Oct 072015

Back in July I reported on some unusual figures in the US Office of the Comptroller of the Currency’s (OCC) quarterly report on precious metal derivatives. In the OCC’s 2015 Q1 report they had a figure of $75.62b, which was a huge increase from the 2014 Q4 figure of $22.42b. However, in the latest OCC report, the 2015 Q1 figure has been revised down to $26.94b. I have cut and pasted the two figures from the previous report (in green) and the new report (in red) below:


No doubt someone on the internet will claim this is a cover up, but looking deeper into the numbers it seems to be a case of a fat finger keying error. If we go back to the 2015 Q1 report it is clear that the main contributor to the $75.62b figure was Citibank, as can be seen from Table 9 below:


At the time Nick Laird of Sharelynx.com contacted the OCC and questioned them about the dramatic $42.048b and $10.970b figures (circled in blue) and the OCC confirmed that the figures were correct and provided the following screenshot of Citibank’s schedule RC-R as proof (relevant figures circled in blue):


I noted the large difference between Citibank’s centrally cleared derivative contracts figures versus their over-the-counter figures, saying it made “me wonder if the red figures are a fat finger and belong in a different row”. Looking at those figures in blue in light of the revised 2015 Q1 figure of $26.94b leads me to think the numbers are in the right row, just that the decimal place was keyed wrong by two points, that is, the figures were overstated by 100.

If we divide the centrally cleared figures by 100 and add them to the over-the-counter figures, we get revised Citibank white precious metal derivatives totals of:

Maturity < 1yr = 3,083,000 + (38,965,000 / 100) = 3,472,650
Maturity 1-5yrs = 703,000 + (10,267,000 / 100) = 805,670

If you plug these figures into Table 9, the Precious Metals All Maturities figure reduces from $75.62b to $26.88b. That is very close to the OCC’s revised figure of $26.94b and the remaining difference is probably the usual sort of minor adjustments the OCC has made to its figures in the past. So it doesn’t look like there was any real increase in silver/platinum/palladium derivatives in Q1 and nor was Citibank cornering that market.

There is one other dramatic change in the OCC report – a 75% reduction  in commodity derivatives:


No revision was made to the Q1 figure and almost all of the Q1 increase, and Q2 decrease, was JP Morgan, so we assume the +$4 trillion figure stands. Very unusual.

Jul 022015

On Tuesday I posted on the latest US Office of the Comptroller of the Currency’s (OCC) quarterly report and how it showed a reduction in total precious metal derivatives. I took Zero Hedge to task for jumping the gun on posting before seeking further information but now I have to admit I was guilty of that myself, the result being I was wrong to claim of a 29% reduction in precious metal derivatives.

Zero Hedge and I both made the same mistake, which was to assume that “precious metals” in Table 9 of the OCC’s report included gold, which I think is a reasonable assumption as “precious metals” is commonly understood to mean gold, silver, platinum and palladium. The OCC confirmed to me by email that Table 9 excludes gold and gave me a snapshot of a section of the call report where, at least to the banks filling out the form, it is clearly identified that the “precious metal” category should exclude gold.


The OCC did agree to my suggestion to footnote Table 9 to clarify that the precious metals notional does not include gold but did not answer my question as to why gold was now included with FX, simply stating that “banks began reporting combined FX and Gold notionals because of changes to the call report, specifically schedule RC-R that require the two be categorized together”.

Since the Perth Mint is not a bank and doesn’t deal in these sort of risky derivatives, I have no insight into the reason for such a bureaucratic form change. My best guess is that it has something to do with Basel II and changes to “harmonize the [US] agencies’ rules for calculating risk-weighted assets to enhance risk sensitivity and address weaknesses identified over recent years”. A look at this notice of proposed rulemaking (the new rules started in Q1 2015) shows that Gold and FX are grouped together with the same derivative contract conversion factors (see below from page 196) – note also that the column names match exactly the reporting categories in Tables 8, 9, 10 and 11 of the OCC derivatives report. So we can blame the Basel bureaucrats for decreased transparency in gold market.


A small mercy is that we still have a continuing dataset for the white PMs (ie silver, platinum and palladium). Zero Hedge noted the 237% increase in this category, which is certainly unusual for these markets which are smaller and often less liquid than gold. A comparison of the Q4 and Q1 OCC reports reveals that 92% of the $53.1 billion increase in the notional value of white PM derivatives was attributable to Citibank, who increased from $3.9b to $53.0b.

In an email response to Nick Laird of Sharelynx.com who questioned this dramatic change, the OCC confirmed the figures were correct and provided the following screenshot of Citibank’s schedule RC-R as proof.


This reveals that almost all of the increase was in Citibank’s centrally cleared derivative contracts (circled in red) rather than in OTC markets (in green), which total $3,786m, very close to Citibank’s Q4 figure of $3,901m. That does make me wonder if the red figures are a fat finger and belong in a different row.

Bloomberg note that “Citigroup has been expanding in derivatives as it rebuilds trading businesses that suffered after the financial crisis led the firm to take a government bailout” and that it has “pursued a risk-managed expansion off of a low base to bring the firm in line with competitors” so maybe Citibank has been active in the white metals as part of this “rebuilding”. If that is the case, then it is possible that they have been doing the same in gold.

This then leads to the question of whether we can estimate the value of gold derivatives for Q1 2015 based on historical relationships between the gold, the white PMs and FX figures. The chart below shows the notional value of gold derivatives against three other OCC report figures.


Gold has been declining in relative terms over the past 15 years, but the last few years are more stable. If we take the average of the last two years for each of these relationships, then the estimated gold derivative notional value in dollars and tonnes would be:

Gold as 0.045% of Total Derivatives = $91,360m (or 2,394 tonnes, red star)
Gold as 0.433% of Gold+FX = $137,491m (or 3,602 tonnes, green star)
Gold as 3.644 times White PMs = $275,531m (or 7,219 tonnes, orange star)

To put those estimate figures into context, I’ve shown them as stars on Nick’s gold derivatives in notional tonnes chart below.


I think the red and green are realistic probabilities but the orange 7,219 tonne figure, based on the large Citibank white PM’s Q1 figure, just does not seem possible given there was no commensurate response in the gold price. This leads me to think that Citibank has only been active in the white PMs and it is not correct to extrapolate this action out to gold.

As to Citibank having a notional $53 billion worth of white PMs derivatives on their books, industry sources tell me that “Citibank are not that active in PGMs so if the figure is real, it would have to be silver and it would have to be something very unusual.” I have an email in to my precious metals contact at Citibank but he is on leave at the moment so no response at the time of this post. I’m not expecting a comment as Citibank declined to comment on the OCC report to Bloomberg, but I’ll post any update/explanation if I get it.

Jun 302015

UPDATE 2nd July: Please see this post for a correction to some of the statements I made below, it is unlikely that total precious metal derivatives have declined.

Yes, you heard that right, the US Office of the Comptroller of the Currency’s (OCC) latest quarterly report show a reduction in total precious metal derivatives from $106,293 million to $75,620 million. This is at odds with Zero Hedge’s post first, ask questions later approach where they say that there was a “237% increase in the total amount of precious metals (which include gold) contracts in the quarter”.

Zero Hedge get their percentage from a graph from the report (see below) that does show a big jump. However, if you look more deeply into the Table 9 figures they also featured, it becomes clear that the chart is most likely a mistake.

In the picture below I have drawn lines from the tables to the chart, to show where the chart is getting its figures from. For Q1 2015 you can see with the green lines which Precious Metals figures from Table 9 their chartist has used. Note that in the Q1 2015 report this table refers to the total of gold AND the other precious metals.

For Q4 2014 you can see with the red lines that the chartist has used the figures from the column named “Precious Metals”, mistakenly thinking that because this column has the same name as those from the Q1 table, that the figures refer to the same thing. This is not correct, as the Q4 Table 9’s “Precious Metals” means the other white metals only.

OCC PM Derivatives

A true apples with apples comparison would add up both the gold and precious metals figures (see purple circles) from Q4 2014 . If you do this you get a total precious metal derivative notional amount of $106,293 million for Q4, which declined 29% to $75,620 million in Q1. Such a decline doesn’t make for much of a meme that the OCC is trying to hide a radical increase in gold derivatives.

Indeed, if you take the last twenty years or so of OCC gold derivative figures and eliminate the effect of changing metal prices to convert them into notional tonnes to see what has really been going on in the gold market, you get a chart like the one below.



You may be surprised to see this chart showing a very dramatic decrease in bank gold derivative activity, as the impression most gold commentators give is that gold speculative derivative activity by banks is huge and increasing. A fair part of that decrease is the result of miners reducing their hedging (see here for a comparison of the miner hedge book versus OCC derivatives over time).

It is unfortunate that the OCC has decided to lump gold in with the other precious metals as it means we will no longer be able to estimate the notional tonnage of gold derivatives with any accuracy – yet another blow to transparency in the gold market. But whatever the reason for the change, it certainly isn’t an attempt to hide any massive increase in gold derivatives.