As the oil price has fallen back from its high of $147 a barrel last month, bank shares have rallied. Like the gyrations of a giant barbell, one end (banks) rises as the other (oil) falls. Just a few weeks ago the barbell looked so violently unbalanced as to threaten the hapless victims under the end marked 'banks'. Indeed, without the retreat in oil, the slide in confidence across this sector might have been even more devastating than it proved in mid-July.

Earlier this year investors bid up dollar-denominated oil futures as a hedge against a falling dollar and inflation. Any sign of a stronger greenback is often enough to give pause to a rally. The central banks' actions fed investor sentiment that economic growth is slowing in the developed world, cutting demand for crude.

So how much further and longer has the oil price to go? Does this signal the turning (at last) of the great commodity cycle and the possibility – one dare not describe it as hope – that inflation and central bank interest rates may start to ease earlier than thought? Might this not mark, if not the end of the credit crunch, then the end of the worst?

The answers may turn out to be both reassuring and the cause of further unease. An early point to note is that the oil price spike to that $147 peak may not have been as significant for the 'real' economies of the US, Europe and the UK as had been originally feared. According to Opec, the average price of its basket of crudes rose from $128.33 a barrel in June to $131.22 in July. The cost of oil to end-users does not match market price gyrations on a daily basis but reflects contracts struck at various times in the past. So there is a time lag on the effective change in oil prices as they work through the supply and distribution chains.

Economist Stephen Lewis of Monument Securities points out how long that time lag can be. In 2007, the average price of US oil imports was $64.28 a barrel, slightly under the $69.08 average Opec price, reflecting differences in grades of oil. In January the average price of oil imported in the US was $84.09, while the average Opec price was $88.35.

As oil prices began to rise in the early months of this year, the gap between US import and Opec prices widened, as the import prices lagged the market prices. By April the US import price was $96.81 whereas the Opec price was $105. In May the US import price was up at $106.28 but the Opec price had risen further to $119.39. The pattern suggests that the delivery prices were lagging market prices by about one month.

This matters, because while the drop in the US market price of oil of $30 from its peak of $147 has been dramatic, it will probably have a disappointingly limited impact on the economy. Lewis points out: "The oil price was around its elevated level for less than three weeks. There would hardly have been time for it to become embedded in the cost structure of the global economy. The negative effects from expensive oil that finance ministers and central bankers have noted in the economic data largely reflect a build-up in costs consequent on earlier phases of the surge in crude oil prices. So far, the scale of the latest falls in these prices is not likely to have been sufficient to reverse much of that damage."

Even so, it seems likely the oil price will fall further in the short run. Consider previous gyrations in the price. Over the past six years, sizeable setbacks have punctuated an upswing that began with the Brent blend crude price at around $20.

Four major market corrections stand out. The post-Iraq war setback in 2003 took the Brent price down from $34.94 on March 10 to $23.23 by April 29, a decline of 33.5%. Then, in 2004, the price slipped from $52.28 on October 22 to $36.77 on December 13, a fall of almost 30%. After Hurricane Katrina in 2005, Brent crude was priced by November 18 at $52.84, some 21% below the $66.80 at which it had traded on August 31. Finally, after Israel's military action in Lebanon ended in 2006, the oil price declined by 29%, from $78.26 on August 9 to $55.82 on October 3. The current pullback to $118 marks a decline of 19.7% – shy when compared with previous retreats. Further, it has only been 28 days since the oil market reached its last peak. Past market corrections have all lasted longer. In 2003, the correction took 51 days; in 2004 it took 52 days; in 2005, 79 days; and in 2006, 54 days.

This would suggest that a further retreat in the oil price to below $100 a barrel, while delivering a big psychological fillip to the market, would not be much different from previous corrective market moves seen at earlier stages in a long-run oil price upswing. This would douse the perception of a historic 'peak' or fallback marking some critical change in the direction of oil and commodity prices.

It is tempting to see the oil price fall as a product of a sharp drop in consumer and business demand. It is certainly true that retail consumption in America, so long associated with the gas-guzzling Hummer and a motoring population heroically wedded to the high-consumption 4x4, has fallen.

America's highways have indeed become the road less travelled this summer. Vehicle miles travelled on all US roads fell 3.7%, or 9.6 billion fewer miles, in May compared with the same period a year ago. This was the seventh month in succession that vehicle miles in the US have fallen.

And America's car giants have taken a big hit. General Motors recently reported a net three-month loss of $15.5bn (£7.8bn) as North American sales fell by 20%. The car maker has had to launch a further round of redundancies after announcing cuts in production in the wake of falling sales of vehicles, especially SUVs and trucks.

Ford also reported that sales of SUVs were 54% lower and trucks and vans were down 18%. Toyota also sold 32% fewer SUVs than in July 2007 and 33% fewer light trucks. Passenger car sales were 8% down. The oil spike has changed driving habits and car-purchasing behaviour.

However, big though it is, America is not the only consumer. It would require a collapse in growth rates outside the OECD area to create a general weakness in the global economy strong enough to undermine demand for oil. The OECD accounts for only 56% of total world oil demand.

Moreover, even if there were a slowdown in global economic activity, upward pressure on the oil price would probably persist because the supply position is not as comfortable now as it was in earlier years.

So where is the price of crude likely to go? With mounting evidence of a slowdown in growth around the world, with the economies of China and India also having to throttle back double-digit inflation growth rates, global growth is definitely slowing, and with mounting worry that this slowdown is going to extend far into 2010. That would suggest a further downward move in the oil price.

The hope is that this will deliver a big stimulus to households. After all, pump prices in America have recently been scraping $4 a gallon. When the price was around $100 earlier this year, the national average US gasoline pump price was just $3. That reduction, reckons veteran US economist Ed Yardeni, "would surely boost real incomes, consumer confidence, driving and shopping".

That, for the reasons discussed, is by no means automatic. The problem is that the continuing squeeze on banks' ability to lend, combined with retrenchment by consumers as unemployment rises, suggests deepening economic weakness ahead. And it will be felt not just in America but also in big exporting developing countries round the world. Falling oil would be a sign not of recovery on the horizon, but that the slowdown is global and is intensifying to make America's problems worse.