Silver & Gold Data Feast!

A rare treat for gold and silver data junkies with the CFTC's new COT report...

IN THIS SPECIAL EDITION of the Got Gold Report, we take a look at a new reporting format for some crucial gold market data, writes Gene Arensberg from Atlanta, Georgia for the Gold Newsletter.

US regulator the Commodities Futures Trading Commission (CFTC) recently "disaggregated" the data it reports in its weekly Commitments of Traders Report (COT) for several markets, including gold and silver. Basically, what the regulator has done is to further divide the reported data into five categories of traders instead of the previous three categories.

We COT mavens could easily get spoiled with this new reporting format. It's so new, in fact, we are not really sure yet how to analyze all the fresh data! But we are glad the CFTC has also given us about three years of back data, reported in the new format, so we can get used to crunching these new figures while continuing to study the "old" method alongside.

That time period includes some of the most interesting and traumatic events of the last decade for precious metals. So we view the new trader data sets as potentially very useful going forward.

For those who aren't familiar with even the old COT reporting format, the CFTC used to class reportable (meaning large) futures and options traders as either "commercial" or "non-commercial" – essentially taken to mean "industry" versus "speculative". (We'll leave the full descriptions to the CFTC here.) Whereas the new reporting plan further divides the reportable players into the following categories:

• Producer, Merchant, Processor, User: Examples include Gold Bullion producers, bullion banks, refiners, large bullion users;
• Swap Dealers: Swaps are used to hedge or lock in positions by trading different instruments (say, a Spot Gold purchase for a forward sale) at a set time in the future. Swap dealers pick up a fee for arranging the trade;
• Managed Money: Portfolio managers, hedge funds, institutional funds, etc;
• Other Reportables: Other large traders the CFTC doesn't include in the first three;
• Non-Reportable Positions: Small traders (i.e. private investors) or traders who are not required to report their positions.

Over the next few months we intend to get more familiar with the new reporting scheme here at the Got Gold Report. But we have been using the new data for about three months now, so we've already developed some theories which have apparent predictive value using the new format.

Until we get some more "mileage" with the new ideas, we intend to continue operating on the basis of the "old" methods for at least the next few months. In essence, in our published reports, we tend to focus on just one aspect of the old COT reporting format: The net positioning of the large commercial traders. As long-time readers know, over the years we have become comfortable with the signals of that one aspect of the COT report, especially when compared relative to the total open interest.

We usually call that the LCNS.TO, or the "Large commercials' net short relative to the total open interest" in Comex Silver or Gold Futures. The idea is basic and simple. In a nutshell, as the metal moves higher in price it becomes more attractive for the players on the commercial side (i.e. in the industry) to sell short or to hedge it, and vice versa. But very large changes to those veteran traders' aggregate positions from week to week often send timely, more short-term signals to us as speculators.

That is what we currently focus on at Got Gold Report. Here is what the silver version of relative commercial net short positioning looks like as a reminder and as a benchmark for the graphs to follow in this special report.

With the new CFTC reporting, there is even more data to crunch, with perhaps a bit more transparency. We'll be reporting more about that in the coming weeks and months, but for now we thought readers might be interested to see graphically just how much information is available in the new COT reports.

Below is just a very small sample of what an experienced trader might look at using just one series of the CFTC's new "disaggregated" COT reports. Please note that all of the charts below are from just one market – silver.

Before we begin, just a quick word about the simple chart template. It is a standard two-axis comparison chart found in most spreadsheet software. We use the cash market closing data for silver in all the charts for comparison. We are using the CFTC data raw, with no adjustments. However, in some cases, net positions are calculated using the gross long and gross short positions with spreading bets (which cancel each other out in the "net" positioning) ignored.

First, let's compare the price of silver with the total Comex open interest. Remember each open contract is a promise to deliver 5,000 ounces of silver at a future date and price.

Can we see a correlation? Yes, clearly the open interest tends to rise as the price of silver rises and vice versa. We still don't have a good explanation for the big drop in open interest in April of 2008 although some analysts have tied it to the Lehman or Bear Stearns "take-unders".

Next, let's look at something in the new CFTC format. Let's examine the Producer/Merchant/Dealer/User (PMDU) net positioning relative to the silver price. Obviously this is the bulk of what used to be the "commercial" category, and it includes the largest of the largest hedgers and short sellers of futures today.

Can we see a correlation in this simple comparison? Well, very generally yes. When the PMDU net positioning is extremely net short (low on the graph means a greater net short position) we can agree that it is generally, but not always, a less interesting time to take new silver positions and vice versa.

Perhaps what is most notable are the exceptions, such as in November of 2007 among others. Be cautious; we barely have three years of data. But in this case it is similar to the old "commercial" category.

Next, using the new format we can look at the positioning of the new category called Swap Dealers. Removed mostly from the previous "commercial" category, Swap Dealers report long, short and spreading positions. Spreading positions are basically what they sound like – positions that offset each other. If we eliminate the spreading positions we can use what's left over long and short to arrive at a net position to see which side the Swap Dealers are exposed to in the silver market.

Here's what it looks like graphically:

For most of 2006 Swap Dealers held more exposure to silver on the short side, but notice that, as a group, they actually went slightly net long in exposure just as silver plumbed its lows in August of 2007.

Perhaps that is a "tell" we might want to file away for later consumption. Note also that once silver fell off Panic Cliff in 2008, the Swap Dealers tended to show a modestly net long exposure to silver, especially after September 2nd. And perhaps more importantly, notice that as silver recovered back up to the early 2008 trading range in the back half of 2009, the Swap Dealers had yet to take on any significant net short exposure.

That may or may not be predictive, but chart wonks like us find that interesting nonetheless!

Next, using the new data, we could look at the net positioning of the group the CFTC now calls "Managed Money". Once again we ignore the spread positions and focus on the net difference between long and short positions. What emerges is the net exposure of portfolio managers, hedge funds, and other very large players who have been mostly on the long side over the last 3 years. However, as with other large categories, it is always important to remember that these positions likely represent the collective actions of a number of traders acting through and taking advantage of the largest actors in the market (as opposed to these very large traders merely trading for their own accounts). It is highly unlikely that any one person acting for the very large traders actually controls their entire positioning.

Can we see a correlation here? Yes, several, but the most obvious is that the better times to buy silver have been when managed money has already been driven away from the long side – after being "run to cover" by the hedgers and short sellers if you will.

Perhaps not as apparent at first glance, see if you can find periods of very large, spiky changes to this indicator and the near-term effects of the changes playing out in the silver price. (Hint: Look both to the upside and the downside...)

Next we might examine the new category the CFTC calls "Other Reportables". These are traders that are large enough to be required to report positions, but don't fall specifically under the first two broad categories. We could look at the long positions, or the short positions or the spread position size in isolation if we wanted to, but today let's look at the net positioning with the spreads taken out.

We caution not to rely too much on this particular graph, because of its relative size to the others. Having said that, we find it interesting, if not predictive, that the net positioning of the Other Reportables remains well under its 2008 near-top congestion and it appears just lately to still be rising, even though silver has corrected harshly.

What about the net positioning of the "little guy", the traders too small to fall into any of the other categories? The CFTC lumps all the rest of the trading into a category called "Non-Reportable Positions" – just as it did before the COT changes – and basically imputes the positioning of smaller traders together after accounting for all the reported long and short positions.

The graph above shows the net positioning of the smallest traders on the Comex market for silver futures. Although we have personal experience which says that some of the non-reportable traders are indeed quite large, an overwhelming majority of this category usually trades less than 100 contracts at a clip. (Most non-reportables trade fewer.) So there is no reason we cannot use this category as a proxy for what small futures traders are up to.

The most striking feature of this graph is, of course, the panic-driven exodus from the silver market by small traders in 2008. To quantify the blue line during that period, on July 15, 2008 smaller traders showed a net long position of 21,866 contracts or about 15.2% of all contracts open. By November 11 they were only net long 8,338 contracts or 62% less nominally, and down to just 8.8% of all contracts open. (It proved a spectacular buying opportunity if ever there was one!)

Notice also that only just recently have small traders had net long positioning in the same general neighborhood as before the 2008 panic. In hindsight, perhaps it is remarkable that small traders actually returned to the silver futures battlefield inside one year, given the devastating carnage of 2008.

How about a different metric to look at in isolation? One can surely do that with the new CFTC data. For example, what if we wanted to see the number of spreading contracts held by Swap Dealers? Maybe someone has a theory that swap dealers tend to put on more spreads near turning points or maybe they do so when prices have gotten too high in their expert opinions.

No problem, we can chart that too:

A higher number of spreading contracts means that there are a larger number of offsetting positions IN FUTURES taken by the Swap Dealers. We have no idea what the Swap Dealers are buying futures to hedge, long or short, nor in which market (London's over-the-counter center, Japan's Tocom exchange, perhaps major-producer Gold Mining stocks even). We can't really even know the size of those off-bourse positions, and that same opacity goes for some of the other large traders too.

But isn't it interesting that with three or four glaring exceptions, most of the spikes higher in the size of the Swap Dealer spreading contracts presaged at least a significant dip, if not an outright plunge, for silver prices? Maybe that's another "tell" we need to file away for future utilization.

These simple comparison graphs each tell a story to an experienced trader. Although there are probably dozens of other ways to look at and compare the data in the new COT reporting format, including comparing the various individual data categories to each other, let's look at just one more single metric today.

The graph below shows just the nominal short position of the "little guys", the outstanding number of bearish contracts held by Non-Reportable traders. One quick take-away is that we usually hear how the small trader always gets the shaft while the big guys hammer them. Well, take a good look at the graph below. See if you see what this author sees...

That's right, at least some of the small traders have been nicely on the short side ahead of all but a few of the big price plunges for the second most popular precious metal. And at least a few of them seem to uncannily cover their shorts just ahead of – or right after – major turning lows, too. That's fun to see graphically, but again it may or may not be predictive.

Even though the chart above is a weekly look (since the CFTC reports are published each Friday evening and cover the preceding week-ending Tuesday) there is a bit too much "noise" for some folks. So we added a 7-week moving average to the short position data to smooth it, and as a suggestion for others to watch.

Looking to Buy Gold at the best possible prices, and hold it at minimal cost with the utmost security? Go to BullionVault now...