Right after I last updated the gold enthusiasm index last July, gold went near parabolic and peaked in about two months. In that post, I wrote that enthusiasm for gold continued to wane, implying from a contrarian view that gold was not caught in a bubble and certainly had not peaked. If I had updated the "gold enthusiasm index" during the run-up, I would have finally observed a sudden ignition of gold enthusiasm. That surge would have perhaps flagged a potential topping pattern (even if temporary given the strong longer-term fundamentals for gold). As it is, I am reporting many months after the fact.
The upshot is that it seems the index can accomplish what I want it to do: signal when interest in gold is outpacing the price gains in gold. I assume that periods where sentimental frenzy runs too far ahead of gold's price tend to generate an unsustainable surge in price. Here is a review of gold sentiment.
As a reminder, I use Google Trends on "buy gold" and "sell gold" as a proxy for interest in buying and selling gold respectively. Google Trends provides a relative index indicating search volume and quantity of news articles using the selected terms. The chart below clearly shows a spike in interest in BOTH "buy gold" and "sell gold."
click to enlarge images
In August, search and news interest in "buy gold" surges to "all-time" highs
Source: Google Trends
At the peak in interest, Google highlights the following article: "Soaring prices: Is it a good time to buy gold?" from NDTV.com on August 11, 2011. On that day, gold completed a near-parabolic rise that was followed by two days of heavy selling, six days of heavy buying, and then another three days of selling that erased all the gains from the previous six days. On September 7th, the manic volatility finally took gold to its final (intraday) all-time highs before another swift and heavy sell-off confirmed the peak. By the time of the September lows, interest in "buy gold" had also returned to the top of the previous "normal" range. News articles followed the same pattern.
Interestingly, Australia led all regions in "buy gold", followed by Singapore and then the United States in third. The United States ranked first in "sell gold." Phoenix, Arizona led all U.S. cities in "buy gold" and "sell gold." I will be watching regional dynamics more closely in the months (and years) to come to determine whether regional differences provide additional insight into gold sentiment.
Since the Chinese are large buyers of gold, I also looked at Google searches in Chinese. This is a poor proxy for Chinese interest given Google is not the primary search engine, but I have to use it for consistency with the rest of my metrics. I used Google translate to convert "buy gold" into 买黄金 and "sell gold" into 出售黄金. "Buy gold" does experience a spike in August, 2011 but it is no bigger than a similar spike in February and much smaller than a spike in late 2009. In other words, the pattern for searches in Chinese differ from English searches are not as well-correlated to the price of gold. Note that China and Hong Kong represent what looks like 99% of all related search activity.
"Buy gold" interest in Chinese was highest in late 2009
Source: Google Trends
As a reminder, the gold enthusiasm index is a stylized metric that is simply the ratio of the relative Google search volume index for "buy gold" and the price of SPDR Gold Trust (GLD). The ratio can surge if search volume on "buy gold" surges much faster than GLD's price gains or if GLD drops in price much faster than search volume drops. Either way, I assume such a divergence signals buying interest is "too high" relative to the price of GLD. When gold is rising, I expect a correction in gold prices. When gold is dropping, I expect a correction in buying interest (that should eventually lead to a bottom in gold).
The gold enthusiasm experienced a sharp run-up into a local peak on August 8, 2011. HOWEVER, by the time of the final peak in gold, the enthusiasm index had already dropped back to new lows. While it has had several mini-spikes since, the overall trend has returned to a continual decline in enthusiasm for gold.
Gold enthusiasm surges ahead of September's peak in gold but is now back to the on-gong downward trend
Source: Google Trends and FreeStockCharts.com
I gathered two key lessons from this exercise: 1) Check the gold enthusiasm index more frequently, especially during dramatic price swings in gold, and 2) consider the full context surrounding changes in the index as relative changes seem to be much more important than absolute ones.
It is entirely possible that the spike in "sell gold" was the more important signal. Unfortunately, the historical pattern for "sell gold" has only provided one other significant change, and that change still coincided with a similar surge in "buy gold." In the first quarter of 2008, "sell gold" spiked along with "buy gold" to new highs at the time. That quarter turned out to be the highs for gold for another 18 months and preceded gold's last collapse by six months. The gold enthusiasm index represents an on-going learning experience, and overall I remain encouraged by the results so far.
To be clear, I still believe the fundamentals for gold remain firmly in place. The intermittent spikes gold experiences are distractions and even noise on the way to much higher prices as long as monetary policies remain loose around the globe. I do not expect this gold enthusiasm index to identify some absolute and/or final top in gold's price, only temporary or short-term peaks. (The jury is even more tentative on using this index for identifying buyable bottoms). I imagine that gold's long-term, secular run will only see its largest run-up once faith in paper money truly takes a plunge, and, presumably, interest in "buy gold" surges like never before and at an unbelievably breath-taking pace.
My archives of previous articles on gold include discussions on trading strategies and additional observations on sentiment. For example, in "A Gold Boom Without the Euphoria," I summarize a Planet Money segment which reveals a Nevada gold mining town's surprisingly sober response to its gold-driven fortunes. I discuss the limitations of this sentiment analysis in "Still No Clear Excess Of Enthusiasm for Buying Gold" (October, 2010).
While I have recommended switching to a core position in gold miners like Goldcorp (GG) and using GLD to trade around that position, I have since decided to retain core positions in BOTH equities. I am using the Market Vectors Gold Miners ETF (GDX) and incremental purchases in GLD to trade around my core positions.
Be careful out there!