1. Context - like with financial ratios you need to be aware of the context of each mine contributing capacity to the curve. How do these mines differ? Now, you can do things to overcome these obstacles, i.e. Not simply saying that a mine can supply at cost, but you can identify other constraints on these miners and include them in your formula, i.e. They can only supply power stations with certain coal quality specifications. These are really complicated algorithms. It gets even tougher when you have mines producing different blends, and if the fix depends on variability in the ore or product prices. The reality however is that the 'blends' are fixed by contractual undertakings. Another issue is 'destination'. It is one thing to say you have produced a cost curve, but one needs to know to what destination. If to Asia, one needs to consider the characteristics of the port (panamax, capesize), as well as the freight rates that will impact on competitiveness of different routes. The freight market is pretty volatile, so you might wonder how useful these curves are. Of course shippers will fix rate for some sense of certainty. Power stations and steel plants also have an important control over supply because they like 'stable' or certain quality coal or iron ore to ensure predictable furnace geochemistry. But what about those other buyers, like cement plants who are far more flexible in their fuel use? They drive the market in every which way, and they don't care about quality; its the price for 'useful heat content'.
2. Prices - Prices are set at the margin. This is another indicator of how difficult it can one to use cost curves. Consider however that many investors are interested in only the long-run supply cost curve to establish the long run outlook for mines. Of course if they are interested in the long-run, they are interested in the trends.
3. Future proofing - Is it possible to build future assumptions into a cost curve, so that they can become useful indicators. i.e. The curve might become dated if a company announces a new mine, or the closure of one. The problem is less the prospect of a new mine, but the prospect of price changes at the margin causing the deferment, not just of one mine, but 5-10% of the market capacity. What if brownfields capacity is identified in that time. There is a certain utility to such curves, but you mind wonder at just how complex their models would have to be. If you are a senior mining executives, I wonder if you look at the detail.
You might be forgiven for thinking that supply-cost curves are just economic rationalism. The problem is many people just look at the curve....and not the data, and what the data means. Anyway, if you want to explore these types of issues, I suggest looking at the following website - as supply-cost curves are their 'speciality'. Iron ore markets are a little simpler in terms of their dynamics.
Profit from mining with Global Mining Investing eBook Global Mining Investing Community.