The global scrap market supplies 40% of ferrous material used for steelmaking. Scrap is consolidated up a tier of processors. World prices for scrap are set in the USA and transmitted through dockside trading. British and American scrap markets have become more closely cointegrated over time, with fewer price differences owing to short run seasonal variations.
Scrap prices regularly halve and double. The scrap market is also affected by episodic price 'bubbles' and occasional price crashes. These are rare and transient features of the market. The rise of US minimills and growth in Far Eastern demand have not caused any structural breaks in the behaviour of markets. Statistical tests suggest there is no long run relationship between scrap prices and energy prices.
Recycling works on a global basis. There is scope for closer collaboration between steelmakers, scrap suppliers and manufacturers to promote re-use, remanufacture and recycling through careful initial design and material selection.
Get full access to this article
View all access options for this article.
References
1.
On the relationship between scrap prices and subsequent ore prices see J. Aylen and K. V. Albertson: ‘The world market for scrap and iron’ in ‘The steel industry in the new millennium, Vol. 1, Technology and the market’, (ed. RanieriR.AylenJ.), 61–74; 1998, London, Institute of Materials.
BarrettR.: ‘Profits not size drive US steel scrap today’, Met. Bull. Mon., March2001, 46–47.
4.
SpanosA.: ‘Probability theory and statistical inference: econometric modelling with observational data’, 457; 1999, Cambridge UP.
5.
HartP. E.: ‘The size and growth of firms’, Economica, 1962, 29, (113), 29–39; S. J. Prais: ‘The evolution of giant firms in Britain: a study of concentration in manufacturing industry in Britain 1909–1970’, 1976, Cambridge University Press for National Institute of Economic and Social Research. Testing the log of firm size is normally distributed is complicated. The Jarque–Bera test of normality does not work well on small samples and it is difficult to gather information on the size of smaller scrap firms to increase the sample size. The two largest UK scrap firms have multinational operations and are not statistically comparable with domestic scrap processors. A test on the censored distribution of the 30 largest domestic scrap firms gives a value of JB·1.33 which is distributed x22 with a probability value of 0.51, against an alternative that firm size is not log normally distributed. We can tentatively accept (domestic) firm size in UK scrap processing is consistent with Gibrat’s Law.
6.
Normal profit has a particular meaning in economics. It is the lowest rate of return necessary to keep resources in an industry. In a competitive economy there is an incentive to move resources into industries earning above normal profits and withdraw them from sectors earning below normal profits.
7.
StoneJ. K.: ‘The iron and steel industry and the commodity market’, Iron Steelmaker, February1977, 36–38.
8.
BoultonL.: ‘Trading trash in cyberspace’, Financial Times, 3 April1996, p. 19
9.
Metal Bulletin: ‘Scrap’s Japanese future’, 7 March2005, 6–7.
10.
Monthly auction markets are more volatile than the overall market. Prices for US factory bundle auctions reported by American Metal Market show almost twice the standard deviation of US No. 1 HMS Composite Prices over the months between January2000and March 2005, a representative period characterised by both volatility and stability.
11.
CookeH.: ‘SBB steel futures guide – possible effects of the LME’s proposed contracts on the steel industry and its main raw material suppliers’ athttp://www.steelbb.com(click ‘futures’), last accessed 9 February2006.
12.
HoefferE.: ‘How good is the cheaper scrap?’, Iron Age, March1994, 25.
13.
MillmanS.: ‘Quality steel from the EAF’ in ‘EAF technology: state of the art and its future evolution’, 21–36; 2000, Brussels, International Iron and Steel Institute, Committeeon Technology. For a counter view see J. Eliasson, T. Sizwecki and B. Hutchinson: ‘Processing of copper-containing steel via strip casting – a laboratory evaluation’, Steel Research International, 2006 (forthcoming).
14.
HabrakenL.Lecomte-BeckersJ.: ‘Hot shortness and scaling of copper-containing steels’ in ‘Copper in iron and steel’, (ed. Le MayIainMcDonaldL.Schetky), 45–81; 1982, New York, Wiley.
15.
GruverG.GiarrataniF.: ‘Modelling geographic ferrous scrap markets: regional prices and inter-regional transactions in the United States’, J. Regional Science, 2005, 45, (2), 313–341.
16.
US Bureau of Mines: ‘Metal prices in the United States through 1991’, 1993, Washington D. C., US Department of the Interior, US Bureau of Mines. The importance of looking at ‘mega trends’ – the longest time available time series going back as far as possible – is stressed by S. Makridakis, S. C. Wheelwright and R. J. Hyndman: ‘Forecasting methods and applications’, Chap. 9, 3rd edn, 1998, Chichester, John Wiley.
17.
For the UK case seeWorswickG. D. N.: ‘British raw material controls’, Oxford Economic Papers, April1942, (6), 1–42and P. D. Henderson: ‘Government and industry’, in ‘The British economy in the nineteen-fifties’, (ed. G. D. N. Worswick and P. H. Ady), 330–331, 335–336; 1962, Oxford University Press.
18.
For evidence on a five year cycle see M. Evans and S. B. Walton: ‘Time-series properties and forecasts of crude steel consumption in the UK’, J. Forecasting, 1997, 16, (1), 47–63.
19.
AlbertsonK. V.AylenJ.: ‘Modelling the Great Lakes freeze: forecasting and seasonality in the market for ferrous scrap’, Int. J. Forecasting, 1996, 12, (3), 345–359.
20.
AlbertsonK. V.AylenJ.: ‘Forecasting using a periodic transfer function: with an application to the UK price of ferrous scrap’, Int. J. Forecasting, 1999, 15, (4), 409–419.
21.
BoswijkH. PFransesP. H.: ‘Unit roots in periodic autoregressions’, J. Time Series Analysis, 1996, 17, (3), 221–245.
22.
AbramovitzM.: ‘Inventories and business cycles, with special reference to manufacturers’ inventories’, 1950, New York, National Bureau of Economic Research; K. V. Albertson and J. Aylen: ‘Forecasting the behaviour of manufacturing inventory’, Int. J. Forecasting, 2003, 19, (2), 299–311.
23.
HyllebergS.EngleR. F.GrangerC. W. J.YooB. S.: ‘Seasonal integration and cointegration’, J. Econometrics, 1990, 44, 215–38in ‘Modelling seasonality’, (ed. HyllebergS.) 1992, Oxford UP. We confine our test to nominal scrap prices because conversion to real prices using seasonally adjusted price indices – the only series available for the US – may introduce spurious seasonality.
24.
EngleR. F.YooB. S.: ‘Forecasting and testing in cointegrated systems’, J. Econometrics, 1987, 35, 143–159. On the importance of including seasonal dummy variables in the CRDW test in this context see K. V. Albertson, J. Aylen and K. B. Lim: ‘The power of the Durbin Watson test when errors are PAR(1)’, J. Statistics Computer Simulation, 2002, 72, (6), 507–516.
25.
DickensCharles: ‘The life and adventures of Nicholas Nickleby, containing a faithful account of the fortunes, misfortunes, uprisings, downfallings and complete career of the Nickleby family’, 1838–1839, Chap. 1.
26.
GarberP. M.: ‘Famous first bubbles’, J. Economic Perspectives, Spring 1990, 4, (2), 35–54 and P. M. Garber: ‘Tulipmania’, J. Political Economy, 1989, 97, (3), 535–560. These two papers cast doubt on the irrationality of these early bubbles.
27.
HunterW. C.KaufmanG. C.PomerleanoM.: ‘Asset price bubbles: the implications for monetary, regulatory and international policies’, 2003, Cambridge, MS, MIT Press.
28.
GilbertC. L.BrunettiC.: ‘Speculation, hedging and volatility in the coffee market’, 1993–96, n.d., London, Queen Mary and Westfield College Economics Department.
29.
An explanation for the scrap price boom in1974, see Committee on raw materials: scrap and the steel industry, Sections 8.1–8.20; 1983, Brussels, IISI
30.
See, for example, ‘Auto factory bundle prices rocket to $298 per ton in US on strong demand’, Metal Bulletin, 5 September2005, p. 15. Reported scrap stocks seldom fluctuate, which does not accord with casual observation of processors’ yards. In general, scrap statistics are limited, apart from published prices and customs statistics. Evidence on scrap use is scarce and seldom accurate. Detailed statistical coverage of ferrous foundries is limited up to the early 1990’s and ceases thereafter in the USA and the UK. There is no reliable data worldwide on foundry use. So, both total scrap consumption and scrap supply are understated.
31.
A view taken byAndersonD. F.: ‘Scrap as a commodity’, Ironmaking Steelmaking, 1987, 14, (3), 130–135.
32.
Millman (op. cit.), AylenJ.: ‘Choice of process route in steelmaking’, Ironmaking Steelmaking, 1990, 17, (2), 110–117.
33.
BarnettD. F.CrandallR. W.: ‘Up from the ashes: the rise of the steel minimill in the United States’, appendix C, 1986, Washington, The Brookings Institution.
34.
SumnerM. T.: ‘A cautionary note on cointegration testing’, Applied Economics Letters, 2004, 11, (5), 275–278.
35.
FransesP. H.: ‘Time series models for business and economic forecasting’, part 3, 1998, Cambridge UP.
36.
AlbertsonK. V.AylenJ.: ‘Modelling the Great Lakes freeze’ (op. cit.). Some forecasting software packages, such as forecastpro, do not allow estimation of lags this long even though indicated by the data.
37.
ThomasM. C.EvansM.DaviesR. L.: ‘Forecasting and scenario planning in the steel packaging industry: case of tinplate petfood cans’, Ironmaking Steelmaking, 1999, 26, (6), 463–469.
