March 03, 2015 - 12:22pm EST by
2015 2016
Price: 20.30 EPS 0 0
Shares Out. (in M): 90 P/E 0 0
Market Cap (in $M): 1,827 P/FCF 0 0
Net Debt (in $M): 83 EBIT 0 0
TEV (in $M): 1,910 TEV/EBIT 0 0
Borrow Cost: General Collateral

Sign up for free guest access to view investment idea with a 45 days delay.

  • Aluminum
  • Metal
  • Basic Materials


I am short Century Aluminum. Century is the best pure-play on upstream aluminum production. Other aluminum companies like Alcoa, Kaiser, Constellium, and Norsk Hydro have a bunch of midstream assets (ie processing) or other business lines. Century operates five smelters in the US and one in Europe. Century is the most beta you can find on the price of physical aluminum. I highlighted “physical” for a reason...this is important, Century receives the LME plus the physical premium.


I actually started working on Century as a long. I became curious about aluminum after reading about how automakers and aircraft OEM’s are increasingly substituting aluminum for steel for strength and weight reasons. So I looked at a couple aluminum companies figuring this would be relatively simple to understand. Several conspiracy theories and senate hearings later and I realized the aluminum market was way crazier than I first assumed.


Midwest Premium


Why bury the lead - let’s discuss the Midwest Premium. What is this anyway? When a smelter like Century sells aluminum to a customer, the contract is based on LME plus the local premium. In the case of the US, almost all smelters are located in the midwest (smelters are old and were built near access to cheap coal power) so the premium for the entire US is the Midwest Premium. Other geographies have their own premia subject to their own local market dynamics.


One would imagine the premium would mostly be the cost to store and transport it. Since those costs probably shouldn’t fluctuate much, one might also imagine that the Midwest Premium probably wouldn’t fluctuate much over time. And guess what, you would be right. For several decades, the Midwest Premium hovered around $100/mt. It did this in bull markets, it did this in bear markets, it did this when the price of aluminum was high, it did this when the price of aluminum was low, interest rates, etc. Sure there were some relatively minor movements based on a variety of factors, but generally the Midwest Premium stayed constant. This fundamentally makes sense...again, the premium shouldn’t be much more than the cost to store and transport aluminum.


Then in 2010, the Midwest Premium started rising. Pretty soon, it started rising very fast. Recently, it went vertical, abruptly spiking to record levels. This charge for essentially shipping and handling had strangely increased 400% from $100/mt historically to over $500/mt. Below is a chart of the Midwest Premium stretching back over 15 years for some perspective on how unusual this move was.



When commodities charts look like that, it’s seldom anything beyond a short-term anomaly. Baltic Dry index, rare earths, whatever…and the idea that there is a new paradigm in storing aluminum seems awfully silly. For some further perspective on how unusual this is, the Midwest Premium had historically only been about 5% of the price of the LME aluminum price, but at $500/mt (note that the chart is in pounds and there are 2,200 pounds per metric ton) it was over 30%. So what happened in 2010? Well, before I tell you, let’s talk about how the premium is determined.


Detroit Metro


For all intents and purposes there is just one warehouse in the US that controls most of the physical supply of aluminum: Metro International Trade Service (Metro). Metro is a network of warehouses in Detroit that for whatever reason (I’m guessing inertia combined with the likelihood that metal storage is a crummy business) manages to house most of the US aluminum supply. Metro is registered and regulated by the LME, thus allowing delivery under LME contracts and a variety of other LME regulations. The Midwest Premium is basically set based on the comings and goings of aluminum at Metro.


...Back to the Midwest Premium


Okay, so what happened in 2010? Wouldn’t you know it, Goldman Sachs bought Metro. Of all the gin joints, right? There was nothing crazy going on, they simply saw a good trade opportunity. As the owner of Metro, Goldman got paid from storage fees which are typically pretty small, something like $70/mt. With the aluminum market in contango, they saw a nice opportunity to profit, but they also figured out a way to turbocharge the trade.


The first thing Goldman did was start offering inducements for people to store their aluminum at Metro. This was basically a one-time bonus of a $100-200/mt to put your aluminum in their warehouse. Now that might seem a little strange, I mean, how does a company that charges X per year make any money by paying 2-3X upfront? Not exactly a good payback before operating expenses.


The second thing that Goldman started doing was enforcing LME’s load-in/load-out rules. The gist of these rules is that an LME-registered warehouse only has to load and unload a certain amount of aluminum per day. Oddly, the LME rules were such that the load-in rate was much higher than the load-out rate, and the load-out quota was a shockingly low 2,000mt, which is only a pallet or two’s worth. This means an LME warehouse could legally trap aluminum if they were so inclined. Now the inducements start making sense, because you can afford to pay those amounts upfront if you figure out a way to lengthen the storage life. In fact, this is what started happening. So an actual customer/owner of the aluminum at Metro would need their aluminum for something. Say, Coca-Cola to make cans. Let’s see how the conversation between Coke’s procurement guy and the manager of Metro would go:


Coke: “Hey, can we get 20,000 tons of our aluminum? Marketing is all bulled-up on this Kardashian-Edition Coke we’ve got coming out, and we need to do a giant canning run. You don’t even want to know what the secret ingredient is.”

Metro: “Gee, we’d love to except we just hit our daily load-out quota so we’re done for the day. Why don’t you try tomorrow? Love the Kardashian idea, by the way.”

Coke: “What do you mean? It’s 10am. It’s our aluminum and we kinda need it.”

Metro: “No I hear you, but we’re done for the day, all our drivers went to the MGM Grand.”

Coke: “Seriously, if we don’t get our aluminum Kanye West is going to go postal.”

Metro: “Sorry, take it up with LME. It’s Miller time. Thank you, come again.”


The final thing Goldman started doing was getting cute with some of the many financial owners of aluminum, as in the commodity desks of other investment banks. Those 2,000 tons of aluminum that Metro “hit their quota” with? Half the time it was aluminum owned by another investment bank who agreed to move it from one Metro warehouse to another. Boom, quota met, and no one who actually uses aluminum can get their hands on it. There are stories about Metro truck drivers just aimlessly driving around the Metro facilities with a load of aluminum.


How bad did it get? One way this is measured is the days to destock one’s aluminum, or the load-out queue. Before being under Goldman’s control, Metro averaged about 40 days to destock one’s aluminum. Obviously there are some logistical constraints to unloading the entire US’ aluminum supply, so 40 days sounds about right I’d say. How many days did it take until very recently? 700. Yes, it went from taking just over a month to get your aluminum out, to two years.



So imagine this. Aluminum inventories are near record highs and the LME price of aluminum is blah because we’re still producing too much aluminum globally, but because of these strange circumstance, no one can get their hands on enough aluminum. So what happens? The Midwest Premium goes bonkers because the Coke’s and Ford’s and Boeing’s of the world literally can’t get enough aluminum ingots delivered to their factories despite very high inventories and generally weak LME pricing. These consumers of aluminum were forced to bid up the physical price of aluminum and draw down whatever inventory they could get their hands on. LME was irrelevant, they needed physical.


Something similar happened in Europe as well, and wouldn’t you know it but Vitol, the owner of Europe’s mega aluminum warehouse, managed to copy Goldman’s trade over there. There were a few lawsuits about collusion that ended up getting dismissed. It wasn’t collusion, just a crowded trade.


Why Short Now?


There were two recent catalysts that have fundamentally changed the picture and make Century an interesting short. First, after a major uproar, LME changed the load-in/load-out rules so that they’re balanced. It’s much harder for warehouses to play games that cause a logjam to get aluminum out.


Second, and as I see it this is the more important one, Goldman sold Metro in January. Call me old fashioned, but I believe Goldman is strictly about the money. They didn’t sell Metro because of political pressure or anything else, they sold Metro because the trade was over. And looking at it, what a masterful work of art. They bottom-ticked the buy and top-ticked the sell. I hope someone forwards my writeup to the folks on the commodities team over there because seriously, I am in awe.


Implications for Century


In technical terms, Century has a mega ass-ton of operating leverage to the price of aluminum. Century has capacity of about 700,000mt in the US and another 300,000mt in EU. The EU capacity is currently subject to a tolling agreement so it somewhat distorts the financials (the agreement basically excludes the cost of alumina from both price and cogs, so it overall lowers Century's reported revenue/ton and cogs/ton) but the tolling agreement is rolling off in 2016 and converting to a direct agreement. The cash difference is relatively minor, so my financials will treat the EU capacity as if it’s direct rather than tolled for the sake of simplicity.


Let’s look at the last two quarters. Century has an even 1 million tons of annual capacity. Their cogs/mt is about $2,100 and their opex/mt is about $60. LME averaged about $2,000/mt while the Midwest Premium averaged about $500/mt. During each of the last two quarters, Century has earned about $80-85m in ebitda because of 15-16% gross margins. As a tip-off, the only other time Century has had these kinds of margins was in the 2006-2007 period when China was white-hot and aluminum inventories were at all-time lows.


Century’s fourth quarter call was surprisingly dramatic with respect to the outlook for the Midwest Premium. I would pull two quotes from the CEO:


“On the Midwest premium, I think you heard us say we don't think $0.24 is a good number. We think it's coming down. I mean, let me just be blunt, it is coming down. Anybody who's in these markets knows that the deals that are being transacted have been done – number one, the posting price has come down. Midwest is below $0.24, Europe is falling more as we said, and actual business that's being transacted is happening below those posted prices.”


“Delivery premiums remain strong through the fourth quarter, but have begun to come down over the last couple of weeks as those of you have seen who follow these markets. The EU has been the hardest hit due to the relatively weak demand in that region and building imports principally from Russia. The duty paid premium in Europe is now posted at $435 a metric ton, but we have information and others do that business is being transacted well below that, perhaps in the range of $400 per metric ton, plus or minus $20 to $30. It's a wide range given the lack of liquidity in these markets.”


At current LME spot price, and if the Midwest Premium continues to fall to say $350/mt for the year, I estimate Century will generate an annual ebitda loss of $80m. Keep in mind that since the second half of 2014, LME has fallen $200/mt and the Midwest Premium has fallen $100/mt, and Century has about $90m of ebitda exposure to every $100/mt change in the all-in price of aluminum. Here’s the real problem though. If the Midwest Premium reverts to historical norms and LME stayed the same, I estimate Century would generate an ebitda loss of about $220m. $2,100/mt cash costs (dd&a and opex cancel each other out) and an all-in price of $1,900/mt on 1m in annual volume will do that. At that point, Century will become a restructuring situation. Given the dynamics of the Midwest Premium, it doesn’t strike me as the type of thing to fall a little bit from its all-time high and then level out. This is the kind of thing that goes right back to where it was at for eternity.


Meanwhile, analysts are doing an up-and-to-the-right analysis for a highly cyclical commodity producer and what’s been an unambiguously terrible business for decades. The street is at $375m and $450m in ebitda for 2015 and 2016, respectively. Cyclicals: cheap at the top.


Let’s recap the 30k foot view of the trade:


  1. The Midwest Premium was at $100/mt for forever

  2. Goldman bought Metro and got cute with operations

  3. The Midwest Premium made unprecedented move to $530/mt

  4. Goldman sold Metro and rules were changed

  5. The Midwest Premium goes back to $100/mt????


I can’t figure out why #5 doesn’t happen. If it does, ceteris paribus, Century is in trouble.




I’m not a commodities trader. I’m an equity guy which for better or worse makes me the big picture person on aluminum. I’m sure there are a thousand things about aluminum trading that I’m oblivious to, so be prepared for some comparatively smart person to ask me a question about something and my answer is the online equivalent of a blank stare. Or I'm just plain wrong on something.


The Midwest Premium isn't a liquid market and there are some timing lags, so what's reported on Bloomberg might not reflect reality or be immediately reflected in Century's financials. This is why the stock reacted so poorly to their fourth quarter call, because the management commentary was clearly much more grim than what's been readily available from other pricing sources.


The LME price could go sharply higher even if the Midwest Premium fell. I don’t have a strong, or really any view on this. I accept it as a risk and move on.

Century is a fairly well-run company. Management correctly focuses on operations since they are purely price-takers. They have done a good job lowering their electricity costs which is the largest input to aluminum production. They might be able to extract some further improvements in this area.


The industry will likely reduce capacity if pricing gets bad.


Century is a volatile stock. Be prepared for it to go up 20% in a week for reasons you can't figure out.

I do not hold a position with the issuer such as employment, directorship, or consultancy.
I and/or others I advise do not hold a material investment in the issuer's securities.


Midwest Premium falls

2       show   sort by    
      Back to top