All that glitters is not gold

[HenryTo]

This is from three days ago but I think it is definitely worth posting. Mark Hulbert is now bearish on gold based on his increasingly bullish readings from the HGNSI survey despite declining gold prices during the last six weeks. This makes another good case that the US$ should continue to go up here.
------------------------------------------------------------------------------------
All that glitters is not gold

By Mark Hulbert, MarketWatch
Last Update: 5:50 AM ET May 27, 2005

ANNANDALE, Va. (MarketWatch) -- Contrarian analysis may have acquitted itself very well in the stock market over the last couple of months. (Read my column on this subject from earlier this week.)

But it has fallen flat in the gold market. Sentiment among gold timers fell in mid April to close to a record low, a development that was bullish according to contrarian analysis. (Read archived column.) And yet gold bullion is today trading for some nine dollars per ounce less than where it traded then.

That's bizarre enough, but there's more. Since mid April, while gold bullion has been declining, the average gold timer has become more bullish. That's highly unusual, as the typical pattern is for the average timer to become more (less) bullish as the market rises (falls).

Consider the latest readings from the Hulbert Gold Newsletter Sentiment Index (HGNSI), which reflects the average exposure to the gold market among a subset of short-term gold timing newsletters tracked by the Hulbert Financial Digest:

On April 19, when (38099902: news, chart, profile) gold bullion was trading at $427.25 per ounce, the HGNSI stood at minus 30.4 percent, only 0.9 percentage point above its multi-year low of minus 31.3 percent.
As of Thursday's close, with bullion trading at $418, the HGNSI stood at 12.5 percent. That means that the average short-term gold timer tracked by the HFD has increased his gold market exposure by nearly 43 percentage points in the face of a significant decline.

How would a contrarian interpret these developments?

I think they are bearish for the gold market. The timers' reaction over the last month betrays a significant undercurrent of stubborn bullishness - which, history teaches us, is rarely vindicated.

It's reminiscent of how stock market timers reacted in the weeks immediately following the March 2000 stock market top. In the wake of the first 10 percent decline from that top, the average short-term market timing newsletter significantly increased its equity exposure. As I concluded at that time, such stubborn bullish is rare - and bearish.

I know, I know. Only a little more than a month ago I was declaring there was excessive pessimism among gold timers. How can I turn around and now say that they are showing signs of stubborn bullishness?

Because that's what the data show, improbable as it otherwise seems. And over the last six weeks, the data unambiguously show that the average gold timer has aggressively retreated from a very bearish posture.

And even though the gold timers' consensus from mid April turned out to be right, that in and of itself is not a good reason to expect it to be right this time around as well.

On the contrary, as contrarians might say, the consensus more often than not turns out to be wrong.