Investors' hopes for an early recovery in the mainland steel industry look overly optimistic. Since the Hong Kong stock market bottomed in late October, the shares of locally listed mainland steel companies have rebounded handsomely, with both Angang Steel and Maanshan Iron & Steel comfortably outperforming the H-share index (see the first chart below). The rally was fuelled largely by a bounce in mainland steel spot prices from November's lows (see the second chart), coupled with high hopes that Beijing's massive stimulus programme of infrastructure spending will lift demand for the metal. Those hopes may be disappointed. Although on the surface the outlook for steel company earnings does appear to have improved, the underlying picture remains bleak. Even before the onset of the economic slump, China's steel industry was plagued by overcapacity. After the rapid expansion of recent years, mainland mills now have the capacity to turn out more than 600 million tonnes of steel a year. Last year, with capacity utilisation running at an average of 80 per cent, actual production came to about 500 million tonnes. Domestic demand, meanwhile, amounted to about 450 million tonnes. Much of the excess was exported, but even so, the result was a rapid build-up of inventories. As demand cratered in the final quarter of last year, flattened by the twin blows of the mainland's domestic property downturn and the crumbling international economy, mainland steelmakers responded by halting production at many mills and slashing prices in an attempt to shift stocks. The result was the downward spiral that saw steel spot prices halved in just a few months. Not surprisingly, steelmakers' profits, already under pressure from last year's 85 per cent rise in Australian iron ore prices, suffered badly. Earlier this month, China's largest steelmaker Baosteel said last year's earnings fell 32 per cent compared with 2007, while the second-largest, Angang, said its profits dropped 55 per cent. Things now appear to be improving. Since the beginning of December, steel prices have recovered modestly and production has been stepped up. Yet the improvement is largely illusory. The recovery is the result mainly of the last quarter's vicious destocking rather than any genuine pick-up in user demand. Beijing's stimulus measures will help. Its programme of stepped-up infrastructure spending is expected to generate extra steel demand of about 70 million tonnes this year. But that is unlikely to compensate for the drop in demand elsewhere. With private sector construction moribund, new car sales stalled and shipyards facing cancellation rates of up to 40 per cent of their orders, overall domestic demand is still likely to fall this year. Exports will be no answer. International demand is suffering even more and protectionist calls are mounting. This week, the US House of Representatives approved a measure requiring that government-funded stimulus projects must use only US-made steel. Earlier this month, the European Union slapped a 25 per cent duty on imports of steel reinforcing bars from China. Nor should producers expect too much from their annual ore price negotiators. Mainland steel companies are believed to be demanding a price reduction of about 40 per cent, while the big miners, after cutting their own production, hope to restrict any cut to just 20 per cent. Beijing has promised to help the industry by blocking the expansion of production facilities. But even that is a double-edged sword. Steel mills use up a lot of steel. Stopping their construction will suppress demand further. Nor is trying permanently to close old or smaller mills likely to significantly reduce capacity. Beijing has tried that before, only to run into fierce opposition from local governments desperate to protect jobs. As a result, the outlook for mainland steel producers remains a picture of high costs, excess supply and weak demand. Investors beware. As a number of readers were kind enough to point out, the headline on yesterday's Monitor column should have read 'Manipulated? Yes, but maybe not undervalued' instead of 'Manipulated? Yes, but maybe not overvalued' as published. Our grovelling apologies.