Introduction

In April 2013 a sudden and inexplicable collapse in the value of gold triggered considerable debate about its implications, and (because of gold's growing strategic significance) several scenarios are considered below.

No particular scenario seems adequate on its own to explain what happened. Rather a combination of factors seems likely, including:

• a 'flash crash' effect (associated with computerised trading in a complex financial market) that accentuated changes that had other causes;
• a widespread perception that gold prices were due to collapse (as they did in the 1980s) - a perception that arose because differences in the current situation that have their origins in profound differences between Western and East Asian financial systems were not widely appreciated;
• increasing pressure on the US Federal Reserve (because of QE's unpredictable and potentially destabilizing effects) to phase out the aggressive quantitative easing that has helped keep the global economy afloat by providing access to cheap credit (especially to governments) and also boosting asset values ; and
• a possible collapse in the value of 'paper gold' assets if arrangements like fractional reserve banking have existed in 'paper gold' markets with no reserve arrangements to backstop a run on physical gold.

If so then the 'canary in the goldmine' seems to be predicting severe global financial and economic disruption (eg rising interest rates facing heavily indebted governments / institutions; the possibility of significant losses affecting financial market players that might resurrect the credit freeze that unknown counterparty risk generated after 2008; and no viable method for counter-cyclical macroeconomic management in the face of a severe financial crisis).

In late July 2013 it appeared that a crisis could be emerging for bullion banks (ie major banks that create a market in 'paper' gold) because demand for physical gold was strong while supply was short. They could only meet the physical demand by paying a (currently small) premium to acquire physical gold (a most unusual position). The premium for physical gold was reportedly very much greater in Asia. This constituted a 'run' on those banks - which could potentially lead to a need for another round of bank bail-outs  [1]

This possibility was reinforced in April 2014 when the establishment of a physically-settled gold futures market in Asia was seen to indicate that the Western makers of paper-gold futures markets (whose physical gold holdings had disappeared) might be bankrupted [1] - though by that stage it was likely that significant institutions had eliminated their exposure to losses from failure of 'paper gold' markets.

In July 2015 there was an overnight crash in gold prices associated with an apparently coordinated large-scale (57 tonnes) sale at a time when liquidity was likely to be limited.

 A Report: Powerful speculators have launched an unprecedented attack on world gold market as US Fed prepares to tighten monetary policy. Anonymous funds sold 57 tonnes in Shanghai and New York at a time when liquidity was likely to be limited - so that confidence would erode. This happened after China stated that its gold reserves had increased by much less than had been expected. The prospect of tighter interest rates can be hazardous for gold - as this is encouraging capital inflows to the US seeking higher yield and reversing the process that may have led to gold boom. However some believe that China's stated gold reserves are significantly under-stated as it does not wish to be threatening $US as China seeks 'reserve currency' status from IMF. Central banks were net sellers of 400 tonnes of gold pa til early 1990s. Emerging economies have been net accumulators because they owned little gold and wanted to reduce $US / paper currency dependence. They were buying 400 tonnes pa but this has now reduced to 100 tonnes pa. Russia reserves fell as result of Ukraine crisis and oil price fall. Vulnerable states may now have to sell gold reserves when US Fed starts tightening interest rates in earnest - as emerging markets have $4.5tr in $US denominated debts [1]

CPDS Speculation. This event occurred in a very complex / unstable international economic / financial environment and could have had diverse commercial and geo-political motivations.

A relationship with the also-engineered gold price crash in 2013 was neither impossible nor obvious. The latter had seemed likely to reflect a desire to protect the holders of paper gold in Western financial institutions from ruinous demands that they produce physical gold which they had little access to though they were contractually committed to give 'gold' to someone who had bought it from them. It may be that an inability to engineer a physical gold squeeze had reduced 'Asian' motivation for accumulating physical gold.

Or the latest event might simply reflect a market view that gold had experienced a boom in earlier years as a byproduct of QE - and that, as the era of easy money policies was over, gold would fall to something like its pre-boom value.  However the way the sale was organised implies that the goal was to depress prices rather than to generate capital from the sale of anyone's gold holdings.

Or the latest event might reflect a view that China was telling the truth about its gold reserves - and thus that strong Asian demand could not be relied upon to underpin the value of gold. The above article suggested that emerging economies had reduced their accumulation of gold reserves.

However if China's gold holdings were larger than stated and it were intended at some stage to launch a gold-backed currency as an alternative to (or as well as) gaining 'reserve currency' status for the yuan from the IMF, then crashing the value of gold could render that tactic ineffective. And, if a VERY significant share of China's foreign exchange reserves were in gold, then crashing gold's value would accentuate the risk of a financial crisis that China faces (and that other emerging economies potentially face). There might thus be a relationship between the engineered gold market crash and the crash on China's stock market that had occurred a few weeks previously and which China's government had 'pulled out all the stops' to reverse.

On the other hand, it was suggested below that China had not accumulated sufficient gold reserves - and needed to buy more. In that case it would have been China that had a motivation to arrange a rapid fall in the price of gold.

In any event it seems highly unlikely that gold prices would fall too far as prices (eg about $US 1100 per oz) were approaching the typical cost of new gold production.

Other Views

China made a statement about its gold hoard - which showed only a modest (ie 60%) increase. It is only pretending to be transparent. Understating China's reserves justifies lower prices - and this suits China as it is buying.  China will play by the rules until it joins the IMF's 'SDR club' - but it will then break those rules. Many assume that China is seeking to escape from the fiat money system by creating a gold-backed currency - but this is wrong it wants to join that system, China can't become a reserve currency without the 'plumbing' (eg a deep liquid bond market dominated in that currency). The latter does not exist. There must also be the systems associated with a bond market (eg networks of traders). China could not launch a gold-backed currency even if it had 10,000 tonnes of gold - because this would only amount to 6% of their money supply and to run a successful gold standard 20-40% would be needed. China's goal is neither a gold-backed currency or a reserve currency. So its goal must be to make it's currency  important enough to global trade to give it power over others . China wants to join the 'club' which requires that it behaves itself as far as US is concerned (eg maintains its currency peg). It has launched its own development bank to put pressure on the IMF. The IMF will probably admit China to the 'SDR basket' later in 2015 [1]

The fall in the price of gold has triggered a rise in demand for physical gold - and a doubling of its premium over paper gold [1]

Scenario A: Another Salvo in the 'Currency Wars

The present writer's initial reaction (see Another view of gold) was that the gold price collapse seemed likely to have a relationship with the (so called) 'currency wars'.

Aggressive quantitative easing (QE) appeared to be being used (particularly in the US) as a way of reversing financial imbalances / current account deficits (noting what Japan's easy credit and carry trades had done prior to the start of the GFC) as part of a long term contest between Western-style 'capitalism' (ie profit focused investment) and the (so-called) 'financial repression' of East Asian 'authoritarian family-states'.

And gold seemed to be becoming increasingly strategically significant for national foreign exchange reserves - noting: politically-motivated proposals for gold-based currencies from various Islamic countries; claimed efforts by various countries (most notably China) to acquire gold (as an alternative to holding foreign exchange reserves in the form of $US assets); and Germany's frustrated attempt to repatriate its gold holdings. 

Frustrated Repatriation: Germany reportedly attempted to arrange for its gold reserves to be repatriated from storage elsewhere - but then found that this would take many years to put into practice presumably because that gold had been leased to financial institutions by other reserve banks who had been storing it on Germany's behalf as the basis for a 'paper gold' trade.

Thus it seemed possible that efforts may have been put in place (eg by the US Federal Reserve) to defend the $US as the world's reserve currency (and thus Western-style profit-focused financial systems) by attempting to devalue the gold alternative and thus make it less attractive.

However that scenario became less likely as it emerged that (perhaps-predictably) it was the value of 'paper gold' that was collapsing while bullion (physical gold) was still in strong demand. Triggering a collapse in the value of 'paper gold' would have been contrary to the Federal Reserve's perceived interests.

Observer's Reactions

Observers offered many widely differing explanations of the gold price collapse - which will at times reflect their own commercial / political agendas (and the circulation of ‘disinformation’).

For example:

• the collapse in the price of gold is not justified by facts in the public arena (as its fundamentals still seem sound) - so this seemed to be evidence of market rigging [1];
• Fed panicked when value of gold rose to $US1917 - as this challenges $US value relative to other currencies. Individuals (rather than countries) started exiting $US. The Fed's dependent banks shorted the paper gold market. However those in China / Russia / India want to buy gold at lower price. Fed is creating $1tr pa but world is moving away from using $US as reserve currency. Fed's attack on gold indicates that hard times are coming [1]
• there are indicators that the gold price fell because of major 'paper gold' sales, and that there is a simultaneous rush to buy physical gold at the lower prices that are now available (as indicated by queues of buyers overtaxing the capacity of bullion dealers) [1].
• premiums over the official price are increasingly being charged for physical gold [1];
• the collapse was caused by the US Federal Reserve and the Bank of Japan. The recent equities boom has priced in an economic recovery that does not exist - with the world stuck in a trade depression with manufacturing over-capacity and a record global savings rate of 25%. But excess supply does not explain gold's collapse. Threats by the Fed to end quantitative easing began in February (based on concern that the longer QE lasts the harder it will be to extricate from). And it had been assumed that the Bank of Japan's QE would revive the yen carry trade - but there are signs that this did not happen. Japan may in fact be exporting deflation because of trade effects [1];

• those with significant short positions may be manipulating the price lower (as long suggested bythe Gold Anti Trust Action Committee) because they were concerned about the lack of inventory in Comex (physical gold) markets due to increased demand for physical metal since February and concerns about counter-party risk in holding paper gold. Futures traders have been selling far more paper gold than there is physical gold in Comex inventories [1] ;

• the gold price fell because of speculative onslaught by major financial institutions (with support of US Federal reserve). There was naked short selling (ie sales not supported by any physical metal) of 500 tonnes of paper gold. There was at the same time a massive demand for physical gold. The paper gold market could disconnect from the physical gold market [1];
• Japan caused the gold sell off because big players were forced to sell to meet margin calls associated with Japanese bond holdings that arose because 'Abenomics' (ie quantitative easing in Japan which involves bond purchases in excess of those by US Fed in an economy 1/3 the size) increased the perceived risks associated with Japanese bonds. Selling gold was the easiest way to meet these margin calls - and was seen to be overvalued. US and European institutions reacted to this sell-off by ceasing to buy. The sudden collapse was a consequence of computer-driven speculation [1] ;
• It is uncertain whether gold price collapse was due to 'paper' markets (eg futures markets) or the exchange traded funds (ETFs). ETFs are listed investments whose price tracks the price of sharemarkets, bonds, commodities (eg gold). There is about $2tr invested in ETFs. Most ETFs are backed by physical gold, but some overseas may use derivatives to track the price of gold. Others see the futures market as the explanation - because the paper market (futures) is huge and is where the price of gold is set. Hedge funds buy or short the paper - and there is much more trading in paper than in physical gold. Traders' bets on futures are usually leveraged - so price movements can be magnified. EFTs are less likely to be to blame - as when they are sold nothing is added to the supply of gold. Fundamental forces drive the gold price - though ETFs can speed up the change. Gold 'bugs' hoped that QE would generate inflation and weaken the $US thus forcing gold prices up. But inflation has not appeared, and $US is strong. Thus investors want out of gold. Two gold ETFs listed in Australia are backed by physical metal [1];
• Eastern central banks are trying to buy large quantities of physical gold, but there is none available. The gold price collapse simply reflects short selling in the paper market. Those trying to push down the price are constrained by the lack of physical supply. A massive rebound effect is likely [1];
• the current situation seems like gold collapse in 1980s. After 10 years of rising prices, gold is no longer the preserve of a few specialists. But optimism evaporated, and bankers have faced sell calls from panicked clients. There have been warnings of problems. The boom was driven by fears about financial system which encouraged investors into gold as a safe haven. The ECB's promise to do 'whatever it takes' to save the euro also eroded enthusiasm. Indian gold buying fell sharply in 2012 as its economy slowed. China's demand fell after surging in 2011 . Weak demand from countries accounting for 40% of global consumption meant that buying by western investors became more important. But this did not happen. ETFs whose creation a decade ago allowed easy investment in gold began dumping their holdings - selling 182 tonnes in the first three months. Cyprus bailout deal did not boost gold prices - as it would have previously. Cyprus's gold sale to fund bailout became a rout - and perhaps presages widespread monetisation of gold reserves. Gold bugs believe that QE will boost demand for gold as alternative currency, but despite large QE programs inflation remains weak - and deflation fears are rising. QE might generate inflation - but not for many years. Demand for physical gold has surged - and dealers ran out of stock [1];
• there is increased disagreement about whether the price of gold is being manipulated. Gold is seen as protection against government use of the printing press. GATA sees price manipulation as an attempt to defraud gold holders - whereas it is more likely to be simply an attempt to protect the value of currency. Gold in 1971 was grossly undervalued - because its value had been held down by government. It then rose too fast to 1980. Inflation can be beneficial in making it easier to pay off debts - while current low inflation rates make this difficult. The extent of QE recently makes serious inflation possible - some see inevitable. A high gold price is seen as indication that this is happening - and thus money printers seek to control gold prices. However they can't do this forever [1]
• the Fed is rigging markets. It rigs the gold market to protect the US's exchange rate which is threatened by QE. It is adding to the supply of dollars faster than there is any demand for them. If $US falls, inflation would rise and Fed would lose control over interest rates [1]
• long seen as safe in times of turmoil, gold now looks shaky. Analysts have called for an end to past gains. Some see strong consumer demand in China as restoring strength, while others expect a repeat of 1980s' meltdown. Factors responsible include: Cyprus sale (and expectations of similar sales by other European countries); large sales by ETF; cooling of fears that QE would lead to inflation; electronic trading could have led to fast changes; large amounts have been taken from gold ETFs recently. 171,000 tonnes of gold have been mined so far. Gold is hard to value because its price if affected by many factors, and it has limited uses. In recent years economic uncertainty has elevated its price, while growth was moderate. One analyst saw outflow of hot money as all that was behind the fall. Retail investors have bought gold because of the price fall. A disconnect has emerged between physical market and Wall Street's 'paper' gold. People would not be rushing to buy if this was a bubble bursting. People are concerned about counter-party risk - that doesn't affect gold [1];
• believers give many reasons for expecting higher gold prices - inflation risk, currency collapse, flight to safety.  But there is no way to know what value is right ('Gold: look out below', 16/4/13)
• no one can explain implosion of gold price. Theories include: possible early end of QE; shift to a 'risk on' environment; dumping gold by Cyprus; receding inflation expectations. The problem is that financial are opaque and sophisticated - so authorities to tell what is happening. Exchange traded commodity funds were big players in the rout and possible that macro hedge funds (via ETFs or directly) were also involved. There were net sales of 480 tonnes of gold over two days (about 20% of annual mine supply). Most of this was purely synthetic exposures (purely financial bets, not physical sales) - and some would be highly leveraged thus adding to steepness of the fall. Strangely over the week the Japanese yen strengthened, despite its aggressive monetary policy easing that had previously seen the currency fall. Hedge funds had created a massive carry trade - shorting the yen while going long gold. When the gold price broke these positions had to be reversed. There had been a net built-up of short positions by fund managers during 2013 and an accelerated outflow of funds from ETFs. Gold is not a liquid commodity, but gold securities (via ETFs or derivatives) are. Given the probability of leverage with carry trades and ETF exposures, margin calls would have accelerated the crash once it started. This does not explain why the fall started, but the dynamics and structure of global markets probably explains why it was so rapid [1]
• it is 10th anniversary of founding of ETFs which played a role in gold's price surge and rapid fall. At their peak gold ETFs held $150bn of gold. From a start in 2003 ETF's metal holdings grew rapidly to 85m ounces in 2012 - and then the sell-off began. Over past 6 months ETFs have experienced outflow of $10bn. Fundamental reasons for decline in gold price involve market having been over-excited about the inflation risk associated with massive money printing, and tanked when theis did not eventuate. ETFs had a role in rise and fall of gold price by making it easier to get in and out of gold market. They may have transformed markets once dominated by investors in real products into markets dominated by investors / speculators [1]
• it seemed initially that gold had sold off because central bank efforts to revive the global economy were working. It later appeared that the sell-off was for the opposite reason (ie because the world was sinking into recession / depression) [1]
• Gold sales by reserve banks played a major tole in pushing prices down to $250 / ounce in 1999. Brighter economic outlook has encouraged investors to seek riskier assets. Gold is less valued as an inflation hedge, as there is little inflation. US economy has been improving and that boosted $US, and made QE less needed. Global economy also looks better. Others suggests that gold's fall reflects reducing budget deficits worldwide and thus less reason for gold to rise due to expanding reserve bank balance sheets. Eurozone is in austerity and US has cut back. China is slowing faster than official data indicate. Falls in gold have accompanied past financial crises. Gold bugs see its rise as due to currency debasement ['Gold fever', AFR, 20/4/13]
• Easier access to gold via EFTs has encouraged use of gold for speculation with leverage, and this may have contributed t crash ('Speculators tarnish gold', AFR, 20/4/13)
• There is disagreement between those who believe that gold is overvalued (due to lack of income / appreciation) attributes and others who see it as a protection against other risks. Recent volatility illustrates artificial pricing in western market-based economies. Central banks are aggressively boosting their balance sheets and disruption of economic systems seems possible. Rising prices generate feedbacks (ie further price rises) which break link between physical demand and supply. The catalyst for gold price fall was recognition (from events in Cyprus) that quantities of gold could be suddenly dumped. However the fact that reserve banks have separated financial markets from economic fundamentals is also a factor. The global economy is in a stable disequilibrium - with an expectation that reserve banks can deliver higher growth. If this doesn't happen there will be more to worry about than gold prices  [1]
• Recent slide in gold price has generated substantial demand for bullion that will bring disaster on central and bullion banks. In West gold is seen as an investment. However West holds little of world's 160,000 tonnes of gold. Investor behaviour is quite different to that of hoarders. Analysts tend to look at the issue in terms of economic theory - which does not understand growing demand for protection from monetary instability. Hoarders in Asia value gold more than those in the West. This puts paper currencies at risk. In 1974 world had 80,000 tonnes of gold - with 37,000 tonnes in monetary for - with the rest as bullion / coins / jewellery (about 30,000 tonnes was in West). By 2000 after 20 year bear market very little of the latter was held in West. Since 2000 demand from India and China has grown rapidly. Western bullion markets have thus been on the edge of a physical supply crisis for years. Nearly all new mine production has been absorbed by non-Western hoarders / central banks. This must eventually make futures / forward gold markets fail at some point. Recent price crash may indicate that this has happened.  Undervaluation of gold by western investors led to gold migrating to Western and Asian hoarders. Physical demand is increasing with every price drop. Ther eis a growing pricing conflict between the futures / forward markets (that don't involve physical settlement) and the physical market. Analysts look at gold supply in terms of mining and scrap supply (when nearly all gold mined is available at some price). The price of physical gold is suppressed by paper market below what it would be. Bullion has been transferring rapidly from capital markets to hoarders - leading to acute shortages in capital markets. Running unallocated gold accounts )whereby a clients gold is taken onto a bank's balance sheet) is profitable under stable conditions - especially where gearing is used. Paper foward positions may thus be many times more than physical bullion available. Firms data is not available - though there seems to be a huge synthetic short position. Thus any price rise could be catastrophic - drain bullion supplies and escalate the risk. Reserve Banks have been using their gold to support the market (via leasing arrangements) to protect the value of their currencies. However there was a shortage of physical gold before the recent price collapse. If this was engineered by central banks - it backfired spectacularly. Central banks may not be able to support bullion markets by intervention much longer - and will need to rescue the bullion banks other ways to maintain confidence in paper currencies. Gold held by struggling eurozone nations won't support markets for long - and may already have been sold. Another financial crisis is very likely.  [1]

Preliminary examination of available indicators suggests a number of other scenarios.

Scenario B: Another Flash Crash

A large, sudden (and quickly reversed) crash in the US stock market in 2010 was seen to have been due to idiosyncrasies of computerised trading systems that generate feedback effects. Something like this might have been a factor in gold price crash.

Scenario C: A Repeat of the 1980s

A crash in the value of gold like that which occurred in the 1980s has been suggested by some observers.  After a boom associated with rapid inflation in the 1970s, governments instituted many inflation fighting measures which saw gold lose value [1]. In particular this started after Paul Volker took over as the Fed chairman with a determination to fight inflation by tightening up on interest rates.

Against this scenario is the fact that high inflation (which was seen to be a factor in the 1970s' gold boom) has not been a factor in recent years. Rather gold price increases seem more likely to have been a consequence of concern about the world's main reserve currency (the $US) because of the rapid rise in government and other debts (and in the Federal Reserve's balance sheet), partly as a consequence of international financial imbalances and QE.

Moreover East Asian economies are now a significant component of global economy, and involve financial systems that are incompatible with the global financial system - thus implying that one or the other is likely to fail (see Structural Incompatibility Puts Global Growth at Risk and Understanding East Asian Systems of Socio-political-economy).

The relevant impact of those systems involves:

• so-called 'financial repression' to direct 'family-state's' savings to production with limited regard for profitability;
• generating cheap exports that keep inflation under control in trading partners; and thus
• permitting trading partners to run easy money policies to boost asset values as the basis for strong debt-based imports.

This 'virtuous circle' was a major factor in underpinning the long boom that preceded the GFC, but has arguably passed its use-by date.

And China, various Islamic countries and others seem to be keen buyers of physical gold (eg as a way of ensuring that foreign exchange reserves are not held in $US, or as a basis for a possible gold-based currencies as an alternative to the $US).

Gold's 'flash crash' in 2013 was nothing like its price decline in the 1980s - but key differences were arguable 'below the radar' of many market participants..

Scenario D: Phasing out Quantitative Easing

A plausible reason for an unexpected collapse in the gold price combined with growing signs of equities' market instability would be an expectation that reserve banks (especially the US Federal Reserve) might significantly reduce quantitative easing [1].

This could happen (for example) because of recognition that:

• aggressive QE has been and is generating unwanted adverse side effects (distorted investment decisions and asset bubbles); and that
• these risks were also being amplified by Japan's equally aggressive QE - which (assuming that that Japan's financial system still limits the flow of credit to households) would mainly result in carry trades and asset bubbles offshore.

The gold price collapse started, it has been noted, at the same time that notes from a US Federal reserve meeting were accidentally (?) released indicating that a more 'hawkish' approach to quantitative easing might be gaining support [1]. Moderating / eliminating QE would: (a) probably leave authorities with no useable tools for counter-cyclical macroeconomic management; and (b) allow market forces to apply to the high debt levels that have been created by many years of easy credit. The latter might in turn:

• send interest rates much higher in the short-medium term and thus bankrupt numerous institutions / governments and result in massive losses (in bonds and equities) that could again dislocate financial systems / economies worldwide;
• reduce the rationale for holding gold as an alternative to fiat currencies;
• promote faith in the $US in particular when the dust settles, and thus in Western-style international institutions (ie those based, for example, on a rule of law and profit-focused accounting principles that facilitate initiative by 'rational / responsible individuals' as compared with the 'financial repression' and social relationships that are the foundation of non-capitalistic market economies in East Asia's 'family states');
• make it impossible for the trading partners of East Asia's neo-Confucian economic / financial systems to provide the credit-fuelled demand needed to protect financial systems that don't take profitability seriously and thus give rise to financial crises, economic dislocation as well as social and political instability across East Asia. (most significantly in Japan and China);

While the consequences of moderating / elimination QE would be severe, it might be that there is no no alternative.

Scenario E: A 'Paper Gold' Wipe-out

A ‘paper gold’ crash would be possible if many owners don't actually have a right to anything (as there is not enough physical gold to support all 'paper' entitlements) at the same time as there is a strong and growing demand for physical gold.

Reliable information about ‘paper gold’ does not seem to be readily available.

About Gold and 'Paper Gold'

The Wikipedia article on gold indicated that:

• up to 2009, 185,000 tonnes of gold had been mined (valued at $US8.8 tr at $1600 / oz);
• world production in 2009 was 2700 tonnes;
• about 50% of all gold produced has come from South Africa;
• China overtook South Africa as the world's top gold producer in 2007 (with 276 tonnes);
• gold is used for jewellery (50%); investment (40%) and industry (10%);
• India is the world's largest gold consumer (buying 800 tonnes annually / 25% of world's gold, mostly for jewellery). India imported 400 tonnes of gold in 2008. Indian households hold 18,00 tonnes of gold;
• China consumed 817 tonnes of gold for jewellery in 2012 (just behind India) having greatly increased its consumption between 2009 and 2011 - in line with a similar global increase;
• in 1968 the gold pool collapsed, and a two tier pricing scheme was introduced (til 1975) under which gold was used to settle international accounts at $35 / oz, while the private price fluctuated;
• the world's largest gold depository (US Federal Reserve NY) holds 3% of all mined gold

The Wikipedia article on Gold as and Investment indicated that:

• most of the gold ever mined still exists as bullion or jewellery - and thus can come back onto the market at the right price;
• at the end of 2004 central banks held 19% of above-ground gold;
• the 1999 Washington agreement on gold limited gold sales by members to 500 tonnes pa. This was extended for 5 years in 2009 with 400 tonne pa limit;
• Russia started adding to its gold reserves in 2005. In 2006 China (which held only 1.3% of reserves in gold) announced intent to increase its holdings. India also increased its gold reserves;
• gold prices tend to rise and fall inversely with interest rates;
• gold ETPs (exchange traded products) are an easy way to gain exposure to gold. However such products, even when backed by physical gold, carry risks - eg some have been compared with mortgage backed securities due to their complex structure;
• gold certificates allow investors to avoid the risks associated with holding bullion, but have their own risks and costs. Banks may issue either allocated or unallocated certificates. Unallocated certificates are a form of fractional reserve banking and do not offer guarantees in the event of a run on the metal. Allocated certificates correlate with specific bars, though it is impossible to tell whether banks may have allocated the same bars several times;
• gold accounts are are available - on a similar basis;
• derivatives such as gold forwards, futures and options trade on various exchanges or over-the-counter;
• as of 2009 COMEX gold futures have experienced problems in delivery of bullion - suggesting to some that COMEX may not have the gold inventory to back its existing warehouse receipts;

Wikipedia article on Gold Reserve provided details of gold reserves claimed by various countries (which are not independently audited). It also noted that gold is held as: jewellery (52%); central banks (18%); investment bars (16%); industrial (12%); and other (2%).

Wikipedia article on Gold Certificates referred to both the certificates that reserve banks have historically issued entitling holders to a portion of their gold holdings. and also to current options for investing in gold by acquiring certificates (ie ‘paper gold’) from banks that may either involve ‘allocated’ physical gold or unallocated gold. It also suggested that the latter involves a form of fractional reserve banking (whereby banks traditionally lend many times more than their own reserves) and thus that such certificates don’t guarantee that holders will receive an equal exchange of metal in the event of a run on ‘paper gold’.

Manipulation of gold markets is alleged by the Gold Anti Trust Action Committee (eg as outlined below) – but is officially denied.

Gold is the worst understood financial market. Most official data is misinformation. The IMF allows member nations to count gold they have leased out as if it is still in their vaults. China announced in 2009 that its gold holdings had increased from 600 to 1054 tonnes. Saudi Arabia reported an increase from 143 to 323 tonnes since 2008 - though later suggested that it had long held this. Some believe that both China and Saudi Arabia have accumulated more than they claim. In 2009 Germany admitted that much of its gold reserves were held else where - but did not say how much of this had been leased out. In 2009 the US Fed stated that it had a secret gold swap arrangement with other countries. Thus it is not certain that US has the 8200 tonnes it claims. It has not been audited for 50 years. It is unclear how much gold the major ETFs hold. While they report data, it seems unreliable. They won't say where their metal is. And the custodians for the major ETF's are international banks which have big short positions on gold - which gives them a major incentive to suppress the price of assets that they are holding on behalf of investors who are hoping for gold to rise in value. How much gold do ETF's themselves hold, and how much of it is encumbered. No one knows. The biggest physical market is the London Bullion Market Association. It publishes statistics on gold and silver traded by its members that exceeds the amount of metal that exists. The London bullion market is a fractional reserve gold banking system built on the assumption that most gold buyers will never take delivery of their metal. Most gold sales by LBMA members is highly leveraged - but it is not clear how much traded gold does not actually exist. Paper gold has been invented.  [1]

In the event that, as some have suggested, there are fractional reserve arrangements involved in gold markets, then (as for normal banking) there would be a risk of institutional failures in the event of a 'run on the metal' (ie attempts to claim more physical gold than is actually available) unless there were some sort of 'reserve banking' arrangements in place. And for physical gold this would be hard to arrange if the 'physical gold' that those who make the 'paper gold' market is actually gold leased from reserve banks and the reserve banks can't create any more with the stroke of a pen.

As 'paper gold' and physical gold markets are apparently moving in opposite directions (ie investors want to sell paper gold products while there is a strong demand for physical gold at the same price), there is a possibility that the gold flash crash of 2013 could be partly due to suspicion / recognition that there is not enough physical gold to go around, so that many owners of various forms of unallocated 'paper gold' will find that they don't actually have a right to anything.

If so, a collapse in the value of ‘paper gold’ might generate a financial crisis like the sub-prime crisis in 2008 though of currently-unknown severity (with unallocated gold 'paper' playing a role similar to Collateralised Debt Obligations in 2008). Major financial market players might have unallocated ‘paper gold’ on their books (who knows how much) that becomes essentially worthless overnight. This could give rise to another credit freeze because of unknown counterparty risk, and a legal frenzy as losers sue the institutions (eg banks) who created the ‘paper gold’ market.