rates. Gold prices rallied on the day although the long run-up to the announcement and yesterday?s leak had left the anticipated impact largely priced in. This is consistent with what our work showed in the US, with the impact of QE priced into gold prices upon announcement (or prior when it is widely anticipated like with US QE3). Importantly, our worked showed that it is the QE stock announcement that mattered most for gold, not the actual flow (see Gold cycle set to turn on improving US recovery, December 5, 2012).

While this would imply that the gold price reaction has already largely occurred, the ECB?s statement kept the door open to expanding the program further if warranted by the inflation path. As a result, further increases in the program size, should inflation and inflation expectations fail to rise, would likely provide additional support to gold prices. Importantly though, this potential support would likely take time to materialize, with current gold prices probably already reflecting the ECB QE program for now.

SNB: The impact of the SNB shock is less clear on gold prices: while negative deposit rates on CHF bank accounts creates an incentive to shift domestic savings into other safe-haven assets, gold?s historically higher price volatility (up until this week), low correlation to the CHF, and own

negative cost of carry will likely limit the rotation into gold. Second, with the EURCHF cap removed, the Swiss franc has again become a safe-haven asset for foreigners and can compete with gold for safe-haven inflows, although following this week?s move, the CHF currently screens even more overvalued relative to our economists? GSDEER fair value estimate. Net, our bias is that the rotation away from CHF and into gold remains limited, although our conviction is low.

Net, absent a reversal in the US and global recovery, we expect only limited further upside to gold prices despite the recent European and Swiss monetary shifts, as these are likely already largely priced in. Nonetheless, we also see limited near-term catalysts for gold prices to retrace their recent gains. As a result, we raise our 3-month gold price forecast to $1,290/toz, in line with the current forward curve.

As we move forward through 2015 however, we believe US gold prices will resume their decline given our economists? expectation that the shift to above-trend growth in the US will continue, with easing financial and lending conditions and lower oil prices helping growth in the rest of DM sequentially improve. The key view imbedded in this forecast is that, relative to 2014, the increase in US rates actually occurs in level terms, with the beginning of the US rate tightening cycle as the catalyst starting from 3Q15. Our updated 6- and 12-month USD-denominated gold price forecasts are $1,270/toz and $1,175/toz. Ouryear-end 2015 forecast is $1,190/toz and $1,000/toz for year-end 2016.

Beyond this rate view, our forecasts reflect our expectations that: (1) ETF gold holdings will resume their decline once US rates increase (with that relationship holding well over the past three years); and (2) central bank (ex-China) gold purchases continue although at the more modest pace observed recently. A slowdown in Russia?s strong gold purchases would create downside risk to our gold forecast.

A repeat of 2014 with a further slowdown in Europe, Japan and continued low inflation would likely create upside risk to our updated gold price forecasts. Nonetheless, we continue to believe that a material slowdown in global economic growth would be required to push the gold price sustainably higher from current levels.

Importantly, while our near-term conviction for lower gold prices has declined, our confidence in lower gold prices for longer has increased on the back of the lower expected inflation in coming years and the deflationary impact on gold?s mining marginal cost of production of lower energy prices, a stronger USD and the shift to the Exploitation phase of commodity supply cycle. As a result, we are lowering our 2016 and beyond USD denominated gold price forecast to $1,050/toz from $1,200/toz previously。

FX and energy cost deflation are relieving cost pressure

Similar to other mining commodities, the key developments for gold that have helped to drive cost deflation over the last six months are lower energy prices and FX depreciation among producer countries (excluding the US).

Energy inputs directly account for around 10%-20% of mine cash costs while the price of crude oil is down by almost 50% and Newcastle coal down by over 25% in 2014.

Owing to ample storage capacity and the need to keep investment in shale sidelined, we forecast the oil price to remain near current levels for most of 2015. While we subsequently expect rising

prices, our expectations for long-term prices have also been revised lower. Accordingly, the majority of the decline in energy input costs should prove persistent.

Wages and other local currency denominated costs make up a further 30%-40% of mine cash costs. We estimate that, on average, producer countries exchange rates vs. the US$ have declined by 12% over 2014. Based on GS forecasts1, we expect a further 7% depreciation over 2015 and 12% cumulative depreciation over 2016, as interest rate differentials with the US$ are expected to widen further.

Combining these FX and energy price dynamics with our mine-by-mine cost model, and using the 90th percentile of each cost curve as a proxy for the marginal cost of production, we calculate that marginal cash costs will decline to around $850/toz, from $1,000/toz in 2013 (a 15% reduction). Our preferred measure of “All-In” marginal costs2 looks likely to shift down to around $1,050/toz, from around $1,200/toz in 2013 (a 12.5% reduction).

Further cost reduction likely in the ?Exploitation? phase of the commodity cycle

As we have previously outlined (see Investor Returns will survive the productivity comeback, April 24 2014), the ?Exploitation? phase of the commodity supply cycle typically features additional sources of cost deflation, most of which are only beginning to take effect for gold:

There have been signs of lower domestic wage settlements in local currency terms.

The clearest example of this is in South Africa, where 2014 compensation in the mining sector was almost flat year-on-year, down from sustained double-digit growth rates over most of the 2000s.

With employment growth having rationalized since 2013 and output continuing to gradually move higher, there are now some tentative signs of increasing labor productivity. However, there is still some way to go to reach the levels of labor productivity prevailing in the late 1990s.

We expect to see capital productivity increase and greater discipline imposed on new capital expenditures.

While ore grades are expected to continue to decline over the foreseeable future, we do not ignore the cumulative effects of ongoing technological progress. In fact, while ore grades have been steadily falling for over 100 years (-1.4% on average), this has not prevented a continued rise in global gold production (+1.6% on average) far in excess of the rise in the real gold price (now +0.6% on average, but zero or negative for most of history).

PGMs: Restructuring remains key as demand disappoints; Cautious on platinum, palladium remains our favoured exposure

2014 was a disappointing year for platinum despite the five-month strike which took c.40% of global mine capacity off-line. Reasons for the 11% decline in USD platinum prices through 2014 include:

1) Inventory release and above-ground stocks: The Big 3 producers (Anglo Platinum, Impala and Lonmin) stockpiled inventories as they expected an extended period of industrial action, which they liquidated to service customers. This, coupled with end-user inventories, above-ground stocks (which includes inventories at recyclers etc.) kept the market amply supplied during the strike and the ramp-up period.

2) Lower demand: 2014 started with optimistic forecasts for the European auto market. However, autos sales disappointed again as Europe logged sluggish macro growth. Although the recent downturn in oil prices could be beneficial for platinum as this stimulates consumer demand for autos.

3) Rand weakness: The ZAR is a key determinant of Pt prices. Given 80% of mined platinum comes from South Africa, the marginal cost of production is set in ZAR.

Thus, a weaker ZAR implies a weaker USD price and vice versa. ZAR weakened 12% in 2014, which had a considerable impact on USD Pt prices.

4) Lower gold: Platinum prices are highly correlated with gold prices. Strong macro data points emerging from the US throughout 2014 increased the chance of an earlier rate hike by the Fed, which resulted in the gold price remaining muted throughout the year and, in turn, affected platinum prices.

5) No restructuring: The Big 3 producers promised significant steps to take off marginal production during the five-month strike. However, after the strike, the focus shifted towards ramping up swiftly rather than delivering on the restructuring.

Platinum is currently trading close to the marginal cost of production -- we estimate c.$1,250/oz (at spot ZAR 11.7 vs. USD). However, we highlight that this can change very swiftly in tandem with movements in the ZAR. Also, with continuing USD strength, further downside in the ZAR could put more downside pressure on the USD price of platinum.

Our base case for European auto demand is lower-than-trend growth. Our autos equity analysts believe growth in car volumes for Western Europe in 2015 will be only mildly positive at 1.5%. Beyond that, they forecast a modest recovery, albeit at a still-depressed level – they forecast 2015-18 European car sales growth to average 18% below the 2007 peak level. QE by the ECB and a lower oil price could be upside catalysts for auto demand and, by extension, platinum demand in Europe.

Restructuring of the supply-side a likely outcome from sustained low prices Returns for the major miners are below the cost of capital and have been since 2011. Anglo Platinum, Impala and Lonmin all have some form of restructuring option, but have so far been unwilling to be the first-mover on large-scale capacity reductions.

If the demand situation does not improve markedly, we expect one of the major producers to announce mine closures by the end of 2015, which would act as a positive catalyst for prices, in our view. Impala?s strategy review is expected with its 1H15 results in February.

Platinum: We remain cautious despite positives from auto growth and Euro 6

Over the near term, we see some modest upside potential in platinum prices given that, at current levels, the price is below the marginal cost of production. Over the longer term, we retain our cautious outlook on the metal – our view is driven by:

1. Rebound in mined supply and rise in secondary supply: We expect an increase in both primary and secondary supply. Mining is a resilient sector and the only way companies can curb costs is by producing more ounces (which spreads fixed costs over more salable material). As such, we expect a rebound in secondary supply. Also, we see an increase in new production (in South Africa) and also see secondary supply (from recycling) rising as more vehicles from 2002-06 (which had high Pt loadings) are replaced.

2. European auto outlook: Given the weak macro data coming out of Europe, major car companies

have trimmed their outlook for the region (Daimler recently cut its 2015 sales forecast to 3%-4% from 4%-5%). Commentary from other majors – Volkswagen and Ford – has also not been reassuring. Our autos team forecasts tepid growth in Western European car sales for 2015 of 1.5%. As such, we expect autocat demand to remain depressed and we do not see significant upside.

3. Euro 6 – an incremental positive, not a game changer: Consensus we believe is pricing in a strong uplift from Euro 6 implementation in 2015. We believe that it will be an incremental positive and does not present significant upside to demand.

4. Reduced investment demand: Even in 2013, when the platinum market was in a c.1Moz deficit, it was only in a small deficit of 70koz pre-investment. We expect investment demand to plateau as prices remain depressed. Sluggish European auto market takes away positives from Euro 6; recent oil rout could be a potential catalyst

Auto demand is c.40% of total annual demand for platinum and over 60% for palladium. Platinum is significantly more exposed to Europe due to its required use in diesel engines and Europe being c.50% a diesel market. Palladium is more exposed to petrol (gasoline) markets, such as the US, China and Japan.

Going into 2015, our autos equity analysts expect Western European sales to continue to remain sluggish – growing at c.1.5%, with zero growth expected in Germany, the UK and Spain, the most important markets for platinum. We expect Europe auto demand to increase c.2% to 20.4 mn units by 2015. On Pt loadings, we expect a small spike (as has been the case historically when regulations change) to 2.25g/car, which translates into an additional demand of just under 100koz – only 1.2% of total demand.

The recent downturn in oil prices could, however, spur vehicle demand. Oil is down by more than 50% over the past six months. Lower oil prices tend to spur vehicle demand particularly – heavier vehicles have higher Pt loadings. Although we expect the autocat demand to be fulfilled easily over the short term it could be positive sentiment-wise for platinum.

Mine supply to rebound as mines ramp-up and new mines come online; recycling increasing too During the strikes, there was widespread anticipation of restructuring and closure of marginal/loss-making mines. However, when the strikes ended, focus shifted to swift rampup and talk of restructuring faded as companies sought to avoid further confrontations with unions. We expect mine capacity to rebound in 2015 as ramp-up is completed. Also, the new mines coming online – PTM?s WBJV, Northam?sBooysendal and RBPlats? Styldrift will add c.400koz by 2017 which should see supply rebound to 2011 levels. Longer term, we see a significant increase in platinum supply as newer projects, such as PTM?s Waterberg and Ivanhoe Mines? Flatreef project (both are currently in PFS stage), reach commercial production.

Further, we see an increasing contribution from recycling. Recycling has increased by over 70% over the past 10 years and we expect a further 500koz (c.9% of total demand) to be added as cars manufactured between 2003-07 (a time of peak PGMs per car), reach the end of their useable lives and are recycled.

Platinum in a deficit -- albeit an investment-induced one; pre-investment it remains in surplus; ETF outflow a downside risk to price

Excluding investment demand, the platinum market has been in surplus and we expect a much smaller surplus as investment demand plateaus. Investment demand picked up significantly in 2013 on account of new the South African ETF. However, 2014 was a disappointing year as prices failed to appreciate when the strikes ended. As such, total ETF additions in 2014 were just over

100koz.

Some key points on ETFs: 1) ETFs are non-yielding, with price appreciation the only source of return; 2) unlike gold, which behaves more like a currency, with financial demand the largest demand category (c.70%), the price of platinum is determined by the supply/demand, and investment demand is only 5%-10% of total demand; and 3) the over 3.0 mnoz of platinum and palladium that are in the global ETFs are, in effect, aboveground refined inventory that can come back to the market at any time, but would do so typically when prices rise.

The potential implications are twofold: 1) if prices stay weak and there is no significant inflow into the global ETFs, platinum would remain in surplus; and 2) if prices remain depressed there is a chance the platinum ETF market could see a significant outflow similar to the one that was seen in gold.

ZAR weakness a key reason for weak USD Pt prices; to continue to affect prices Almost 80% of world?s platinum comes from South Africa. As such, the marginal cost of production for platinum is set in ZAR. Therefore, a weaker ZAR allows for an even weaker

USD price. One of the factors that has seen a weak USD platinum price since late 2010 has been the weakening rand which has moved from 7.5:1 to close to 11.7:1.

Going into 2015, our economists expect the ZAR to continue to face headwinds in the form of a strengthening USD, which we believe will lead to a lower USD platinum price.

Customer stocks – a downside risk to price; jewellery a price-capping mechanism We believe customers have been building stocks as prices have fluctuated over the past few years and the majority of supply remains at risk. The risks to supply (e.g. strikes, power disruptions) are not new and customers, such as the catalyst fabricators and auto producers, have been building stocks. Dips in USD prices will have created opportunities for customers to build stocks and weather potential supply disruptions. The implication of this is that higher-than-previous above-ground stocks could act as a buffer to higher prices for both platinum and palladium. Although estimates of above-ground stocks range from 1- 2.5Moz, it is very difficult to measure them given it includes inventories at recyclers etc.

Jewellery demand has increased by c.600koz from 2010 and we expect it to remain strong. However, we do not see this demand as a swing factor given it is highly sensitive to prices.

As such, we see it as more of a price-capping mechanism than a price-setting one (as we see increased buying when prices fall, providing a floor, and decreased buying when prices increase, providing a ceiling).

Platinum: Sluggish auto growth offsets positive from Euro 6; ZAR weakness limits potential upside

1.5%.

However, they forecast a modest recovery to trend growth over 2015-18, with European car sales will be in on average 18% below the 2007 peak level.

(c.US$400-600 per car) and this may see car companies put more resources into small-capacity petrol (gasoline) cars, which is more beneficial for palladium.

-18 as cars with high

platinum loadings built between 2004-07 are recycled in the years ahead.

it in 2015 and in a structural surplus from 2016. Investment

demand dropped significantly in 2014 down from c.850koz in 2013 to just over 100koz in 2014. Given that we forecast a subdued price environment in 2015, we expect limited inflows.

With the market being in a fundamental surplus and a weaker rand, we expect the USD price for platinum to decline in 2015 to US$1,250/oz. Thereafter, we expect a weakening ZAR to soften the impact of nominal inflation; as such, we expect the metal to average US$1,350/oz and US$1,425oz in 2016/17, respectively.

Upside risks to prices could come from production stoppages due to industrial action or power disruptions. The recent rout in oil prices could also be beneficial for platinum as consumers buy new, bigger cars which would lead to increased autocat demand.

Palladium: Markets in sustained deficit, driven by strong gasoline demand; Oil price decline, Russian supply upside risks

Palladium is one of the few precious metals that escaped a big sell-off in 2014. Palladium appreciated c.13% in 2014 in USD terms. This was due to a combination of two factors: (1) the strikes in South Africa; and (2) the geopolitical tensions in Ukraine (Russia is the source of more than 40% of the world?s palladium supply and concerns surrounding security of supply have grown as the Ukraine situation has evolved).

Structurally, we believe palladium will be in a sustained period of deficit and, as such, remain bullish on the commodity because:

Given the recent decline in oil prices there has been an increase in US auto sales – which is primarily a gasoline market. According to November auto sales data, four of the five top sellers were pickup trucks/SUVs – demand for these vehicles was 9.3% vs. a 1.3% increase in car sales. These cars have higher Pd loadings and as such this is a near-term positive for the metal.

The majority of growth in new vehicles will come from China and India which our global autos team expects to add 10.6 mn vehicles by 2018. These markets are predominantly gasoline, favouring palladium. We highlight that there is an upside risk to this number given the lower oil prices which could spur demand.

Emission regulation is tightening in emerging markets and catching up with European and US standards, which should see palladium per vehicle increasing over time.

The geopolitical tension in Russia/Ukraine also adds to upside pressure on palladium prices. Note that Norilsk is the source of c.45% of the world?s palladium and any potential threat to this source would present a significant upside risk to our palladium prices. In addition, media reports (Reuters, May 16, 2014) have stated that the Russian precious metals and gems repository, Gokhran, is looking to buy palladium from Russian producers – with Norilsk Nickel reported to be the likely seller.

Russian stock sales have come down materially over last 10 years. Also, the supply from other producers – Stillwater and Norilsk – is capped and as such we expect little upside risk from there. Substitution of platinum in diesel vehicles by palladium continues as technology advances continue.

Heavy-duty diesel (HDD) legislation changes in Europe are most likely to be supportive of higher palladium loadings as they were in the US in 2007.

Given the increasing demand from gasoline markets coupled with the fundamental deficit the we believe the palladium market is in, we expect a sustained increase in the prices. As such, we forecast palladium to average US$863/oz in 2015 rising to US$988/oz and US$1,050/oz in 2016 and 2017 respectively.

Iron ore: A painful war of attrition awaits

Back in 2004 when iron ore was selling for US$35/t CFR China (in 2015 $ terms) Australia and Brazil counted only a handful of established producers supplying the seaborne market. Following decade of high prices and significant investment, that number now exceeds 20.

The year 2014 was a period of transition when iron prices dropped 50% to US$68/t but the adjustment required to balance the market is far from complete and the market is facing a sustained period of oversupply. Balancing the market in the face of slowing demand and strong supply growth will result in war of attrition at the top end of the cost curve that should continue beyond 2016. The number of seaborne producers is bound to shrink over that period. Cost deflation + excess capacity = further downside

A sooner than expected decoupling from the Chinese cost curve and growing evidence of cost deflation lead us to downgrade our price forecasts to US$66/61/60/60 over the period 2015-18. In the short to medium term, prices will continue to underperform relative to costs in order to incentivize mine closures and project delays over the period 2015-17. This will unfold while marginal production costs are falling due to weaker commodity currencies, lower input costs and rising efficiency. Following on previous analysis3 of supply trends across different commodities over the past 30+ years, we believe that the marginal cost of production will decline in real terms over a sustained period from its current level of US$75/t. However, prices must remain below marginal costs over the period to 2017 in order to incentivize the closure of surplus capacity.

We have also downgraded our long term price forecast to US$60/t (real). Significant overinvestment to date will ensure that the market is well supplied while demand from the Chinese steel sector is maturing, and we believe the need to incentivize the next round of large greenfield projects is at least a decade away. Prices must undershoot during 2015-17

At approximately 6% of annual seaborne supply, the scale of the adjustment required each year during the period 2015-17 is testament to both the scale of overinvestment in previous years and the sudden deceleration in Chinese demand. The process to displace marginal iron ore capacity will be shaped by two basic market dynamics. First, prices must undershoot relative to marginal costs in order to incentivize mine closures. Placing a lossmaking mine on care and maintenance involves significant costs that any management team would try to postpone, ranging from the default on supplier contracts and the layoff of employees to the loss of option value on the mine itself. In order to bring forward the decision to cut production, the operating loss per tonne must be such that it drains the balance sheet of the company and/or undermines the confidence in an eventual return to a break-even position. Coal markets have recently illustrated this dynamic: in 2014 prices undershot by approximately 20% relative to our estimates of marginal cost, and mine closures often came after several consecutive quarters of operating losses.

Second, the seaborne and Chinese iron ore markets have decoupled. Until recently, high cost mines in China acted as the marginal supplier and seaborne prices converged towards the Chinese price for domestic concentrate. However, this relationship has broken down.

The buffer provided by Chinese iron ore mines that are both high cost and price sensitive provided scant support to prices when the market fell into oversupply during Q2 2014, and now that buffer has largely been displaced. In our view, China should play a less prominent role in balancing the

iron ore market in the future and seaborne prices will therefore be determined by competition among marginal seaborne producers.

Metallurgical coal: Mean reversion at play

Commodity cycles last 25 to 30 years on average, and as the wheel turns, market dynamics change only to revert back to their original state. Given their long timeframe relative to human memory, some changes can seem irreversible until they are eventually proven otherwise. We argue that three instances of mean reversion are at play in metallurgical coal: 1) price reversion to marginal cost in a deflationary environment, 2) steel demand reversion back to sustainable levels in China and 3) market share reversion for US exporters at the top of the cost curve. Prices will eventually converge towards a lower marginal cost level

Production costs are falling as mining productivity starts to improve. Meanwhile, the gradual displacement of marginal capacity (mine closures since 2013 are equivalent to 10% of current seaborne supply) is gradually flattening the industry cost curve; extending these trends into the future, we estimate that marginal production costs for premium hard coking coal could drop to US$130/t by the end of our forecast period. At the moment, coal prices have undershot relative to marginal cost. The persistent oversupply in the Chinese domestic market is likely to continue during 2015-16 and many producers in China and other supply regions are operating at a loss; the decline in prices has run ahead of cost deflation. However, prices should revert towards marginal cost even as costs are dropping.

On that basis we downgrade our premium hard coking coal forecasts for 2015/16/17 to US$116/125/130/t FOB Australia. We also downgrade our long term forecast to US$130/t (down 18%) to reflect the combined impact of sustained deflation and a delayed return to inducement pricing.

India emerges as a bright spot to offset a mature Chinese market

Like iron ore, metallurgical coal is exposed to a maturing Chinese steel market. The Chinese economy has been transformed in the past decade by rising wealth and the annual migration of approximately 20 million people from rural to urban areas. GDP per capita has more than doubled since 2004 while the urbanization rate now matches Romania?s. In our view, Chinese steel consumption has increased to unsustainable levels and is bound to decline eventually. China retains considerable influence over seaborne prices given its role as the marginal consumer; import volumes are highly price-sensitive. However, metallurgical coal is also exposed to growing demand in India. India stands out among other emerging markets both in terms of its population and its steel stock. On a per capita basis, the stock of steel in the Indian economy (0.7 tonnes per capita) lags well behind China (5.6 tonnes) and developed markets like the US (13 tonnes). In other words, India is 25 years behind China in terms of steel consumption, and the long term demand outlook for steel raw materials is positive. US coal will play a more modest role

On the supply side, production costs are falling rapidly in all major supply regions except for the US, where producers are hindered by a strong currency and by more limited potential for productivity gains. Moreover, competition is increasing in the seaborne market because the world is shrinking as a result of cheaper freight. In our view, the combination of low prices and increasing competition from other supply regions will continue to marginalize US producers; the

US accounts for 44% of production capacity closures announced since 2013, well above its 19% share of the seaborne market in 2014.

Over time, their market share is likely to revert from its 2011 peak of 28% back towards its pre-bull market level of 11%. The outlook for US metallurgical coal is therefore challenging.

In our view, mine closures will continue; in some cases, corporate restructuring may enable some loss-making mines to resume production once their cost base and balance sheet has been reset.

Thermal coal: Reaching retirement age

Just like a worker celebrating her 65th birthday can settle into a more sedate lifestyle while she looks back on her past achievements, we argue that thermal coal has reached its retirement age. Coal is the single largest source of energy for the power sector, and it remains the fuel of choice in many emerging markets and power plants have long operating lives so changes in the fuel mix can only happen gradually. Nonetheless, the golden years for thermal coal demand and prices are clearly behind; we argue they are unlikely to return.

Regulation and technological innovation are undermining coal demand in the power sector, and the prospect of peak coal consumption post 2020 is real. Past experience in previous commodity cycles would suggest that spare capacity in the thermal coal industry will eventually be exhausted, giving way to another iteration of the commodity cycle and another bull market but we argue instead that the market will not require another investment surge in greenfield mines and infrastructure if exploiting existing assets more efficiently and low cost expansions are sufficient to satisfy demand until it peaks. As a result, thermal coal prices are likely to track marginal production costs for the foreseeable future at a time when cost deflation is in full swing, and we downgrade our price forecasts accordingly.

The next decade will be different as regulatory risks increase

Progress in international negotiations on climate change has been slow, but regulatory risks for new coal plants are increasing. Power utilities considering how to allocate their capital spending across their energy portfolio are therefore faced with a conundrum: coal plants are often profitable under current market conditions, but the prospect of tighter regulations increases the risk profile of coal. We believe that, at a minimum, regulatory risks are encouraging greater diversification in the fuel mix of many countries. Meanwhile, demand growth in China has slowed down sharply after years of overinvestment in energy-intensive sectors of the economy. We believe demand for thermal coal will continue to grow at least until 2020 but the pace of growth will continue to moderate until both global and seaborne demand eventually peak unless coal-fired generation with carbon capture exceeds expectations and becomes available at a competitive cost. Past investment is enough to satisfy future demand

On the supply side, we believe that thermal coal output will continue to expand on the back of the recent period of overinvestment in production capacity. Just like output growth during the 1980s and 90s was possible in spite of falling prices because of previous investment, future supply growth will be supported by higher operational efficiency and by latent capacity along the supply chain. For example, mining productivity in the Australian coal sector has emerged from a lost decade during which output per employee declined by 46% and output per unit of capital stock declined by 67%. According to our estimates, total productivity has improved by 6% over the period 2013-14. These trends are likely to be replicated in other supply regions, and past

experience shows that productivity growth can be sustained over long periods. As a result, output can grow in spite of a reduction in total employment and a sharp slowdown in capital expenditure. Low prices are unlikely to hold production back. In fact, commodity producers have an additional incentive to grow supply. For example, Indonesia has discarded its policy of restricting coal exports as a way to support prices, and the forecast for domestic consumption was recently cut. Meanwhile, Glencore stopped production for three weeks at its Australian operations late in 2014 in the hope of alleviating the oversupply in the market; the Newcastle index fell nonetheless.

Potash: A detour on the way to normality

Potash prices exceeded our expectations in 2014. Commodity markets with excess production capacity usually trade at, or below, the level of marginal cost but potash still is not a normal commodity two years after the break-up of the Canpotex/BPC duopoly. Other mining commodities with similar market fundamentals saw prices undershoot significantly; instead, the decline in potash prices ended prematurely and profitability in the industry remains attractive. Nonetheless, prices are likely to soften in the medium to long term as the potash sector becomes more fragmented and the pricing power of large producers erodes further.

Admittedly, some producers appear to behave in a more competitive manner. For example, Belaruskali is looking to enter the US market for the first time, and together with Uralkali it is now producing at full capacity in line with the strategy adopted by low cost producers in markets such as iron ore. However, Canpotex members have chosen to keep capacity utilization rates relatively stable and some of the pricing power previously enjoyed by the duopoly still remains, as indicated by the rebound in prices during 2H 2014. This benign outcome for producers was supported further by a significant increase in demand from markets such as China, India and Brazil. Market fundamentals improved further following a temporary closure at Agrium?s Vanscoy mine, followed later in the year by the accidental flooding and potential loss of Uralkali?sSolikamsk 2 mine.

Pricing power to erode further as production capacity runs ahead of demand The recent strength in potash prices is not sustainable in our view. Import volumes responded to falling potash prices in the period to mid-2014, but this temporary boost to demand should wear off. Moreover, farmer economics are not supportive of above-trend demand for fertilizer because the 2014/15 outlook for crops is positive in key producing regions such as the US and Brazil, and the stock-to-use ratio is bound to increase. We therefore expect 2015 import volumes to contract modestly by approximately 1Mt from 2014. On the supply side, Vanscoy is back in production while Uralkali can compensate for the loss of Solikamsk by increasing production elsewhere in its portfolio. Meanwhile, the commissioning of new capacity in Canada and other regions is on track to add 11Mt in global production capacity by 2018, outpacing the expected growth in demand and pushing average utilisation lower.

Our fundamental view is unchanged: as the share of global production capacity held by independent producers increases, the pricing power of the old duopoly will be eroded further and potash will increasingly behave like a normal commodity. The transition from duopoly to fully competitive market is gradual and can take several detours, but we believe the ultimate destination is not in doubt. We therefore mark our 2015 forecast to market (up 11%) but leave our 2016 forecast unchanged. We also downgrade our long term forecast by 9% to US$325/t CFR Brazil to

reflect the following:

Cost deflation is well entrenched in the mining sector. Operating costs should decline with weaker currencies (in particular the Russian and Belarusian Ruble), cheaper freight and the prospect of a long period of improving mining productivity, in line with other mining commodities. The capital intensity of new projects should also decrease relative to current levels of c. US$1,000/tpa.

Despite our constructive view on long term demand, the need to incentivize new capacity via high prices may be more than a decade away. Even in the absence of large projects such as Jansen in Canada, global capacity utilization is likely to remain in the low 80%s at least until 2020.

Mineral sands: Trapped in quicksand

Someone unlucky enough to fall into quicksand will feel rather uncomfortable but, contrary to the usual depiction of quicksand in movies, he is unlikely to sink and drown. Instead, the victim is able to keep his head above water but movement becomes difficult; patience and perseverance are required to escape. In our view, this analogy is appropriate to describe the mineral sands market. Demand destruction and supply discipline in balance

The key drivers of the mineral sands market in 2015 are essentially unchanged from last year. Demand for finished products such as ceramic tiles and paint is undermined by consumers? efforts to reduce intensity of use continue and by weakness in the Chinese construction sector. Producer margins remain under pressure, but a highly concentrated industry has been able to prevent a steeper price decline by idling excess capacity.

Profitability in the mineral sands sector has gone full circle; a mine that was barely breaking even under low contract prices in 2010 saw EBIT margins rise to 66% in 2012, only to fall back to ~5% this year.

We downgrade our short and long term price forecasts

These market dynamics, with ongoing demand destruction offset by supply management by the largest producers, may continue for the foreseeable future with limited impact on pricing provided that the large producers continue their strategy to support prices at the expense of volume. In our view, such a strategy could come into question if weaker than expected demand or stronger supply from Tier 2 competitors was to upset the balance. In the absence of such disruption, we assume that prices have found an equilibrium level and that the deflationary forces in the broader mining sector will impact mineral sands for the foreseeable future. On that basis we downgrade our price forecasts for the period 2015-17.

We also update our long term price forecasts for zircon and titanium feedstocks, in line with the methodology adopted in other mining commodities. In particular, we assume that: Current prices are close to the level of marginal production costs for most products.

Cost deflation will impact the mineral sands sector via weaker currencies, lower input costs (contractor rates, energy, etc.) and improved productivity.

Given ongoing demand destruction and existing overcapacity, the need to induce greenfield projects via higher prices is well past 2020

Our revised long term price forecasts reflect the average over an initial period of marginal cost pricing, followed by a period of inducement pricing. In the absence of adequate data to develop an inducement cost curve, we have used the average capital intensity of recent projects to estimate

future development costs.

On a relative basis, the outlook for zircon and chloride grade feedstocks is more attractive than that for sulphate grade feedstocks. Partly because of the higher level of industry concentration in those market segments and because of lower supply growth from greenfield projects. However, the supply chains of these products are interconnected, and a price decline in one segment could easily upset the existing balance either via direct competition (e.g. chloride versus sulphate) or by upsetting the supply discipline of leading producers. The possibility of such a scenario where idled capacity comes back online and displaces higher cost competitors implies downside risk to our price forecasts.

本文来源：https://www.bwwdw.com/article/tf7r.html