Both gold and silver hit fresh price highs mid-week, continuing the three month rally that began at the end of May, but by week’s end gold ended lower by $7 (0.5%), while silver retained the bulk of its gains, ending the week a sharp 92 cents (5.3%) higher. As a result of the distinct outperformance of silver, the silver/gold price ratio tightened in by more than five full points to 83 to 1.  I can’t call the sudden outperformance of silver surprising; more surprising is in what took so long.

After lagging behind gold for most of this year, over the past few weeks silver has caught up with gold’s performance year-to-date. I’m still clueless as to what the very short term will mean for gold and silver prices, but despite the sharp recent tightening in the silver/gold price ratio, 83 to 1 still represents an undervaluation for silver that must be corrected sharply over the long term. It’s not so much that the ratio is now ten full points from where it was just weeks ago, it’s more what could be the possible legitimate reason it’s still as high as it is? That’s a trick question, as the only reason for silver’s extreme undervaluation – COMEX positioning – can hardly be considered legitimate.

While the price moves have been dramatic, even more dramatic have been other developments, both physical and derivatives-wise, particularly in silver. These developments, quite frankly, have come so fast and furious that it has been hard to keep them in proper perspective. In fact, the price moves look kind of puny compared to the flow of verifiable data.

In the space of two or three months, we’ve seen the establishment and dismantling of the largest concentrated long position in COMEX silver futures history, the physical inflow of more than 100 million oz of silver into a few silver ETFs (the most in history), and the sinking into a financial hole by the biggest shorts in COMEX gold and silver that has been much deeper than any in history. How the open loss situation of the big COMEX shorts gets resolved will be the biggest price factor ahead and that resolution is still unknown, although it is fairly simple to measure.

Price weakness in gold since Wednesday brought relief to the big shorts since that day, but not for the week. While the 7 biggest shorts on the COMEX were better off by $500 million from where they were on Wednesday, they still ended the week worse off by $200 million from where they stood on a combined gold and silver basis from last Friday. When the dust settled yesterday, the 7 big COMEX gold and silver shorts were $4.6 billion in the red on an open and unrealized basis or just over $650 million per trader on average. Of course, I’m excluding JPMorgan from these calculations, since it is impossible for the bank not to come out ahead no matter what prices might do.

Aside from the three historical developments over the past 2 to 3 months, there were quite a few surprising developments this week in a host of areas, including the new Commitments of Traders (COT) report.

The turnover or physical movement of metal either brought into or removed from the COMEX silver warehouses increased to 4.2 million oz this week, higher than the average over the past two months, but not quite up to the average over the past 8.5 years. Total COMEX silver inventories declined by 1.2 million oz to 311.9 million oz, which while still very close to the all-time high set a couple of weeks ago, was somewhat surprising in that inventories generally rise going into a major delivery month. No change in the JPMorgan COMEX warehouse for the 9th week.

There was an unusual transfer of more than 11 million oz of silver in two separate COMEX warehouses from the registered to eligible category, which was unusual because going into the first day of delivery of a traditional delivery month (Friday for the September contract), there are normally transfers from eligible to registered in order to make delivery. The most plausible explanation for the counterintuitive category move was that as the end of the month approached, it made sense for the holder of the silver in the registered category to switch to the eligible category which is cheaper to hold in terms of storage charges. I have little doubt that the holder is JPMorgan which, in addition to the 153.8 million silver oz it holds in its own COMEX warehouse, owns as many as 100 million oz in other COMEX warehouses, or 250 million oz (80% of the total COMEX inventories of 312 million oz).

The first two days of the September COMEX deliveries featured heavy deliveries in both silver (a traditional delivery month) as well as gold (a non-traditional delivery month). In silver, a total of 5481 contracts (27.4 million oz) have been delivered (or redelivered) over the first two delivery days, more than in any full delivery month this year. In COMEX gold deliveries, 1521 contracts have been delivered over the first two days, more than in any non-traditional full month this year.

The only real interest for me in COMEX deliveries is what JPMorgan’s role may be, particularly as concerns its house account. So far this month, JPMorgan has been the largest stopper of both silver and gold in its house account; stopping 1310 silver contracts or 24% of the total contracts issued and 579 gold contracts or 38% of the total gold contracts issued. JPMorgan has now stopped more silver deliveries in its house account in the first two delivery days of September than in any full delivery month since and including December. Combined with the unusual category transfers this week in the COMEX silver warehouses, the dominant stopping of deliveries by JPMorgan indicate it is continuing to acquire physical silver at every turn.

On the ETF front, “only” 2.7 million oz were deposited into the big silver ETF, SLV, this week, as well as 1.7 million oz into SIVR, the second largest silver ETF. Based upon my back of the envelope calculations, at least 10 million more ounces are “owed” to the SLV and possibly a lot more. Thus, this most recent delay in depositing metal is very different from the massive and timely flow of physical silver over the past couple of months and is suggestive of physical tightness. Yesterday’s month end, also brought an end to the two-week reporting period for stock short sales. While this report is extremely difficult to handicap, I wouldn’t be surprised to see a significant increase in the short position of SLV when the next report is released on Sep 11.

I’m still of a mind that most of the 100 million oz of physical silver that had come into the silver ETFs over the past two to three months came from JPMorgan (for the simple reason it accumulated and held all the silver available for more than 8 years running). My only question was whether JPM made that silver available via straight sales or by leasing it to other Authorized Participants who provided it to the ETFs.

It’s an important distinction because if JPMorgan leased the metal to the other AP’s, even though the metal was placed into the ETFs on an unencumbered basis as far as the trusts are concerned, the AP’s doing the borrowing are very much responsible for returning the metal to JPMorgan. While I still don’t know the answer for sure, this recent delay in deposits strikes me as if JPMorgan turned off the lending spigot or the AP’s decided to be short shares and owing metal to the trusts rather than in owing metal to JPM. Kind of like the choice of being in the short frying pan or the fire. At least in terms of timely metal deposits into the ETFs when due, I don’t sense the same situation in gold, where ETF deposits appear to be timely.

The positioning changes in this week’s COT report were more than interesting, particularly in gold, which was a bit of a shocker on a number of fronts, including what I was expecting (an attempt to say I was wrong by a wide margin). Instead of a big increase in managed money buying and commercial selling, there was a slight reduction in the total commercial net short position. I’m still scratching my head a bit since gold was up about $35 over the reporting week, making new six year highs and total open interest rose by more than 42,000 contracts.

In silver, I thought we would see a bigger increase in managed money buying/commercial selling than the 2000 contract increase in total open interest over the reporting week and despite me hoping I would be wrong due to lower numbers, the numbers were actually a bit larger than I expected.

In COMEX gold futures, the commercials reduced their total net short position by 2400 contracts to 333,800 contracts. Although it was a big surprise that the total commercial net short position didn’t rocket to a new historic extreme, that’s not to say that the position is still not very bearish on a conventional basis, because it is still just that. It goes without saying that if we do move substantially lower in gold, the one and only reason will be the still-outsized commercial short position.

That said, one of the standout features to this week’s COT report was a disproportionately larger reduction in the concentrated segment of the commercial short position. The 4 biggest gold shorts reduced their net short position by 9300 contracts and the next 4 largest shorts bought back another 3000 contracts. This concentrated short covering should make one sit up and take notice.

Another aspect quite noticeable was that this was one of the very few times (if ever) that the commercials, particularly the largest shorts, bought back contracts at higher prices and at a loss. Admittedly, there wasn’t a particularly large number of short contracts bought back and the losses incurred weren’t large, but when the mailman bites the dog, you can’t help but take notice. Clearly, the big commercial shorts didn’t appear inclined to add to short positions for some reason. Is this the start of a broader commercial short covering? It’s way too soon to tell. I have little doubt that JPMorgan joined in on the concentrated short covering.

If the commercials were net buyers over the reporting week, then it’s a safe bet to conclude that the managed money traders were likely net sellers, even though the overall numbers weren’t that significant. The managed money traders did sell roughly what the commercials bought on a net basis, but how they did so was also a surprise. The managed money traders sold 3130 net gold contracts, but that net selling included the purchase of 8573 new longs and the very surprising new sale of 11,703 short contracts. It was the new short sale of so many managed money contracts that was probably the biggest surprise in this week’s report.

On the one hand, the managed money short position was already near historic low extremes and not capable of being reduced sharply as I must have pointed out a gazillion times recently. But that’s altogether different from expecting a big increase in managed money short selling on new price highs. Best I can explain, it’s important to remember that while the vast majority of the traders in the managed money category are technical funds, not all managed money traders are technical funds.

The CFTC’s basis for inclusion in the managed money category is determined by having a money manager controlling the trading of a specific group of outside investors, not the methodology employed in deciding when to buy and sell. A managed money trader who decides trades based upon non-technical factors is still a managed money trader. My best guess for explaining the unusual increase in managed money short selling on a move to new price highs in gold is that a subjective decision was made that gold prices were too high and about to come down. Whether that turns out to be correct or not will be revealed in the fullness of time.

The resultant managed money net long position of 226,716 contracts (258,269 longs versus 31,553 shorts) is still very large and bearish in conventional historic terms. The fact that so many new managed money shorts were added this week is what enabled the commercials to buyback gold short contracts. Therefore it does cloud the picture as to whether the commercials went out of their way to buyback shorts this reporting week or were only being somewhat opportunistic in the face of unusual managed money shorting.

In COMEX silver futures there was no commercial buying on higher prices, as the commercials increased their total net short position by 10,800 contracts to 81,700 contracts. While still not exceeding the recent extreme in commercial shorting of 83,900 contracts on July 30, this week’s commercial short position must be considered historically bearish, although not as extremely bearish as gold’s commercial short position.

To be fair, the price of silver at the end of this reporting week is roughly two dollars higher than what it was on July 30, so a good case can be made that the commercial short position should be much larger now than it was back then. And while I am pegging JPMorgan’s short position to have increased this week by about 5000 contracts to around 20,000 contracts, I was pegging JPM’s short position on July 30 to be at least 25,000 contracts. So in essence, JPMorgan hasn’t really increased its silver short position from July 30 levels, even though silver rose by two dollars. Also, as you may recall, I revised my take on JPM’s short position to be lower than I was estimating when the August Bank Participation report was released. We’ll get the BPR for September next Friday, so let me hold off on firm declarations of JPM’s position until then.

Even at 20,000 contracts or thereabouts, JPMorgan’s net long position in silver is still roughly 750 million oz (850 million oz of physicals minus paper shorts of 100 million oz). With the recent turn up in silver prices to $18+, JPMorgan is now ahead on its net silver position for the first time since beginning to accumulate the metal in early 2011 (Not that I ever doubted that JPM would come out ahead). Plus, it still seems to be accumulating metal based upon COMEX delivery stats.

The managed money traders were net buyers in silver to the tune of 11,892 contracts, comprised of new longs in the amount of 6460 contracts and the buyback and covering of 5432 short contracts. The resultant managed money net long position of 55,788 contracts (84,643 longs versus 28,855 shorts) is firmly in the bearish camp, although by some measures not excessively so. Twice before, in 2016 and 2017, the managed money net long position was at or over 100,000 contracts, more than 40,000 net contracts more than it was on Tuesday. This leaves room for additional managed money buying of significant proportions (as well as room for significant selling on lower prices).

More to the point is that the managed money net long position as of Tuesday is still nearly 7000 contracts less than it was on July 30 and before silver prices climbed by nearly two dollars. While there has been significant managed money net buying of nearly 22,000 contracts over the past two reporting weeks, there was even more significant net selling of just over 28,600 contracts in the two previous reporting weeks (Aug 6 and Aug 13), even though silver was not down in price. I can’t say that the old wash, rinse and repeat cycle has been broken for good, but the old washing machine is making funny noises and seems to have had better days.

Taken with the other quite pronounced developments mentioned earlier (the COMEX Silver Whale, the unprecedented physical metal flows into the silver ETFs and the record open losses to the big COMEX gold and silver shorts), it seems the game may have changed. Whether it has or not should be obvious in price action up ahead. We still could go lower if the commercials can arrange to make a stand to save themselves or sharply higher if they can’t. No one knows which it will be and I continue to play it as if the big shorts will fail (but with some recently created buying power in case we go lower). However, if we do get a sharp selloff, I must believe that will be the last such selloff before we blast higher for good.

Exactly 40 years ago, on Labor Day weekend in 1979, Hurricane Dennis threatened Miami, where I lived, much like Dorian threatens Florida and the southeast US today. Dennis made a last-minute turn to the north, sparing Miami (much like I hope the same for Dorian today). It’s funny how certain events can trip off vivid memories. I was 32 years old and a broker at Drexel Burnham. I had yet to be bitten by the silver bug (that would come 5 or 6 years later), but was fairly active in silver at the time for clients. Izzy Friedman wasn’t my client but we were pretty close and he would call me if his broker was out.

Izzy had a fear of hurricanes worse than most and when one threatened Miami, he would hop on a plane to the first place his travel agent could arrange. On Labor Day weekend in 1979, he ended up in Las Vegas for the first time (on which he would recount later he was shocked at the utter disregard displayed for money in general). Izzy had bought silver several years earlier at around $4.  He took delivery of 4 COMEX contracts for which he paid about $80,000 in total, his largest investment ever. On Labor Day, 1979, silver had run to around $17, its highest price ever and up nearly three times in a year, which put Izzy’s investment in silver valued at close to $350,000, more than meaningful for a man with 4 children approaching college age. Naturally, Izzy was fearful of letting those large gains slip away.

I remember taking his call from Las Vegas that Friday and could tell he was very close to selling and booking big profits. Not having any input in the slightest in why Izzy bought the silver in the first place (it was all Izzy), I knew enough to keep my two cents to myself and just gave him the pricing and news info he asked for, but could sense he was close to cashing in. To make a long story short, he decided to hold and would end up selling four months later at $40, booking a net profit of close to $750,000, which basically set him up for life (he was a frugal man and that amount of money 40 years ago was quite substantial).

There’s no particular moral to this story, aside from the Labor Day hurricane walk down memory lane and the fact that silver was quite close to what it is price today, 40 years later. Back then, silver would run to $50 over the next few months, into January 1980. Nine years ago, on Labor Day 2010, silver was priced around $19.50, having been at $18 late in August, a week or so earlier. By May 1, 2011, silver had run to near $50 yet again. No one can tell you what the price of silver will be several months from today, certainly not me. But I can tell you that something that has occurred twice before in fairly recent times, can certainly occur again. Particularly when there’s a lot more reason for it to do so than previously.

Prayers and thoughts to all those in the path of Dorian, not the least of which are my family and friends in Florida.

(On a housekeeping note, I am switching to the December COMEX contract for silver closing price purposes, given that September is no longer the lead month. Thus silver joins gold in pricing off the December contract. This change adds about 16 cents to the price of spot silver, while the price of gold is now about $10 above spot)

Ted Butler

August 31, 2019

Silver – $18.48       (200 day ma – $15.41, 50 day ma – $16.36)

Gold – $1530           (200 day ma – $1332, 50 day ma – $1459)

Comments are closed.