GOLDMAN SACHS is used to moving prices with magisterial statements that some share is worth buying or, in broker-speak, has been downgraded to underperform. But last week a team of the investment bank’s New York analysts caused the world’s second most important price to move: crude oil prices in Europe and America hit new highs.

The analysts, led by Arjun Murti, offer the stark message: “We believe that oil markets may have entered the early stages of a super spike period, which we now think can drive oil prices toward $105 per barrel.”

Over the past 18 months, the price of West Texas Intermediate, the US benchmark, has doubled to about $57 per barrel. A further rise of up to 85 per cent is asking a lot.

The more substantial fear is that prices will average $50 to $55 per barrel in 2005-06 instead of reverting to an average of $40 a barrel. If so, we are in for a period of sustained high oil prices not seen since 1979-82, when they triggered a worldwide recession.

Mr Murti concluded that investors should buy most US oil stocks. Colleagues at Goldman’s London office were not quite so enthusiastic. They upgraded the oil and gas sector from underweight to neutral. At the predicted rate, they reckon, oil shares would trade at an average of only eight times 2006 earnings, 30 per cent lower than the average forward share rating. That rating is much lower than it would be if oil prices are assumed to fall.

On the consensus view, companies such as BP or BG Group trade nearer to the market average of 12 times projected 2006 earnings. As the London team points out, however, “higher oil prices do not necessarily mean sustained higher returns on capital”.

What sustained high oil prices generally mean is hard times ahead. This is the wood that Mr Murti and his colleagues miss when they examine the impact on the oil trees in their back yard of the price staying above $50 a barrel.

Their argument is that global demand for oil is rising faster than supply can catch up. China, India and Eastern Europe are all growing rapidly and needing more. They do not assume any big disruption to supply but simply note that oil output in Russia, Iraq and Venezuela is unlikely to rise as fast as it could if managed on Western lines and that most other producers are already pumping as much as they can. Oil demand is not so price- sensitive as it once was. Far less is now used for electricity and Americans spend much less of their income on petrol than they did 25 years ago.

American demand will not drop much just because the price of a US gallon has risen from an average of about $1.73 a year ago to $2.21, the equivalent of about 30p a litre. Hence Mr Murti and his colleagues argue that the price would have to spike up to $4 per US gallon, about 56p a litre, as “only a sharp sustained increase in energy commodity prices will meaningfully reduce consumption and recreate the kind of spare capacity cushion that existed through much of the 1980s and 1990s ”.

They do not mention the consequences, which would probably include plunges in US consumer spending, vehicle demand, leisure industries and jobs, just like 25 years ago.

Any non-American is bound to think that there is a much easier way. A tax of $1 per gallon, made in five annual steps and announced in advance, would eliminate the US budget deficit, support the dollar and help to curb oil consumption in a predictable way that allows the motor industry to market new ranges of vehicles using less fuel and maintains consumer confidence. It would also, in future, give the Federal government a weapon to stop future oil spikes hurting the US economy. Sadly, we know that US oil is influential and this will not happen.

Long before petrol reached $4 a gallon in America, demand would have fallen in poor oil-consuming countries, China would probably suffer a financial crash and eurozone output would grow more slowly than a box hedge. The UK should suffer less because oil money will again be recycled through London into Western financial assets and property.

If the Goldman “super spike” did fly our way, most sectors would be a sell. That is the key risk to watch out for this summer.