Signals suggest commodity rally drawing to a close

The single biggest technical feature dominating the commodity markets is divergence on medium and long-term charts. This is afflicting both individual markets and the indices, and, in our view, implies that the rally of the past three years is mature: if these signals are validated in the first quarter, then we would expect 2005 to be the year the commodity rally comes to an end.

At the moment, it is the base metals markets where these signals are most advanced, and we are watching these closely for fresh evidence of a cyclical peak. For those who still prefer a bullish tack, the most pronounced “healthy” uptrend is in gold, and while support at 430 limits the downside, we would still allow for a run at US$479.40.

q Indices — We believe the general picture for commodities moving into 2005 is deteriorating. Divergence on weekly charts warns that the uptrend of the past three years is losing momentum, and if we close out 2004 below 288, then the signal would be repeated on monthly charts as well. Negative divergence on long-term charts is a warning that a major top is forming and implies that 2005 is going to be very different from 2004. Admittedly, we could still see marginal new highs in the first half of the year, but following a low December close, our strategic mindset would completely change from range-trading to selling rallies.

q Gold — Gold has been at the forefront of the general commodity rally and provided the recent breakout above US$430 per oz. is maintained, there is little to suggest that a major top has formed. Provided US$430 continues to provide solid support, we believe the technical outlook is for the bullish bias being maintained into the first quarter of 2005. The obvious targets are Fibonacci resistance at US$479.40 and the psychologically important 500 level. Thereafter, the proximity of the 1980’s highs should slow the pace of the rise.

q Silver — This is a market where the December close is of interest. Below US$6.87 per oz. and it would be a reversal month. We saw the same pattern in April, and the implication, in our view, is that a contracting range is forming and that silver will continue to whipsaw in a US$5.40-8.19 range well into 2005. A weekly close below trendline support at US$6.51 is required to undermine the bullish bias that has prevailed since 2001 and signal a return to the potential range low.

q Platinum — The cyclical uptrend stalled in 2004 and activity became locked in a choppy erratic range. Thus far, trendline support at US$806-816 per oz. is holding, and while this remains the case, we cannot discount the prospect of renewed bullish drift in the first quarter towards US$885 or US$909. However, a closing breach of the trendline would change the outlook dramatically; below US$806 implies 2005 will bring a deeper correction. Initially, following a breach of the trendline, we would expect a decline to at least US$738 and possibly US$672 before the market stabilizes and recovers. At the moment, in mid-range, this is not the most exciting chart, but that will all change if the trendline gives.

q Palladium — The rebound in 2003 off the last retracement level, of the rally between 1991 and 2001, offered some support to bulls. However, the rapid rejection of levels above US$300 and the absence of any buying interest above US$200 over the past six months suggest that the rally was no more than a correction within the context of the cyclical downtrend. We believe this is a warning that the risk of a re-test of US$138.70 in 2005 should not be under estimated. While capped by US$210, the immediate focus for bears is support at US$182 and US$168.80, but if those levels give, then a return to the 2003 low should be expected.

q Aluminum — Divergence on monthly and weekly charts, combined with key weekly reversal patterns in April and October, all point to the uptrend off the 2001 low running out of steam. Our wave count suggests there may still be a final rally toward US$1,902 per tonne, but that would be the fifth and final leg of a “diagonal triangle,” which would indicate that 2005 will be the year aluminum tops out. To the downside, trendline support at US$1,730 is the important benchmark for bears. A closing break of the line would signal that the top is in place and imply a return to US$1,550 per tonne. In 2005, we are looking to sell rallies.

q Copper — In our view, all signs point to the highs of the past 15 years continuing to prove an insurmountable barrier in 2005. The usual suspects — divergence and wave count — point to an impending top . . . if indeed copper has not already topped out. While the metal is supported above US$2,816 per tonne, we will keep an open mind about the prospect of a final surge against the long-term range high. However, following such a move, we would be champing at the bit to sell in anticipation of a test of US$2,480 per tonne, possibly even US$2,260, later in the year.

q Tin — After such an explosive rise, we have rarely seen such a muted response in other markets. The shock (or manipulation) of 2002-2003 has paralyzed investors, and this market has come to a standstill. From the slight positive bias to the range over the past six months, it is clear that demand has not been fully satisfied. The September low at US$8,500 per tonne defines the base of the trading range, with US$9,300 the top. A break of US$9,300 would signal a return to US$9,800. A price below US$8,500 is required to squeeze out trend followers. It is a market to avoid, in our view, until there is a decisive breakout.

q Nickel — The choppy range that has dominated trading this year does not yet appear mature, and we believe the outlook for 2005 is more of the same within the contracting range. In the first half of the year, we will be focusing on trendline support at US$11,750 per tonne, given the rejection of higher levels this quarter. A break of the line would imply a further slump in the first half down to key long-term support at US$10,300 or possibly to the US$9,700 swing target associated with the range this year. If trendline support holds in the first quarter, we would expect the volatility within the range to diminish gradually.

q Zinc — Repeated tests of Fibonacci resistance at US$1,209 per tonne throughout the second half of 2004 were all unsuccessful and caused the weekly Relative Strength Index to roll over: the uptrend off the low in 2002 is losing momentum. Evidence of bearish divergence against such important resistance is the sort of signal we expect at a major high. A low close in December (below US$1,175 per tonne) would replicate the signal on monthly charts and reinforce our belief that 2005 will be a negative one for zinc. In our view, even a breach of US$1,209 would not materially change the picture, given that there is further strong resistance at US$1,245 from the highs back in the 1990s. A break of trendline support at US$1,091 is required to signal the top is complete and target a return to US$915, possibly US$836.

q Lead — Throughout 2004, lead has consistently been the most bullish base metal, but despite making multi-decade highs, the price action is infected by the same evidence of topping activity that is eating away at the other markets, namely divergence and an increasingly bearish wave count. The problem we have with this market is identifying the point at which the uptrend will end. The rally is off our charts and therefore, apart from the psychologically important US$1,000-per-tonne level, we are left to rely on trendlines and projection targets. However, given the signs for an impending top, we are equally concerned that trendline support at US$823 will be retested in the first quarter of 2005; if breached, then we believe lead is heading back down to US$650.

— The opinions presented are the author’s and do not necessarily represent those of the Barclays group. For access to all of Barclays’ economic, foreign-exchange and fixed-income research, go to the web site at barclayscapital.com. Queries may be submitted to the author at phil.roberts@barcap.com