Feb 022016

Since my article on the LBMA Silver Price on Friday, more market participants have come out criticising the process:

  • Afshin Nabavi, MKS: “People are going to lose all faith in the fix if this keeps going.” (link)
  • Brad Yates, Elemetal: “When Thursday’s number came in, people initially thought CME would void it, it was so far out of line with the market. When they endorsed it and it became the official print, the benchmark immediately lost credibility. We had two clients shift business away from pricing on the fix to live pricing.” (link)
  • Simon Grenfell, Natixis: “The new silver price setting mechanism appears broken. It is clearly an issue that the regulator should be looking at.” (link)
  • Grzegorz Laskowski, KGHM: “The large discrepancy between the spot price and the fix is very alarming to us especially that it happened twice in a row. I think the LBMA needs to make every effort to explain why it happened and needs to help to develop a system that would help to avoid these kind of situations in the future.” (link)
  • Unnamed bullion banker: “People are too scared to change their orders in the middle of the process so it got stuck. In the old days, banks would step in and take positions in order to balance the process. No-one dares do that anymore, as then they have to answer to compliance etc.” (link)
  • Unnamed precious metals dealer: “The system is broken. In the old days, if it was out of line, someone would have bought the fix and then sold the futures. It’s a joke — that’s all I know.” (link)

The most damming comment comes from Ross Norman as he is respected enough to be included in LBMA’s oral history project ‘Voices of the London Bullion Market’, who noted that “ten times in the last six months the silver price has been ‘fixed’ outside the trading range of the spot price for that day which is nonsensical. … A benchmark or reference price it is not. … the so called LBMA silver price does not come close to reflecting reality – and it is clearly vulnerable to manipulation – it is therefore effectively invalid.”

It seems the LBMA is certainly feeling the heat, with Ronan Manly noting this recent change to the LBMA’s website:

Ross asks why the LBMA Silver Price oversight committee (on which the LBMA sits) has “not come forward and explain what is going wrong and what they are going to do about – it is after all their job”. However I think there isn’t much they will be able to do about it because, as Ross himself notes, the real problem “is that banks are increasingly unwilling or unable to place corresponding orders where they perceive a mis-pricing because of fears of being accused of abusing a situation and facing the wrath of the regulator or their compliance departments.”

To fix the Fix, bullion bank traders (whether they are direct participants or not of the fix process) have to be able to “buy the fix and sell the futures” when the fix gets swamped by sell orders (or vice versa). The problem is that banks have a wide range of clients who hold positions with them across spot, forwards, futures, options, ETFs and so on. It is therefore highly like that one of those clients would be the loser of any such activity (and others winners) and complain (as did the client Barclays’ trader Daniel Plunkett traded against) about it. Alternatively, the regulator may decide to investigate markets from time to time.

The problem is that when a regulator comes looking at trades after the fact they could construe manipulative intent when no such thought was going through the trader’s mind – who was just arbitraging a market imbalance – and the trader finds themselves fined £95,600 and banned from trading, as Plunkett was.

If you think that traders would not be worried about such an unfair claim against them happening, or that client complaints or random regulatory investigations it would be unlikely, you haven’t been reading enough Matt Levine, who, coincidentally on the day of the silver stuff up included the two following stories in his daily article:

  • Nav Sarao, who is facing a 380-year jail sentence, “may not have had a material, or even any, impact on the bout of equity market volatility in May 2010 that later became known as the flash crash, according to a draft research report by University of California, Santa Cruz and Stanford University professors” (link)
  • Tom Hayes, who is serving 11 year jail for LIBOR rigging, saying that he was “thrilled that the brokers can tonight return to their families and their lives” while also “bewildered that he is now in a situation where he has been convicted of conspiring with nobody” (link)

Now I’m not saying these guys are scapegoats, but as a trader these stories would not make you feel comfortable that you would be given the benefit of the doubt. So would you help keep the fix in line with other markets if it risked you going to jail? No you would not, so the traders sat/sit on their hands. No matter what the LBMA, CME or Reuters say, I can’t see traders keeping the fix in line with other markets unless they got a letter from the FCA and SEC saying they will not go to jail, and the chances of that happening are zero.

The only hope for the fix is if non-banks step up to act as market makers/arbitraguers, otherwise “it can only get worse” as Ross says. Bullion Vault’s article hints that some such trading did happen

“As soon as [the benchmark] was done, [futures market] was hit with arbitrage selling and traded down to 14.07 with 5,000 lots trading. Then bounced back to trade $14.30+.”

however given how far the fix was able to drop, and how the futures market dropped much less, such arbitraging was not enough (but highly profitable to whoever executed it). Possibly more non-bank traders will step into the market to take advantage of the fact that the banks are impotent.

The problem for the LBMA Silver Price as a business is that unless traders do step in, more and more clients will stop placing their orders on it, as some of Elemental’s clients have. As the two-way liquidity drains away the chance of imbalances and out of line fixings increases, which causes more to pull out and so on in a death spiral.

unable to place corresponding orders where they perceive a mis-pricing because of fears of being accused of abusing a situation and facing the wrath of the regulator or their compliance departments. … Since Mitsui the only non-bank amongst the price setters departed – (and therefore the least regulation-bound), we have only banks remaining in the benchmark setting process so it can only get worse.

If the alleged email from the CME to its clients below (as reported by Platts and called “nonsensical” by a banking source) is indicative of how this issue is being dealt with, then the future of the fix is fraught.

“The platform worked as it should, in fact perfectly. It’s as good as the orders the participants enter. If you are a client of a ‘participant’ I guess you should direct this question at them. If you want individual flexibility, become a participant.”

Oct 192015

Dave Fairtex, writing for Peak Prosperity, claims he has The Smoking Gun Proving Silver & Gold Manipulation. In identifying all of the 0.5% or greater one minute price spikes over the past 6 years I’d argue he has proven how infrequent it is.

Dave’s main result is that there were a total of 135 gold events and 869 silver events where the price moved up or down more than 0.5% within a minute, with 66% of the gold events, and 54% of the silver, being downward moves. While the data shows a bias to the downside, Dave doesn’t put any of the results into context, undermining I think what is a worthwhile analytical approach.

Dave’s data set includes every 1 minute period between 4pm-3am US Eastern Time from late 2009. That is an 11 hour trading window over 6 years. The number of 1 minute periods in the data set is therefore:

(52 weeks x 5 trading days) less 10 public holidays) x 11 hours x 60 minutes x 6 years  = 990,000 periods

A total of 135 gold events within 990,000 periods is 0.0136%, or 1 event per 7,333 periods. For silver it is 0.087%. That is not a lot of manipulation events and in the case of gold, Dave shows that oil, gas, wheat and corn had more events.

Dave’s other observation is that when you compare the aggregate dollar change over all manipulation events and express them as a percentage of price, gold and silver have much bigger price effects (-33% and -120%, respectively), compared to a few percent for the other contracts investigated.

However, I see two problems with this comparison. Firstly, the total dollar changes are just compared to current spot price. A $10 change on $1000 is not the same as a $10 change when the gold price was $1900. It would been more informative to have added all the percentage changes for the manipulation events so that apples were being compared to apples, in terms of price impact at the time.

Secondly, just as with the 990,000 period example, a 4% total change for Crude Oil may not be directly comparable to -33% for Gold as the normal volatility for gold and oil differ. Having a total dollar change up and down over the other  989,865 non-manipulation events would reveal the real impact these events had for the overall price change for gold and silver within the time period investigated. Given that the average gold price over the time period is well above $1150 I’d say that -33% may be overstating the case.

The other aspect of the analysis that I think can be queried is the choice of trading window, which is explained thus:

“Most goldbugs like to say that gold and silver suppression attacks occur in the “wee hours of the morning.”  Loosely translated, I take this to mean during non-US and non-London trading hours.  So that’s the time range I will use: 4pm-3am Eastern; from just after US market close through to the London market open.”

If it wasn’t for the fact that Comex and London do trade significant volumes, I’d say this was a US-centric “we are sleeping so it doesn’t matter” view. However, do not those same “goldbugs” proclaim that US and UK trading is all paper, compared to the “real” physical markets in Asia, using charts like this to prove their point:


Below is a table of the key gold trading hubs and their usual trading hours. The highlighted areas are when those countries are operating their trading exchanges.


To argue that 4pm-3am is “when activity is relatively lighter than usual” is to argue that the huge physical markets of China, India, Japan and Dubai don’t matter when it comes to gold and silver prices. By the way, Perth is in the same time zone as China, which means the 300-400 tonnes of gold we trade each year must also not matter.

From the table above you can see that all of the major Asian markets close down their overnight electronic platforms around 3pm New York. This is not coincidental, as that is when CME’s Globex is winding down. I would argue that a more realistic period during which the global precious metal markets are relatively illiquid would be from Globex close at 5pm (New York time) to 9pm when the SGE opens, and maybe 2:30am to 5:00am, between the China/Japan break and the setting of the AM London Price.

I have left a comment on the Peak Prosperity article asking Dave to re-run his program with a change to the trading window and accumulating percentage changes for all manipulation events as well as all 990,000 periods. Hopefully he will indulge. It would also be interesting to see the analysis run for the entire 24 hour window, to get an overall sense of potential manipulation events.

Sep 252015

Alex Stanczyk of the Physical Gold Fund has just posted a transcript of an interview with an executive at a Swiss refinery about the state of the market. It is well worth a read or listen to the podcast. Below are some quotes and my take on them.

“How difficult is it to source the metal you need today? … It is truly difficult. This is also reflected by the price. It is getting more and more expensive to get material out of the market, and also there is less liquidity in the physical precious metals market than there used to be in the past.”

I think the statement that the difficulties “is also reflected by the price” needs to be clarified as later on he says “the price does not reflect the realities at all.” This is not a contradiction! The first reference is to the premium, the second is regarding the spot (ie metal) price.

When acquiring physical, the professional end of the market settles trades by unallocated account debits and paying a small dollar premium for charges, freight etc, rather than buying a physical form on the spot market with dollars. The references to “price” and “expensive” were to that premium above metal. See this post if you want more detail on this aspect of the market.

It was a bit frustrating that the interviewer did not pick up on this reference and drill further – was this a premium on 99.50% 400oz bars, if so were bullion banks refusing/delaying redemptions from unallocated accounts, or was it for other forms or was he referring to loco premiums reflecting freight and funding costs?

I would have also liked to know to what extent that refinery’s feed stock comes from newly mined gold or do they rely more on scrap and 400oz bars. Each refinery has a different mix of source metal and contractual arrangements for supply that can affect their perception of tightness. The Perth Mint is primarily a newly mined supply refinery with scrap being a swing form of supply for us, so we have a strong base of supply.

“The other point is that nobody is interested in any physical delivery at the end. These products are all cash settled. People are happy just to use the spot market as a benchmark, and the product itself never ends up in the physical market.”

He is obviously talking about Western markets here, and he makes an important point and will tie in with my future posts on fractional reserve bullion banking. He goes on to note that this is a dangerous set up as if everybody wanted the physical it “would not be around”. However, this is not a risk as “it looks very much like people are very confident in general financial markets”. The question of course is how much physical reserve exists against bullion bank unallocated accounts versus how much of an increase in physical redemption activity occurs.

“As long as market participants are happy for cash settlements, this can go on forever.”

So true. This was in response to a question about a mismatch between the spot price, which has been low, versus the tightness in the physical market. This idea of a disconnect between paper and physical is an argument that the money in the futures market or other paper markets is somehow not legitimate, that their “view” on price is not valid. Yes markets are more financialised these days, get over it.

I guess some people think that if only futures markets could be banned and everyone had to trade physical, the price would magically shoot up. They forget that if you ban all forms of paper gold you ban paper longs. And in any case, any paper contract can be synthesized using physical and borrow/lend. They also seem unaware that the net position of paper trading is, by arbitrage, reflected into the physical market, and vice versa.

“The flows of metal end up in Asia. It is mainly China, also India, and to some extent the Middle East.”

Same here in Perth.

“Then there is price-sensitive scrap – very opportunistic – coming every now and then out of Asian countries; not China or India, but other countries in the area.”

This is not something you hear talked about a lot, but yes the Asians do actually sell gold but are very good at it. Having a location advantage to Asia compared to the Swiss refineries, we do pick up a lot of this business but as I mentioned above, it is a swing or volatile source of material for us due to its price sensitive nature.

“since the last move up, a lot of scrap has already come to the market, so if the price moves up again, I don’t know how much scrap will be around in order to compensate for the lower volumes coming from the mining industry.”

I note that at the Denver Gold Forum CPM Group saw that a decline in mine production from 2018 was baked into the cake given the lack of exploration. If our refinery guy is right about scrap then the next leg up in the gold price could be quite dramatic. It is hard to call the scrap market as really high prices may be the inducement to get women to look in the bottom of their drawers for that last bit of out of date jewellery, and you can be sure the cash for gold business will be promoting hard in that environment. In any case scrap has not been/will not be a major source of metal sufficient to dent bullish demand too much, when it comes.

“If you see in one of these products a paragraph that references the possibility of cash settlement, keep your hands off.”

Good advice, I don’t think investors really pay attention to the contractual terms of the products and read between the lines, and often it is about what is missing rather than what is there.

Aug 172015

Last week I explained why shortages occur and how they are generally caused by production capacity shortages. Often such shortages of retail coin products are spun by commentators into a shortage of raw gold or silver or a physical-paper disconnect and thus a sure sign that metal prices will rise.

Mike Shedlock, in his very direct style, says that “any time you see articles promoting the difference between physical gold and paper gold you are most likely reading a pile of crap”, referring to the fact that “one can get physical gold near spot rather easily” and giving the example of GoldMoney or BitGold.

Mike’s comments were in response to this article that argued that because “demand for physical metal is very high … yet, the prices of bullion in the futures market have consistently fallen during this entire period. The only possible explanation is manipulation.” Market manipulation is certainly a factor but claiming it is the only possible explanation is taking a limited view of the dynamics involved.

The authors start by claiming that “for four years the price of bullion has been falling in the futures market despite rising demand for possession of the physical metal”. As proof, they mention prior periods of coin shortages. While there have been periods of resurgent demand, the fact is that retail bullion dealers have had a tough time since gold prices peaked and many have scaled back the size of their business. One would not see bullion dealer bankruptcies like Tulving or Bullion Direct in a market with rising demand. I’d argue that scaling back of inventories as demand waned was probably a contributing factor in the recent shortages.

The authors also ignore the rest of the gold market. Certainly Chinese demand has provided a base but Indian demand was affected during the last four years with their Government’s raising of duty levels. We have also lost the demand that came from ETFs, which has swung from a net 1,407 tonnes in the 15 quarters from Q1 2008 to Q3 2011 to minus 712 tonnes in the subsequent 15 quarters, a change of over 2,100 tonnes in the supply/demand balance.

After going over Supply & Demand Curves 101 they then argue that it is “nonsensical” to argue that “the drop in precious metals prices unleashed a surge in global demand for coins” because “price is not a determinant of demand but of quantity demanded. A lower price does not shift the demand curve. Moreover, if demand increases, price goes up, not down.”

This view ignores the feedback nature that price has upon demand. Most certainly price shifts the demand curve – people adjust their personal demand curves over time and price is one input into that. The Perth Mint has a lot of clients who bought gold at around $300 an ounce and I’m sure in 1999 if asked they probably would have agreed they’d sell when it got to $1,000. However, many did not and held on because the trend upwards in the price made them revise their perceptions of what was a fair price.

The authors also assume that gold has a normal demand curve, that is, if price goes up then quantity demanded will go down. However, as Paul Mylchreest asks in this article, maybe gold is a Giffen good, where quantity demanded goes up when the price goes up. Ted Butler makes a similar observation in this article, noting that “managed money technical traders … buying on rising prices and selling on declining prices”. Looking at a graph of retail bar & coin demand as reported by the World Gold Council versus price seems to lend weight to the Giffen good argument.


Apart from two surges, quantity demanded seems to increase, and decrease, with price. If I take the data from the chart above and plot it as price versus quantity demanded, then you get the chart below.


The red line is a line of best fit, and it certainly doesn’t look like the demand curve the authors refer to. This is by no means a proof that gold is a Giffen good, but it certainly points to the fact that gold does not neatly fit into an Economics 101 view of the world and the reality of gold investor behaviour is more complex.

The authors then seek to argue that the only other explanation of a lower price, using Supply & Demand Curves 101 theory, is that supply must have increased. They then claim that “there are no reports of any such supply increasing developments. To the contrary, the lower prices of bullion have been causing reductions in mining output as falling prices make existing operations unprofitable.”

Regarding unprofitability, an analysis of the top mining companies’ financials by Adam Hamilton notes that “the gold-mining industry’s cost structure is far lower than that $1200 number often thrown around … on a cash-cost basis, they could weather an $800-gold anomaly for many quarters”. This may be why GFMS figures show no reduction in mine production, with Q2 2015 tonnage of 786.6 up 3.1% on Q2 2014, with the chart below showing that mine production has been rising all during the “four years the price of bullion has been falling”.


The authors then refer to futures market speculators and manipulations. Certainly, gold’s low liquidity allows for manipulative events, which can affect market sentiment in the short term. However this is not enough to supress the price over the long term, or four years, where market fundamentals are the main driver.

Ultimately, the authors’ problem is with speculators, who they see as “selling naked shorts [as] a way to artificially increase the supply of bullion in the futures market where price is determined.” They get so close to the actual nub of the problem but miss it when they say that “if purchasers of these shorts stood for delivery, the Comex would fail.”

The fact is that the longs are as naked as the shorts. Both are using all their available money to fund their margin requirements so as to maximise their leverage, so the shorts don’t have the metal but the long don’t have the money. That is why delivery rates have always been so low on Comex and there is nothing to indicate that this will change. You have gamblers on one side betting the price will go up versus gamblers betting it will go down. If there is any manipulation, one could argue that it is the policy of low interest rates which encourage such excessive speculative behaviour.

The authors conclude that a “rational speculator faced with strong demand for bullion and constrained supply would not short the market” but as I’ve shown above, demand has been weak and supply has been rising. That is not to say that the short speculators are entirely rational continuing with their trade, as many have noted the market is reaching extremes on a number of measures. The nature of markets these days seems to favour trend following rather than fundamentals, which is fine until the trend changes. One thing I can be sure of is that when it does you will not see any article complaining about speculators “buying naked longs”.

Aug 032015

Michael Pettis argues that a market dominated by speculators tends to be more volatile as it is sensitive to changes (in consensus) in the way news is interpreted. If gold is entirely a speculative market, as I argued in Friday’s post, then we should see high volatility. While gold is more volatile than many other assets and currencies, it is not as excessive as we would expect based on Pettis’ theory. Why is this? I think it is because it is difficult for gold speculators to converge on a consensus view.

Divergence of Convergences

While I agree that the game theory-ish Keynesian beauty contest means speculators will converge on a view of a market, gold is composed of a number of groups with fundamentally divergent frameworks through which they view gold. A US Hedge Fund has a completely different view of gold compared to an Indian farmer, for example. Furthermore, many of these groups do not have a game theory approach – if you mentioned to an Indian farmer whether he is considering what all gold investors think about how all gold investors view gold, he would say “what the?” If participants are not aware of beauty contest dynamics then convergence doesn’t occur.

Note that within each group there is a convergence or agreement about how to view gold and when to buy it, there are just divergences between each group. Hence we have lower volatility as each group’s differing view counterbalance. The above implies that a savvy investor needs to implement a modified beauty contest: identify groups unaware of such game centric thinking and estimate their view of gold then identify game aware groups and then work out their view of the unaware groups plus their view of what game aware groups think other game aware groups think. Confused? No wonder no one can agree on gold.

Divergence Globally

In the latest World Gold Council (WGC) Gold Investor report they say break down the “drivers of gold into key factors: currencies, inflation/deflation, interest rates, consumer spending and income growth, systemic and tail risks, and supply-side factors.” Plenty of areas for divergent views. In addition, the WGC then point out that gold is a global market, so each of those factors differ between countries. I would argue that even if every gold market participant was playing a Keynesian beauty contest there is no way that with such a wide variety of factors to consider, and difficulty for the participants to communicate/signal their view, could any convergence occur.

Divergence Over Time

Adding to the complexity of the gold market is the fact that its structure is changing. As an example, it used to be that Western retail investors would run away when gold dropped significantly, but on the big drops in 2013 and 2015 we saw big surges in buying. Another example – consider the following two charts from the WGC Gold Investor report.


As the make up of participants in the gold market changes, then their influence on gold changes as well. WGC also makes good points on how conventional (ie Western professional investor) views about the influence of the US dollar and real interest rates on gold may be wrong.

Zero Correlation

I think the result of the above may explain why gold has no correlation to many asset classes. In the chart below from the WGC the correlations in red are not (statistically) significantly different from zero.


Those assets with zero correlation to gold are basically equities and debt; the positive ones being commodities, and the big negative the dollar. It shows that there is no consensus about how gold should relate to equities and debt.

Current Market Situation

While I have been arguing that gold is basically a highly divergent market, at the moment I think there is a dangerous convergence on the idea that gold will continue to fall, certainly in the Western markets (as discussed here) but this would also affect sentiment in other markets. It seems to be primarily driven by narratives around US interest rate rises as signalled by the Fed. On that I’ll leave you with this from Pettis’ article, modified to suit the current gold market:

“In this market, you [sell] because you believe that everyone has agreed on the collective interpretation of government signaling. Anything that undermines the confidence you have in the collective interpretation must undermine your decision to [sell], and in fact because everyone is watching everyone else, at some point, this can become a collective decision to [buy].”

Jul 312015

I have written before on how gold is a pure epsilon asset and driven by narratives. This article by Michael Pettis takes a similar approach to China’s recent stock market problems but he makes a number of observations that apply to markets in general. I think these observations have application to gold in general as well as the current state of the gold market.

Michael notes that there are two types of players in markets – value investors and speculators. He says that markets dominated by one or the other type will generally behave differently.

A Market Dominated by Value Investors – Value investors base their decision to buy or sell on their interpretation of a piece of news. Because value investors generally vary widely in their investment strategies and interpretation, “small changes in the way news is interpreted or in market sentiment will have a limited impact on overall supply or demand” and thus prices.

A Market Dominated by Speculators – Speculators base their decision to buy or sell on their expectation of the collective interpretation of a piece of news. Because of the dynamics of the Keynesian beauty contest, in general speculator expectations “tends to converge very quickly” with the result being that “a market dominated by speculators is extremely sensitive to changes in the way news is interpreted or in market sentiment.”

Michael’s point is that where you have convergence, you will have higher volatility. Markets dominated by speculators will, as a result, generally be more volatile. A market dominated by value investors will generally have a wide range of views and be less volatile.

However, it is also possible for a convergence in investment strategies to occur (which he argues is happening in China today) making a value dominated market liable to explosions in volatility. It is also possible for a speculator dominated market to have uncertainty as such high levels that it undermines “the ability of speculators to agree collectively on how to interpret signals” resulting in a wide range of views and thus less volatility (but the potential for a big move once consensus returns).

Michael’s next point is worth a bulk quote:

“volatility can never be eliminated. Volatility in one variable can be suppressed, but only by increasing volatility in another variable or by suppressing it temporarily in exchange for a more disruptive adjustment at some point in the future. When it comes to monetary volatility, for example, whether it is exchange rate volatility or interest rate and money supply volatility, [or gold volatility?] central banks can famously choose to control the former in exchange for greater volatility in the latter, or to control the latter in exchange for greater volatility in the former. Regulators can never choose how much volatility they will permit, in other words. At best, they might choose the form of volatility they least prefer, and try to control it, but this is almost always a political choice and not an economic one. It is about deciding which economic group will bear the cost of volatility.”

The reason I found Michael’s article of interest is because I believe that gold is a market dominated by speculators. By that I don’t just mean evil Comex shorts – all gold holders are speculators. This is what I mean when I say that gold is a pure epsilon asset. That is not to say that we are all gamblers, betting on whether gold’s price will go up or down. In Michael’s conception “speculator” means one who is looking at consensus of what the market thinks, what sentiment is.

Don’t think you are a speculator? Then that means you must be a value investor. But to apply the concept of value investor to gold is to argue that gold has an objective or fundamental value. I don’t see how that is possible for an asset with such a large overhang of stock compared to annual flow, where the withholding of supply by existing holders matters most (as I argue here).

I don’t see how that is possible for an asset that is not productive, that is, does not earn an income. Even for that rare group who can lend physical gold, it is not the gold that is productive, it is the use to which it is put that helps to determine its interest rate (you can’t value a dollar by discounting the cash flows of its interest).

While value investors may disagree between themselves on what a company’s future ongoing earnings will be, they all agree that the method of arriving at “value” is to discount those earnings. But for gold no such discounting is possible. All these “fair value” models of gold, when you look at them are formulas based on correlations to other assets or macro economic variables like interest rates, inflation, etc (eg here and here). No value investor values Apple with a formula based purely on relationships to macro economic variables, simplistically they estimate current and future phone sales etc and the resulting profit.

Understanding whether gold is a value or speculative dominated market is crucial for the rest of my argument, so I’ll leave it there for now as I’m sure I’ll get some disagreement about gold being purely speculative/narrative driven and it might be best to let that discussion play out. If you think gold has a fundamental value, then please tell us what it is and the logic of your calculation. I’m ready to be convinced. On Monday I’ll continue with applying Michael’s ideas to gold and what it means for where we are now.

Jul 292015

Yesterday Chris Powell of GATA criticised an article by Clif Droke on market manipulation. One point caught my eye, where Chris identified “an ‘ipse dixit’, an assertion made without authority” that Clif made, namely that “the market for gold is immensely huge and virtually impossible for any one entity to control its price swings … Even a coterie of interests devoted to pushing gold prices lower would meet with certain failure due to the enormous size and complexity of the market.”

It is one thing for Clif to claim that one entity could not control the gold market, but it strikes me as quite bold to claim a “devoted coterie” could not do it. To assess Clif’s claim we need factual examples of gold market liquidity so that we can “assert with authority” and solve this ipse dixit problem. Being the gold nerd that I am, over the past few years I have accumulated a number of statements about actual gold market liquidity (primarily because I’ve been annoyed with trite statements about how gold is “highly liquid” without any quantification) and thankfully now I have a use for them.

My simplistic understanding of “liquidity” is how much gold one can trade without affecting the market price. Below are some quotes that refer to actual trade sizes by people in the industry and their effect on price, both approximate and quantified.

  • Me, Feb 2012: “One could execute up to a tonne in one lot with a bullion bank at spot. … If someone was silly enough to ask a bullion bank to commit to a price for a 10 tonne lot, then from what I have been able to establish, the bank would adjust their spot price by around $10.”
  • Edel Tully, Nov 2012: “Brazil’s holdings expanded 17.2 tons last month … This is a chunky purchase … and was one of the key factors that gave prices a reasonable floor last month”
  • Mark Dow, June 2013: “Buying $300mm [7 tonnes] of gold can move the market up by less than a percent [$12 an ounce] in a normal market. But when sentiment for gold turns adverse, selling $300mm can drive it down, say, 2-3%.”
  • CNBC, Oct 2013: “Gold lost $25 in two minutes … a single sell order could be the culprit. It appears to have been an order to sell 5,000 gold futures contracts [15 tonnes] at market, Eric Hunsader of Nanex told CNBC.com”
  • ICE Benchmark Administration, Gold Auction Specifications: “Maximum Order Size 100,000 oz [3 tonnes] … This is a simple fat finger test designed to prevent the accidental input of large orders. In the event a user wishes to place an order greater than the maximum size they have the option to break the order down into smaller chunks”
  • Reuters, May 2015: “each interbank call would ‘shift orders of 50,000 ounces [1.5 tonnes] a time’, providing significant liquidity flows … Customers with smaller volumes are probably getting tighter spreads and better prices … but when somebody has size to do, especially with algos out there, it can hit pockets of illiquidity and cause spikes or flash crashes”
  • FT.com, May 2015: “if you want to transact in a decent size, which used to be 100,000 [3 tonnes] to 200,000 [6 tonnes] ounces. That has become harder to get away with without influencing the price unduly”
  • Me, Jul 2015: Gold drops $48 on 22 tonnes.

My takeaway from the above is that one can deal a couple of tonnes without affecting the gold price much – a bullion bank may widen the spread by tens of cents. However, once you get above 5 tonnes you will affect the price. From the quotes we get these figures:

  • $12 with 7t = $1.70 per tonne
  • $10 with 10t = $1.00 per tonne
  • $25 with 15t = $1.66 per tonne
  • $48 with 22t = $2.18 per tonne

What is 20 tonnes worth? At $1,100 per ounce it equals $700 million. Yes that is a lot, but there is a lot of money out there – for example, see this post where I calculated that if just “four groups decided to use only 2% of the $281.7 billion they manage to short gold. That would total 145 tonnes of gold”. Or consider what Ray Dalio said in this Sep 2012 interview “the capacity of moving money into gold in a large number is extremely limited … So the players in the world that … I have contact with, who are — who’ve got money — really don’t view gold as an effective alternative”.

That last quote is probably the best counter to Clif’s claim: the reason Ray Dalio and the coterie of players he has contact with don’t view gold as an effective alternative is because they would move the price, a lot.

Some may argue that such manipulations only affect the price in the short term and the price would recover as the manipulator buys back their position. True, if one wanted to cover themselves within the same day. But consider the July 20th price smash – yes the price has recovered, but it is still $30 below where it was before the selling. If the 22 tonnes was one person they could have bought back their position by now by buying only 3 tonnes a day, which as we have established above, would not move the price.

Chris argues that “is the gold market or any market really bigger than institutions that are fully empowered to create infinite money — central banks?” I would say the quotes above show that one hardly needs infinite money, as the gold market does not have the legendary but vague liquidity that many have claimed it has.

Jul 202015

I was just settling in to write an article on the increase in China’s gold reserves when at 9:30am the gold price got smashed. Initial news reports seemed to put the blame on the Chinese market, with statements such as “bullion fell to as low as $1,088.05 an ounce … shortly after the Shanghai Gold Exchange opened trading” and “According to ANZ, the sudden collapse in gold prices earlier in Asia was due to 5 tonnes of bullion being dumped on the Chinese market” but it started on Comex.

Nanex tweeted this picture which shows the first 4 seconds of the drop in the Comex August futures contract (GCQ2015). Note that it starts just after 9:29am Perth time.


Below is a 1 second time interval chart of the August futures contract from Reuters. The area in the red circle is the 4 seconds of the Nanex chart above, which puts the move into context.


The two green stars are 20 second trading halts. After the first halt the price opened up (the diagonal line) but then got smashed again after which another halt was imposed. This halt seems to have given people time to get their head around the drop and then step in for some bargain hunting and the price recovered above $1105.

Note that all this happened in the space of 1 minute and a lot of the detail that one sees on a Reuters tick-by-tick feed is lost by most traders using lower resolution charting/trading systems. The chart below comes from the CME website and you can see how it just shows the move as one bar.


Note that the volume traded in this one minute was 7,164 contracts, which at 100 ounces a contract is about 22 tonnes. ANZ reported that “half an hour after the market opened we saw 5 tonnes of gold sold through the Shanghai gold exchange, which is way above normal levels.” Reuters report that “more than 1 million lots were traded on a key contract on the Shanghai Gold Exchange”, which is equal to 10 tonnes.

The chart below shows 1 second intervals for the SGE and spot (over the counter) markets.


You can see that traders on the SGE reacted after the Comex price move and bullion banks did not update their spot quotes until 9:30 and a half minutes, 1.5 minutes after the initial Comex drop. This delay would be due to the fact that a bank is committed to deal on its bids and offers on an exchange whereas Reuters over the counter spot quotes are not actionable like an exchange (see here) so they are going to update the exchange quotes first.

Reuters’ article provided the following quote/”explanations”:

  • “It looks like someone was taking advantage of the low liquidity environment at the moment. It’s a bit of speculative selling going on,” said Thianpiriya [ANZ analyst]
  • There were stop-loss orders around the $1,131 an ounce level, said a Sydney-based trader, reflected in a spike in volumes on Comex futures.
  • “I just feel there’s a big push to get gold below $1,100 and then we bounced very quickly,” said a trader in Hong Kong.

While gold subsequently recovered to USD 1115, it is still down $15 from Friday, so objective achieved, we assume. With record short positions it doesn’t look like we have bottomed yet and are still waiting for the uptrend signal.

Jul 142015

China continues to open up its gold market, with plans to launch a renminbi-denominated gold fix and talking with the CME to “list its products and prices on CME, whose members and clients will be allowed to trade the Chinese exchange’s products”. Such news reports are usually accompanied by references to China wanting to “increase its influence in global gold markets” in line with its size in the physical market.

Such developments are generally seen by goldbugs as positive, the narrative being that China is a physical gold market and they love gold and only buy it (never sell), which will mean that will we have a market “that is not distorted by the banks, their proprietary trading, or control of the gold distribution system globally”, according to Julian Phillips, for example.

No doubt Pierre Lassonde will be right when he said that “10 years down the road, the Shanghai Gold Exchange (SGE) is likely to determine the gold price, not the COMEX” but a closer look at the Chinese market reveals it may well be a case of “Meet the new boss, Same as the old boss”.

While the SGE’s most popular contract is the 9999 Gold, the deferred delivery contract (like a forward) also attracts a lot of volume. People also forget that there is a Shanghai Futures Exchange, which has a gold contract no different to Western gold futures.

Consider also the Bank of China International, who recruited a “former Goldman Sachs metals trading chief as an adviser to help it expand its commodities business”, someone who was on the management committee of the London Bullion Market Association and an executive committee member of the London Metals Exchange.

Or how about the Industrial and Commercial Bank of China, who is reported as launching “gold accumulation schemes, swaps, forward hedging, lease/financing, collateralized loans and other financial services” and who “is working with HSBC, JP Morgan Chase, Brink, Metalor and other professional logistics providers”.

Chinese have been involved in using “gold to engage in currency and interest rate arbitrage transactions”, collateral financing trades, or leasing and selling gold “from banks to solve their short-term funding problems in the hope of buying back the gold at lower levels to repay the lease” (see here), so much for Chinese only being buyers.

While the Chinese are indeed large consumers of physical, I think it is naïve to think that as their gold market matures their bullion banks will refrain from the same exchange/over the counter proprietary paper trading activities that Western bullion banks do, or that the same greed dynamic we have seen driving leveraged Chinese stock market investing (via official and hidden margin lending: link) will not occur in the gold market.

This dynamic is no doubt what was behind Pierre Lassonde’s forecast that “when we reach the peak in this gold cycle, the SGE will resemble a casino. The Chinese have a huge propensity for gambling, and this is what will likely propel the gold price to levels that we probably can’t even imagine.”

And then to a crash we probably can’t even imagine.

Apr 302015

In January Steven Saville of The Speculative Investor newsletter wrote an article saying that “focusing on the changes in gold location is pointless if your goal is to find clues regarding gold’s prospects.” While I get what he was trying to debunk, I think he has thrown the baby out with the bathwater in doing so.

His target was specifically those who make the claim that “demand is rising even though the price is falling”, usually in respect of Chinese demand. I agree that this is a nonsense statement as for every buyer there is a seller, so to say that demand is up in China is the same as saying that supply is up in Switzerland.

Steven is correct that “the price trend is determined by the general urgency to sell relative to the general urgency to buy” and that price is the only way to know the relative urgency: “If the price is rising we know, with 100% certainty, that buyers are generally more motivated than sellers.”

So who is the baby that Steven threw out in his article? A clue is in this statement:

“while there could be a reason for wanting to know the amount of gold being transferred to China (I can’t think of a reason, but maybe there is one), the information will tell you nothing about the past or the likely future performance of the gold price”

It is the future price baby. When looking at the past, I could concede that price is probably all that matters, but when it comes to the future price, who bought (or sold) I think matters a lot. Steven himself says in his follow up post that “the most useful information is that which provides clues about the likely future intensity of buying relative to selling”.

Let me explain by an example. When I was a kid, football cards were popular. If you live on the moon, this is what I’m talking about.









I picked the first two names I recognised out of google’s image search. Apparently people collect these things. When I was a kid we used them to flick them and play snap type games and they’d get damaged quite a bit, then I grew out of them and gave them away.

Anyway, according to Steven, all you need to look at is the price of football cards going up. It is not important to know who was buying, say that it was a crazy old millionaire who was buying, stacking them in his house and never selling them – that would apparently have no impact on future prices.

I’d argue that it does matter if the buyers were a wide variety of normal collectors vs a market dominated by one crazy buyer. I’d like to know if he was keeping the cards in good order, or whether they were rotting away in a basement, and how old he was and whether his children had any interest in cards or would dump them on the market when he died. That seems all pretty relevant to the “future intensity of buying relative to selling.” Knowing who was buying in this case would give you an edge on guessing future football card prices.

To bring this back to gold, it does matter to me to know if the buying in size was coming from institutions/hedge funds (ETF flows give an indicator of this) or Chinese/Indians, for example. On the balance of probabilities on past behaviour, I would not want to make a long term bet on gold prices based on hedge fund buying as it is only a “trade” to those people and I could be sure they will sell up at some point.

However, given the known and demonstrated cultural hoarding behaviour of Chinese and Indians, I would have a bit more confidence that this gold would not be coming back into the market in the future. On the other hand, if one thinks that the Chinese economy is going to struggle, then maybe buying will drop off or even dishoarding will occur. Steve may not care to know estimates of how much gold is in China or India, but surely the bigger the hoard the bigger the risk that accumulation may stop (eg China gets enough central bank gold reserves to get into the SDR) or reverse?

Now while India and China are like black holes for gold consumption, I would point out that this is price positive only in the long term. It is a macro factor supporting gold but has less impact on prices in the short run, which is why if you run a correlation between ETF or Chinese flows to price you don’t get much. This is probably what Steven is rebelling against, this simplistic construction of a Chinese demand “meme” as respect to immediate prices.

The fact that people may be making the wrong assessment about the future (lack of) selling intensity of current gold holders, or that they are focusing far too much on short term prices, or that they may be using such “analysis” to pump their products/services (the bathwater), doesn’t mean not knowing who the market participants are (the baby) is not important. Reasoned and considered analysis of the current buyers and sellers and their motivations I think has some predictive usefulness.

Someone who would agree with me I bet would be HSBC and JP Morgan. As the two bullion banks with the greatest market share of the gold trade, they have access to information about where gold is flowing from and to well before anyone else. I find it hard to believe that seeing orders to buy gold (to replace sold coins/bars) from The Perth Mint and other mints and distributors, seeing orders from miners, jewellery manufactures, from scrap merchants and industry users – that all that “who” information is “pointless” for future price prediction.

Some people wonder why bullion banks are so successful and the fractional reserve bullion banking system hasn’t imminently failed. Maybe, just maybe, that knowing exactly who, when, where and how much is being bought and sold both in physical and paper markets is information that has value and gives one an edge.