The Gold Economy

By Brad Zigler

Six trading days and counting. That’s how long the SPDRs Gold Shares Trust (NYSE Arca: GLD) has been overbought.

Not that this condition hasn’t cropped up before. Since its debut in 2004, trust shares have tipped the needle into the overbought red zone 18 times. (Make that 19, counting the current market.)

So, what does “overbought” mean?

This condition—usually a precedent to a price pullback—arises when a security’s (or commodity’s) cost rises to such a degree that an oscillator moves above its upper band.

“Okay,” you say, “So what’s an oscillator?”

An oscillator is a momentum indicator used to identify short-term price extremes. For example, the Relative Strength Index (RSI) is one such oscillator that compares the magnitude of a security’s recent gains to the magnitude of its recent losses, then converts the result into a number ranging from 0 to 100.

When an RSI reads 70 or above, that indicates a security is “overbought”; with an RSI at 30 or below, a security would be considered “oversold.”

Notice that the price action of the SPDR Gold Trust—like gold itself—has pushed its RSI above 70 for the past six trading days.

Now, oscillators like the RSI tend to be poor trade triggers in markets that are strongly trending. They can show a market as either overbought or oversold, even while it continues to trend. But for GLD—and bullion—there’s a strong link between an oversold RSI and near-term sell-offs.

Here’s a compendium of RSI-indicated overbought periods for the SPDR Gold Trust and the price consequence one week later:

Overbought Periods For SPDR Gold Trust (GLD)

Over the past five years (more than 1,250 trading days), GLD was overbought for just 85 days, not counting the current run. The average time spent in this state was five days, so today’s excursion is already “above average.”

There’s a direct relationship between the duration of GLD’s overbought status and the likelihood of a subsequent sell-off. Generally speaking, the longer the period, the greater the resulting price decline.

Here’s why …

Of the 18 times GLD has been overbought, 13 of those, or 72 percent, resulted in lower prices a week out. That’s a statistically significant percentage.

Seven of these were one-day incidents resulting in a mean decline of 0.4 percent. Basically, this means one-day trips into overbought land aren’t very predictive of sell-offs—at least significant sell-offs within a week of the overbought period.

But multiday visits are different. There’ve been 11 incidents where overbought conditions persisted for more than one day. The mean price dip a week after these conditions were resolved was 3.0 percent. Measured on a time-weighted basis—that is, accounting for the duration of each overbought period—the average decline was 4.1 percent.

The fact that the time-weighted average decline exceeded the unweighted mean points to a link between the duration of the overbought status and ensuing price breaks.

This relationship is confirmed when you filter out the two-day periods, leaving only the longer lingerings in overbought status. For excursions longer than two days, the time-weighted average sell-off was 4.5 percent.

So, what can you take away from this?

Well, GLD simply being overbought—even for more than two days—isn’t necessarily a tripwire for a short position if you’re bearish or for a buy-in if you’re bullish. If you’re technically oriented, you’d need to look for confirming signs in other momentum oscillators, or a test of the bullish price trend line.

But, once such a confirming signal is generated, the odds favor the trader or investor shopping for a lower GLD price.

There’s a break in GLD’s—and bullion’s—price due soon. We’re six days into the current overbought status. If you figure that the longest overbought period on record for GLD is 17 trading days, we ought to be looking for a sell-off within the next 11 trading days or so.

You are looking, aren’t you?