It is a challenge to keep track of all the raw materials ballooning in price around the world. The latest obscure entry to the list is none other than potash. This humble material is one of the three essential nutrients used in fertiliser, along with nitrogen and phosphorus. Reserves under ground are ample - 100 years worth or more - but geographically concentrated. Just 12 countries produce potash, and almost four-fifths of reserves lie in Canada, Belarus and Russia.

Until recently, the industry had plodded along for more than two decades. Prices doubled between 2004 and 2006 to $200 per tonne, yet this level was only towards the top end of the historical range. It is since the start of 2007 that a renewed push to boost crop yields has caused demand to soar. Indian buyers agreed to pay $625 per tonne this year. Spot prices of $750 per tonne have been reported in Brazil. With potash inventories at record lows and supply lagging demand, Morgan Stanley expects delivered prices could hit $1,000 per tonne by the end of 2008.

Share prices, naturally, have flown. The market capitalisation of the New York-listed Potash Corporation of Saskatchewan, the world's largest producer, has nearly tripled in 12 months to $60 billion. A similar performance from miner K+S means that it is knocking on the door of Germany's Dax index. Russian-listed Uralkali is up almost six fold.

Valuations reflect this optimism, with PotashCorp trading on 20 times this year's earnings. Yet even if, as expected, earnings double again for the group next year, the oligopolistic returns it is generating look unsustainable long term. If the miner meets consensus estimates for 2009, it will generate operating profit in one year equal to roughly twice the total value of its tangible fixed assets.

Growing appetites in developing countries mean fertiliser demand will continue to rise. Greenfield potash projects take five years to develop. Announced industry capacity plans suggest supply will catch up only in 2012. There is no doubt fundamentals are bullish but given where equity valuations are, the only sane response is to grab a spade and dig.

New Zealand nationalises trains

Even New Zealand, that far-flung crucible of free market principles, has failed to make railway privatisation work. Ownership of rail and ferry services is to pass back into public hands, bringing a conclusion to what has been called "one of the most disastrous privatisations in New Zealand history".

Renationalisation was long on the cards. Under private ownership underinvestment led to accidents. That scared off passengers, lenders and suppliers, forcing the government to step back in four years ago. Since then, the state has owned the rails while Toll Holdings of Australia operated the trains. But even that proved tough, notwithstanding a cash subsidy and in effect monopoly status.

Subsequent performance illustrates the point. In spite of high fuel prices and environment-driven concerns to push freight off ships and on to trains, there has been only a modest increase in usage. Freight volumes rose four per cent last year to 40.5 million tonnes and, while some long-distance lines did well, overall passengers still clock in at fewer than 12 million a year. When costs were low, Toll could hold it together, but the rub came when New Zealand's government wanted recompense for upgrading the track. Passing on the full cost would, with current usage, render the project commercially unviable - a public transport problem familiar elsewhere. Toll shareholders were not too disgruntled by the group's exit from the New Zealand rail business, sending the shares up four per cent in Australia yesterday.

That looks like a rational verdict. Investing in railways has caused heartache to investors since the 19th century, and doing so in a huge country with a tiny population (at four million), New Zealand has half as many people as London) is more hair-raising still. As for the government, doing the right thing by a public service in an election year brings its own benefits.