City was warned over Shell reserves in 2002

The City was warned that Shell might have a problem over how it measures its reserves more than 18 months ago, The Independent on Sunday can reveal.

The oil giant shocked shareholders earlier this month when it downgraded a fifth of its "proven" oil and gas reserves to "probable".

The downgrade raised doubts over the methods used by the company's exploration and production division, whichalso estimates the potential of newly discovered fields.

Edward Ennis, an oil and gas analyst at Zurich-based fund manager Julius Baer, told The IoS that he first challenged Shell's chairman, Sir Philip Watts, about the oil company's methodology at an analysts meeting in June 2002.

Then an analyst with Société Générale, he questioned whether Shell used the "deterministic" method, which usually results in higher estimates of reserves, more than its peers. When oil companies estimate reserves during early exploration, they use the "probabilistic" or Monte Carlo model, which is more conservative. As the field is developed, companies gradually switch to the deterministic model. At the time, Sir Philip, who was criticised this weekend by shareholders for his communication skills, refused to answer the question. Sir Philip was head of the exploration and production division before becoming chairman in July 2001.

Shell insists that the downgrade will not materially change how much oil and gas the company expects to recover. But shares still fell 7 per cent as shareholders questioned other assumptions made by the company. In the last few days, other analysts have requested copies of Mr Ennis's research.

Mr Ennis said: "I raised the question and specifically asked Phil Watts if Shell used a deterministic approach in evaluating reserves when some of its peers were using a Monte Carlo approach, given that the Monte Carlo method generally underestimated reserves when compared with the deterministic method. Phil Watts refused to answer."

Mr Ennis stressed that, over the longer term, any distortion would be smoothed out. "Even if reserves were initially overvalued relative to its peers, the correction would eventually come through in the 'revision to reserves' category as reported by companies. After three to four years, it should all wash out. Basically the devil is in the detail."