Are Gold ETFs Now Being Seen In A New Light?

Peter F. Way

Behavioral analysis of the way market-makers hedge the risks their big-money fund clients require them to take gives us a continuing look at investor expectations for the price of the yellow metal.

The principal investor commitment to gold securities, apart from direct investments in the metal, is in over $60 billion of SPDR Gold Shares (GLD). Other ETFs include iShares Gold Trust (IAU), with $11 billion invested, ETFS Physical Swiss Gold Shares (SGOL), $2 billion, ProShares Ultra Gold (UGL), $0.3 billion, PowerShares DB Gold (DGL), $0.2 billion, and ProShares UltraShort Gold (GLL), $0.1 billion. UGL and GLL are leveraged by their structures to magnify daily price moves by 2x.

To keep our focus on the metal rather than corporate competitive actions, we are excluding ETFs of gold miners’ stocks, the largest investment in which is Market Vectors Gold Miners ETF (GDX), $7 billion.

Our behavioral analysis relies on the need of Market-Makers to protect themselves from adverse price movement risks as they respond to client orders to buy and sell securities while adjusting the positions of their investment portfolios. Long-established relationships between trading-desk individuals provide insights by MMs into the intentions and probable client price targets involved on specific issues.

What the MMs are willing to pay to hedge their perceived risks, and how those protections are structured, can be turned into very logical expected price ranges, ETF by ETF, on a day-by-day basis. This proprietary analysis has been performed without change in methodology for over a decade on over 2,000 widely-held and actively-traded stocks and ETFs.

Since the dominant ETF investments are in GLD, we will keep our focus there. Pictured below is a history of how those expectation ranges have trended on a once-a-week basis over the past two years. The data is drawn from daily evaluations made in “real time.”

(click to enlarge)(used with permission)

This is not a backward-look at actual price ranges. Instead, it is a history of forward-looking price expectation limits, ones worth MMs using part of their profit from a trade-spread fee to buy protection against its happening. There is real-money evidence here, not merely the assertion of an opinion.

What has some directional forecast value in these results is the balance between upside and downside price prospects. Those are easily visualized in the distances from the heavy contemporary-price dot to the upper and lower ends of each forecast range bar.

To provide a standard, single-value measure of that balance, we use the Range Index. It tells what percentage of the entire range lies below the market price at the time of the forecast. A small Range Index implies a large upside potential, and a large RI suggests an exposure to a large drawdown risk.

From the above picture, here is a trace of the GLD market price, accompanied by its Range Index.

(click to enlarge)

The GLD price is intended to track 1/10th of the London gold fixing. Its Range Index in this two years seems to have moved rather concurrently with the price. Since the price has cycled during the period, perhaps there is a useful forecast relationship.

Our surest way to check that out is to go back over a number of years and each day look at how actual prices subsequently changed over comparable periods of days after each forecast. We can then group like Range Index results together to see the average price change performance they produced.

Here is a recent compilation of that, where time periods are in 5-day column intervals. Range Indexes are arranged in rows cumulating from extremes of under 29 and over 49 to meet at a total in the 38 to 40 blue row at the middle.

(click to enlarge)The #BUYS column tells how many days out of the past 4½ years met the description to their left, except the blue row might better say “total” or “average”.

This table’s data is of price changes, expressed in compound annual rates of change to make comparisons between differences in time period columns easier.

What is fairly striking is that in over half the 4-5 years’ days, RIs were below 29, followed by annual rates of gain of +18% to +20% or better. The significance of the magenta numbers in the #BUYS column is that is where the RI is now, at 28.

Not shown is that in holding periods of 3 months or more (65 market days) 3 out of every four instances were gains in price above the forecast day. Shorter periods had gains in at least 2 out of every 3. Even that lesser ratio is at the upper end of the better performances achieved by hedge funds. Extending the holding period to six months or more, nine out of ten were winners.

The obvious remaining tell here is that holding GLD when its RI is above 40 is not a great idea. You likely can buy it back sometime in the next 3-6 months at a 3% to 5% discount, which is like earning at an annual rate of +15% to +25% on your money, very safely (in cash), in the interim.

If you would like to see what the MMs are currently thinking about GLD’s prospects, link to this address. At this writing they have a Range Index of 27 on it, which has happened (or better) 343 times in the past 1256 days (5 years). Those 343 experiences, when closed out either by reaching the top of their forecast range or the passage of 3 months holding (63 market days) produced gains at an annual rate of +32%. Winners made up 72% of those adventures. The worst price drawdowns in each case while being held averaged only -3.3%.


About abwehra group

The Art&Science of Trading Gold
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s