LBMA Trade Data: Turnover Figures for the First Quarter

By Rhona O’Connell, Head of Market Analysis, EMEA & Asia, StoneX Financial Ltd | April 8, 2021

 

Key points:

• Spot and forward gold tonnage over the first quarter of this year was just 8% that of silver in tonnage terms, but more than five times the amount in dollar terms.
• Platinum spot and forwards averaged twice the volume of palladium despite the palladium supply/demand balance being ~35% larger than platinum– underlying frequent anecdotal evidence of palladium’s relatively low liquidity.
• High volumes once again tend to (but don’t always) presage a change in trend, or at least a correction.
• Volume numbers underscore markets’ internal fundamental characteristics.
• Miners almost certainly hedged into the silver spike.

As we discussed in an LBMA webinar in late February this year, an analysis of LBMA turnover figures and setting them against price action and market forces can be very informative. In all four metals, it proved possible, when looking at the figures for 2020, to identify at least nine occasions for each metal when a spike in volume was a precursor to either a change in direction or at least a correction. This has always been felt to be likely to be the case, but it is helpful to see that the data back up the theory – and of course, it makes sense because as sentiment changes, some positions will be closed, and others will be opened.

This analysis of the first quarter of 2021 can corroborate that 2020 hypothesis, but we have gone one stage further this time and also looked at forward volumes along with options and loan/lease/deposit volumes. Looking at the other elements also gives a couple of nice illustrations of how the metals’ fundamentals tie in with trading activity.

Silver spike:

Obviously, the massive spike in silver over the turn of January/February is the stand-out activity for the quarter, but because it is such an outlier, we have also looked at a couple of gold and silver events that, without that silver spike, would have been the highest volume days.

We can’t ignore that spike, however, and in fact, it does throw up confirmation of one particular theory, of which more later in this analysis. A quick recap: this move was initiated by a concerted effort (via options on the iShares Silver Trust) from a retail investor forum that prompted a near-stampede into the metal, including a claim, in a headline of a media post, that the was massively short and the price therefore way undervalued. This was not the case, and there seemed to have been a misunderstanding, not realizing that shorts on COMEX are frequently the other side of a long physical position, financing that physical position by locking in the contango. Either way, silver’s notorious volatility soon took hold and the price shot from just under $25 on 28th January to touch an intraday high of $30.10 on 2nd February, a 21% gain in three trading days.

Those days of heavy volume did indeed presage a change in trend, resulting in silver’s previous mild bull market (which started in late November) being wiped out and starting a bear market that was just about drawing to a close two months later.

There is an important element here which makes this move slightly different from many of the others. Most changes in trend come about because of market fundamentals or financial forces. This one came about because of silver’s spectacular ability to spike up very hard and then retreat even harder, thus scaring off potential investors who became very nervous of being whipsawed or worse. Before we move onto more rigorous analysis of the numbers for the quarter, the important, interesting point to raise about this move is what happened in the forwards. Now it is axiomatic that when heavy volumes go into ETPs (silver ETP holdings rose by 1,237t (equivalent to two weeks’ global industrial demand) on 29th January-2nd February – and carried on rising for a few days afterward), the bullion market will reflect those volumes as the banks have to create the shares for the ETPs. This hits the spot volumes but wouldn’t necessarily be reflected in the forward volumes, but on those three days, the proportion of spot volume to forward volumes contracted very sharply, as forwards ballooned. The average spot/forward volume ratio for silver over the quarter was 2.23:1; in those three days, it was just 2.07:1, and on 2nd February, it was as low as 1.78:1.

What this suggests is as follows:

• Possible hedging of inventory by end-users – but this looks unlikely as the forward curve was steepening at that stage
• Likely hedging from the miners, locking in high prices. Primary silver miners are a distinct possibility here, but it is as likely, if not more so, that base metal producers were locking in strong by- or co-product credits from the spike in silver.

Now we turn to the remainder of the market, stripping out those silver moves.

General:

With the exception of the silver spike, with gold getting involved in the feeding frenzy, there is no clear pattern of dates in which unusual gold and silver volumes coincided in any of the categories that we looked at. What is interesting, though, is that in terms of quarterly average tonnage, gold spot was 7.7% of silver spot; swap/forwards were 7.8%, and total volume was 8.0% of silver. Options and loan/lease/deposit were both 13.4% of silver tonnage. Once prices are taken into account, of course, the picture changes dramatically, with gold total turnover 5.4 times that of silver.

Similarly, there is no coincidence in dates with respect to platinum and palladium. In terms of quarterly average tonnage, platinum spot was 2.23 times that of palladium and forwards, 1.98 times those of palladium. Palladium, however, at roughly 9.6 and 9.8M ounces per annum of supply and demand, respectively, is between 33 and 39% larger than the platinum market. Palladium has something of a reputation for thin liquidity, and these numbers bear this out. The activity in platinum leasing, which is more than three times that of palladium, reflects the nature of platinum’s industrial fundamentals, outlined below.

Individual metals (spot volume except where stipulated – charts show total volume)

GOLD:

Second highest volume day on March 3; equivalent to 37% higher than the quarterly average.

This was part of a period of heavy volumes (24th February-4th March with a daily average of 24.8M ounces), and volume was 28.3M for the day in question, against an average for the quarter of 20.7M ounces, under increasingly heavy liquidation having succumbed to pressure from the 20-day moving average and also breaking through the 10-day. This coincided with heavy liquidation from the ETFs for a daily average over that period of 9.0t, against an average for the quarter of 3.4tpd.

The underlying trading force was the slide towards the psychologically significant $1,700/ounce level; in the background, the relentless strengthening in the U.S. ten-year bond yields would certainly have helped, while more chatter about cryptocurrencies might have flushed out some weak holders also.

Gold forwards: 31st March at 22.4M ounces, 138% higher than the quarterly average.

This may have been some quarter-end book-squaring (although no similar volume surge appeared to occur during 2020); probably more significant here was a sizable jump in bond yields right across the U.S. curve, which may have encouraged market operators to look for contango.

The price trend changed two days later.

SILVER:

Highest volume excluding the spike on Feb. 26; if 28th January to 2nd February is stripped out, 26th February turnover was 55% higher than the average.

This was largely piggy-backing on gold as the latter veered lower towards $1,700 as investors moved away from gold and silver in the face of rising bond rates; as with gold, there were relatively heavy redemptions in the silver ETPs at this stage also. Silver would, without a doubt, have had stops elected after failing a test of $28 and then breaking through the 10-day, 20-day and 50-day moving averages.

PLATINUM:

Heaviest volume on Feb. 16; 138% higher than the average

This was a turning point, as a heavy sell-off developed after platinum failed at $1,350, at the peak of a strong rally that started in early November 2020 for a 58% gain from $830 to $1,341, a $501 rally, itself following a six-year bear market. Loss of momentum aided by a rebound in the dollar and the associated continued resurgence in U.S. yields, as well as overbought considerations with the Relative Strength Index clearing 70, were all contributory factors. The rally had been driven by investors looking through the currently weakened auto market towards tighter emissions limits in China, real-road conditions testing in India and the likelihood, after many years in the wings, of substantial uptake in platinum for fuel cell-powered vehicles, albeit not gaining much traction until the end of this decade.

Meanwhile, the loan/lease/deposit ratio for platinum in the first quarter was the lowest of the four at 11:1, with the others 15 and upwards. This can be explained by the fact that so much of the platinum market involves in-process material, namely for the production of glass (LCD and fiberglass), in the chemical industry (e.g., nitric acid and the formation of silicones), and in the oil industry with petrochemical reforming and isomerization. These are industries in which platinum is used as a catalyst or, in the glass industry, as vessels in which the glass is formed, and so the material is constantly being reprocessed, or spent catalysts are re-refined; consequently, there is a lot of borrowing and leasing in order to keep operations running smoothly.

PALLADIUM:

Heaviest volume on Jan. 8; 300% higher than the average

This heralded a small correction, but the overall downtrend persisted through to the end of the month. Here too, this volume generation was largely technical rather than specific fundamental background news or developments, although growing concerns over the supply issues with semiconductors into the auto sector would have been an underlying contributory fundamental factor. The move saw falls through the three major moving averages (10, 20 and 50-day) on the third day, so here, as with silver’s fall of late February, volumes would have been inflated as a result. This was the third day of a four-day bear phase before a correction and then a resumption of the trend.

While this volume was technically driven and therefore would have contributed to a relatively wide trading range on the day, volume numbers can also (and frequently do) work in the opposite direction. For example, palladium’s range on 29th January was $164 or 7% as compared with $99 or 4% on the 8th, but trading on the 29th was only 28% of the volume. So wide ranges should not always be conflated with high volumes unless there are technical and momentum-driven elements in play also.

Click here to receive monthly Market Analysis updates by email.

The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.

Other Topics

Commodities Gold

LBMA Trade Data

Discover the latest and most comprehensive analysis of the OTC Precious Metals Market available in our monthly and quarterly articles. Written by leading analyst, traders and market researchers, the market analysis uses LBMA Trade Data to provide never before seen insights into the latest trends and patterns occurring within the OTC markets.

Read LBMA Trade Data's Bio