The sequence of events following my suggestions, made in the first quarter of 2004, about the attractiveness of commodity- related stocks provides an object lesson in investment practice.

Anyone holding UK stocks like BHP Billiton and BP will have achieved two key objectives: they will have made money and beaten the broader market. Since the original Serious Money articles that mentioned these stocks there has been a rise of nearly 6 per cent in BHP's price and a rise of nearly 13 per cent in BP's - both against a background of a fall in the broader UK market.

Observing the way in which we have achieved these gains is extremely instructive. In particular, we obtain valuable insights into what to do next. The links with commodity prices are clear and the behaviour of those prices - most obviously oil prices - will determine not only the future performance of these companies, but also the outlook for world economy and all other asset prices

BHP Billiton is a UK-listed global company with broad exposure to commodities. It describes itself as the world's "largest diversified resources company".

Despite the underlying complexities of the business, a significant slice of profitability depends on the behaviour of commodity prices. We all know how those prices have been strong - not just oil - and an examination of this month's results statement shows just how that translates into profits for commodity producing businesses.

Overall earnings before interest and taxes (EBIT) were up 58 per cent in the year to end June, with divisional results including a 304 per cent rise in earnings from base metals production, 34 per cent from aluminium and 18 per cent from petroleum. Even profits from coal - a commodity long since abandoned by most investors - were up 18 per cent. The unusual nature of the commodity price revival is that it has spread to the coal market, eliciting much wry commentary in places like South Wales and Yorkshire.

Commodity price rises led directly to a $3.1 billion (€2.56 billion) rise in earnings (with offsets coming mostly from the weakness in the dollar). The results pose two questions: will commodity prices remain high and is strong profits growth now all in the price?

Forecasting commodity prices is a mug's game, but people who profess some expertise in this area are getting worried that the boom is now over, if for no other reason than the recent rise has sown the seeds of its own destruction.

First, prices depend on demand, which will now slow, not least because of price rises. Second, it is a historical regularity in the commodities business that whenever prices rise, various firms engage in a suicidal rush to introduce new production capacity that almost always ends in a price bust as a result of over-production.

Not for nothing are mining companies considered the ultimate "deep cyclicals" of the stock market. Anyone betting that commodity prices will continue to rise, let alone hold their gains, must also be willing to make the bet that things will be different this time.

BHP seems to think that something is different, and that is the result of China's continued expansion. While they acknowledge the efforts of the Chinese government's attempts to slow the economy down, BHP clearly believes there is something structural going on and China will continue to be a growing source of profits - currently nearly 10 per cent of the total. China consumes nearly one-third of world coal output, nearly 25 per cent of global steel and aluminium output, and 40 per cent of cement production.

And yet it still only accounts for 4 per cent of the world GDP. Nobody saw the China story coming, the consensus is that China must slow and many think a bust is inevitable. Imagine, by contrast, that China's growth continues: the economy will double in size in less than a decade.

Another thing that is different is that the commodities industry is now better managed, with lemming like rushes to over production a thing of the past (at least on the same scale as the past).

Now, many analysts are engaged in a headlong rush to cut back on global growth forecasts. That usually means the time has come to sell commodity stocks. But the key way in which things could really be different this time is if any cyclical falls in commodity prices come at a time when the trend in such prices remains upwards - something that has not happened for at least 20 years.

If BHP is right about Chinese demand, we should look through any short-term weakness in commodity prices and hold the shares. If we want to be really clever, we should sell now in anticipation of some short-term weakness as others make the mistake of assuming another downward trend in commodity prices.

Such tactics should probably be left to traders. Getting caught up in the short term vagaries of the commodity cycle is usually a recipe for an early grave. This is where the history of BHP's share price since we first mentioned the stock is instructive. Initially, the stock held its own but then put in a dreadful performance for a period of about six weeks starting in April. Since mid-May, it has put in a stellar rise, with the market becoming more aware of the profits story.

If we had been traders operating with "stop-losses" we would almost certainly have learned the classic "right idea wrong time" lesson of investing: forced selling at the lows. Timing is the hardest part of any investment strategy and trying to be too clever is almost always a recipe for disaster.

My hunch is that BHP is right: there is a demand story that will underpin higher commodity prices. Expensive petrol is here to stay, even if crude prices lose some of their speculative froth. Commodity and oil stocks currently make an unusual story: both should be regarded as sound long term investments.

The current vogue for getting more pessimistic about world growth is understandable. Like all such trends it will almost certainly be overdone. Only if we start to believe that a global recession is imminent should we revise our belief in commodity stocks. Things have to get much worse before we can contemplate that awful vista.