London, 28 April 2011: Reuters

Market participants are probably over-estimating the level of prices needed to achieve demand destruction and restore oil market equilibrium because their judgement is over-influenced by recent high prices and fails to give sufficient weight to a longer price history.

Psychologists and behavioural economists have long known that humans assign too much weight to recent events at the expense of a longer perspective. One result is the phenomenon known as anchoring -- in which estimates of value or forecasts for the future cluster around current levels or recently mentioned numbers, even fairly arbitrary ones.

In "Priceless: The Hidden Psychology of Value", William Poundstone reviews the numerous experiments showing how estimates and predictions are influenced by recently mentioned numbers -- even when the subjects knew those numbers were entirely irrelevant. Estimates and predictions could be influenced even by random numbers generated by the spin of a roulette wheel.

Anchoring has been observed for inconsequential, low value decisions where individuals have an obvious incentive to take mental short cuts rather than waste time and effort computing estimates and forecasts from scratch. But it has also been shown to affect serious decisions involving high-value items, even for goods and assets with a market determined value, such as real estate.

OIL MARKET ANCHORS?

The question is whether recent price rises and the whole period of high oil prices since 2006 have imparted an upward bias to estimates of the level at which demand destruction sets in, which is sustainable in the long-term, or even "fair" for producers and consumers. There are good reasons to believe anchoring is indeed at work in the oil market. Top-producer Saudi Arabia certainly thinks so.

Last year, Oil Minister Ali al-Naimi spoke approvingly of a price band between $70 and $90 per barrel and observed "There's almost an anchor now for the price" . Other senior officials cited the importance of fostering convergent expectations among producers, consumers, financial institutions and the media to dampen volatility.

As the band came under pressure Naimi lashed out at the media for destabilising the market blaming it for wanting "something we haven't done yet. You want to cause disturbance in the market. The way you ask questions about price, about production, about supply, what you do is become an agent of disturbance in the market"

SUBCONSCIOUS INFLUENCE

Anchors are clearly important in the Kingdom's thinking. But there are indications they influence other participants too. Estimates for demand destruction and market equilibrium make repeated reference to peak prices of up to $147 per barrel (WTI basis) reached in 2008. Even higher numbers, such as the famous prediction of a super spike to $200, are floating around.

The point about anchoring is not that everyone revises their estimate of price/value up (down) to precisely the level of the anchor. Instead anchors seem to exert a gravitational pull on estimates and forecasts. They sub-consciously influence analysts and investors to adjust their predictions and expectations higher (lower) at the margin towards the anchor. There is no direct evidence anchoring effects are occurring

in the oil market at present. It is not possible to control for all the myriad influences on price forecasts. But it would be surprisingly if anchoring effects were not happening given their prevalence in other markets. Repeated upward revisions to price forecasts and the estimated threshold at which analysts think demand destruction will occur (forecast "stretching") suggests some degree of anchoring driven by recent price increases and the spike of 2008 is occurring.

When oil was trading steadily in a range of $65-85, most analysts thought demand destruction would commence at around $100. But as prices have marched higher the threshold for demand destruction has been pushed back to $110, then $120 and now $130 or even higher.

Before the price spike in 2008, current crude and gasoline prices would have seemed unimaginably and improbably expensive. Now they seem comparatively cheap, or at least normal.

COHERENT ARBITRARINESS

Evidence for anchoring is part of a rich literature showing humans are bad at estimating and forecasting prices. Empirical research shows individuals and groups are good at estimating relative prices but not absolute ones.

In the oil markets, most participants know that light products such as gasoline and distillate fuels should command a premium over unrefined crude, while residual fuel oil and bitumen should trade at a discount. Most could identify a range of normal spreads defining the expected relative relationship between crude and product prices.

But there is much less certainty and agreement about what price crude oil, or gasoline or residual fuel oil should command in absolute terms. In the past year, leading analysts have cited equilibrium prices ranging from less than $100 per barrel to more than $130. Over a five-year period, the range of disagreement is even larger.

It is the same for other assets. Observers can agree a larger house is worth more than a smaller one, other things being equal, and perhaps even assign a fixed dollar premium to the difference. But there is much less agreement about what either house is worth in absolute terms, as U.S. homeowners have found to their cost.

Psychologists refer to the phenomenon as "coherent arbitrariness". Individuals are good at ensuring relative prices are consistent with one another but less good at pegging the level of the overall set of prices.

In a world where absolute prices are difficult to estimate, analysts and forecasters reach out for the nearest or most recent anchor, and use that as a baseline, adjusting predictions up or down from that starting point.

POSITIVE FEEDBACK LOOP

Fundamental analysts insist prices for commodities, uniquely among assets, are determined by physical supply and demand. Most would admit both current and expected fundamentals drive market prices. But current fundamentals are not known with certainty and expectations about the future are subject to even bigger errors

Once expectations enter the determination of prices, the question of how those expectations are formed, and the role played by anchoring, if any, becomes crucial. In the absence of a clear understanding about how current oil prices are influencing supply and demand, or how prices will affect the production-consumption balance in the next 6-24 months, producers, consumers and investors are all groping for anchors on which to base their forecasts.

Memories of the recent period of high prices in 2008 as well as the recent rise in prices and some of the more bullish forecasts may therefore be pulling price forecasts and expectations upward. They may be one mechanism contributing to a positive (self-reinforcing) rise in prices and expectations.

Ends --

By John Kemp, Reuters market analyst – for Commodities Now.

The views expressed here are his own.