In an earlier article we analyzed a metric that can be used to measure sentiment in the paper metals markets across the globe. For that analysis, we analyzed silver and found a significant differential in performance in the US markets versus the world markets. We will now use the same metric to analyze the gold market and see if gold also shows that same divergence.
Measuring World Gold Sentiment via the GLD ETF
One way to look at the gold market is to break it down into US trading and foreign trading (which we will refer to as "Overnight Trading" for the purpose of this article). The goal is to understand how gold performs during US hours and overnight hours, and see if there are any lessons or trading strategies we can take from this. We will use GLD as a proxy for the gold price because its data is very easy to retrieve/analyze, it trades during US market hours, and it is a fairly accurate proxy for the gold price.
To do this we will use data related to GLD's open and close prices, and then compare investor performances in three situations:
- Holding gold throughout the period - this is the passive buy and hold strategy.
- Selling gold at US open and buying it back at US close - this strategy exposes investors only to overnight gains/losses and avoids all US trading gains/losses.
- Selling gold at US close and buying it back at US open - this strategy exposes investors to only US gains/losses and avoids all overnight gains/losses.
So let us illustrate by analyzing the following three-day period:
As an investor if you employed the three strategies mentioned above with $1000, you would have the following results:
So for this brief period, gold rose by 2.4%, so investors employing a buy and hold strategy would have made a nice three day gain. Investors who chose to expose themselves only to the overnight market by selling GLD at the beginning of the day (at open) and buying it back at close would have made 1.09%, while investors who chose to sell at US close and buy back GLD at US open, would have made 1.3% over the three day period. These are the results we would expect from these three trading strategies - investors who hold gold for only half the period should expect to return only half the gains or losses.
Analyzing the Data
Let us now analyze the data and see how gold has performed on a year-to-date basis with each of these investment strategies.
GLD started the year at $162.02 (the 12/31/12 close) and ended the year-to-date period at $132.88 (5/21/13).
This is what we expect to see from each of these strategies - about half the losses of holding GLD throughout the year.
Silver and Gold Overnight Divergence
The strange thing is that when we analyzed silver's performance, we found a significant difference in the overnight performance of silver compared to its US performance.
Though gold's overnight performance was a little weaker than its US performance, silver shows a huge difference in its overnight performance compared to its US performance. Investors who bought silver at US open and sold at US close would have only experienced a 5% loss versus 26% for those who held it throughout the year. That is much more than we would expect and is showing us a significant weakness in the overnight markets on the part of silver, but not as significant in gold.
What Does this Information Tell Us?
We are really not sure the exact reasons for this significant divergence in the overnight markets between gold and silver. It may be related to larger players taking advantage of thinly traded silver markets to push the price down and then buying back during US hours. Or it may simply be related to Asian and European traders being bearish on silver, while US traders are bullish (or less bearish).
We think it's probably a combination of the two factors that are contributing to this divergence and feeding on one another. Large traders focused on dropping the silver price during overnight trading (remember it is a much smaller market than gold), can infect other traders with bearish sentiment and thus cause them to also trade bearishly.
But on the gold side price performance is a little weaker in overnight hours, but it is nowhere near as bearish as in the silver markets. This may be due to much stronger physical demand for gold in the east than for silver.
As the table below shows, this weakness in silver did not begin this year but started in 2012.
Let us now give silver investors a little bit of optimism. The last time we saw such weakness (before this current stretch) in overnight silver versus US silver trading was in 2009 - right before the spectacular rise in silver culminating in $49 per ounce in 2011.
Conclusions for Investors
In terms of gold, it does not look like there is a significant difference in performance during overnight hours - a little weaker but this could just be statistical noise. Investors in GLD and PHYS will not be able to see a significant advantage in utilizing any of the strategies mentioned above.
In terms of silver, there is a significant difference in the overnight performance compared to the US performance. Traders in SLV and PSLV may want to use this to their advantage, though we would caution that investors should not try to trade these performance differentials because they could lose a lot of money in the case of sentiment shift. In addition, unless traders are employing large amounts of money, trading fees could quickly reduce any gains to losses or amplify losses.
One thing that investors can do is monitor the overnight sentiment in the silver markets. It may be that once overnight sentiment changes that we start seeing a silver price recovery. Any investors with thoughts on the reasons for this divergence between gold and silver in overnight markets feel free to post them in the comments.