COLUMN-Goldman's oil bulls seize forecasting crown: John Kemp

Wed Dec 14, 2011 7:03am EST

By John Kemp

LONDON Dec 14 (Reuters) - For the second year running, analysts at Goldman Sachs are among the most bullish for oil prices over the next 12 months.

In 2011, the bank's analysts took a risk by making a bold call on prices -- and won big when prices soared. Rivals and customers will be watching closely to see if its call for continued price rises in 2012 proves equally prescient.

Goldman predicts Brent crude will average $120 per barrel in 2012, up from $111 in 2011, with only a handful of trading days left before the end of the year.

The bank's optimism is matched by UniCredit (which is also predicting average prices of $120) and outdone only by CIBC ($123).

Goldman's analysts have consistently been among the most optimistic for the past five years -- earning them the nickname "perma-bulls", and a mixture of envy and admiration from rivals, as prices have regularly surprised on the upside, making the bank more right than wrong.

The bank has triumphed again as the most accurate forecaster in 2011, as strong demand from emerging markets in the first half, coupled with the civil war in Libya and other supply disruptions, sent prices surging 40 percent compared with 2010.

Goldman's track record ensures its forecast has become the benchmark and the one to beat.

GOLDMAN'S BOLDNESS PAYS OFF

Last year, Goldman predicted Brent prices would average $100 in 2011, which was considered aggressive, when the average forecast was just $87, according to a Reuters poll carried out in December 2010.

In the end, the bank's prediction was wrong, but only because it turned out $11 (10 percent) too low. It was significantly more accurate than the next nearest prediction, which was out by $16 or 15 percent, let alone the average, which was off by $24 or 21 percent.

Goldman's analysts have been distinguished by their readiness to break with the herd and defy market convention by predicting prices will change rather than remain roughly the same.

In December 2010, the average forecast for 2011 of $87 was not meaningfully different from the prevailing price ($92) or the forward curve. Most of the "forecasts" were really "nowcasts" (Chart 1).

The 2011 forecasts were also tightly clustered, with a standard deviation of just $5.50 (6 percent), which some would say was unrealistically low, given prices typically move much more from one year to the next.

In contrast, Goldman's bullish call for $100 Brent deviated from the consensus almost three times as much as the average. In predicting prices would change significantly, rather than continue around current levels, Goldman proved correct, though even the bank failed to appreciate how far the market would move.

MOST STICK WITH THE CONSENSUS

Price forecasts for 2012 look very much like those for 2011. The baseline has moved up $20 but most analysts are again predicting little change over the next 12 months.

The average Brent price forecast in a poll conducted by Reuters in late November was $107 per barrel in 2012, virtually identical to the prevailing price of $109 and not significantly different from the $111 average in 2011.

Price forecasts for 2012 are again strongly clustered. The standard deviation is $9.33 per barrel. While the deviation is almost twice as large as at the same time last year, in percentage terms (8.7 percent) the predictions are only slightly more dispersed than for 2011 (6.4 percent) once adjusted for the higher base price (Chart 2).

On average, forecasters expect prices to move less than 4 percent next year. Nearly half the analysts in the survey (17 out of 36) think average prices will differ from 2011 ($111) by $5 or less ($106-116). Less than a third think prices would fall 10 percent to $100 or lower (11/36). Only 3 analysts think prices would rise by more than 5 percent on average.

Such continuity would be unusual. The oil market has rarely exhibited stability for such an extended period. The average rise or fall in prices from one year to the next over the last two decades has been 22 percent (Chart 3).

Prices have moved less than 10 percent on only six occasions since 1990. The last time prices moved 4 percent or less was 2002, and before that 1997. So analysts are forecasting an atypical degree of stability in 2012. It is more likely they will all prove wrong than that they will all prove right.

TAIL RISKS AND CENTRAL TENDENCY

The tight clustering of the price forecasts around the current level stands in marked contrast to prominent warnings about tail risk on both the upside (from supply problems) and the downside (from Europe's debt crisis and slowing global growth).

Published commentaries and many news stories highlight the possibility of sharp price moves if some of these risks eventuate. But the broadly flat forecast distribution for 2012 suggests this is not the central scenario for many analysts in the survey (or even the perma-bulls at Goldman).

Instead, the broadly flat distribution implies: (1) the probability of a large price spike or collapse is not high; (2) any large move is expected to be swiftly reversed; and (3) risks to the upside are largely matched by threats to the downside.

While many commentators talk in apocalyptic terms, almost all the analysts in the survey have a conservative outlook -- one which emphasises unstable stability and more of the same rather than large changes.