One of today’s biggest investment challenges is forecasting the gold price. Gold is the only asset which had been a consistent positive performer in the last five years, but has many characteristics of a typical asset class bubble. The gold bugs are divided right in the middle with the bull camp advocating that the gold price may reach US$5000/oz and the bear camp advocating a reversion back to US$600/oz. The implied assumption of the bull case is that a part of the government institutional and personal reserves and portfolios will be converted into gold. Or central banks will convert their forex holdings into gold. Or at least, a large portion of institutional and personal wealth will be converted into gold.
On the other hand, it can be argued that the gold price has most of the characteristics of a bubble, as perceived wealth is created on the basis of marginal investment in gold. As a marginal dollar invested in gold – either as a deflation or inflation hedge – it creates a wealth effect – and hence, a snowball price effect from speculation. Gold is different from other typical asset bubbles in that it is always perceived as a store of value against other assets, which primarily are consumption assets and hence are effected by oversupply and demand effect which helps to prick the bubble. So, the thesis that gold is a bubble because of creation of a value – wealth effect may not be logically tenable because gold storage is part of, and a consequence of, value effect and hence, traditional speculative wealth may not be able to prick the bubble.
It's tempting to consider gold a bubble. It currently trades at $1370 an ounce, an all-time high. Gold had been the best performing asset in the last ten years. It's survived two asset bubbles' bust. It sells for three times its April 2001 decade low of $255 an ounce. Notwithstanding the parabolic shape of the gold price chart, the gold price does not appear to be in danger of an imminent collapse. In terms of timing, the gold price may not collapse till middle of next year and possibly not till end of 2011. In terms of price, gold can potentially run till US $1500 /oz and possibly till US $2000 / oz. What ultimately may prick the bubble is a pickup in industrial activity, inflation and interest rates, which may reverse capital gains driven, and carry trade. Further, the retail segment will be the first one to exit the gold positions, which may result in a deluge. I outline some of the reasons for this opinion below:
- Different type of bubble: Gold is different than most other asset classes because gold does not have any use and almost no consumption. However gold is produced, it remains in storage. Marginal investment demand is filling the gap left by a decline in Jewelry demand. This is the most common argument used by gold bears. However, gold is perceived as store of value which is deeply rooted in human psychology and history. Hence, investment demand, which could be a symptom of speculation in any other asset class, may be considered a legitimate demand as wealth effect is what gold needs to provide.
- Gold is not rich: Gold is not extremely rich compared to most of the other commodities (relative normalized target price around US $ 1185 / oz). For gold to be in a real bubble, its price must be +40% higher than comparable commodities. The price where I may expect the gold price to inflect will be around of US $ 1700 / oz. So, either the entire commodity complex is mis-priced or gold is not in a bubble.
- Regime Change: Investment access provided by gold ETFs has opened the gates of gold investment to retail investors – which seems to be a game changer – an entirely different scenario. Another reason why gold has become more attractive is because it has become more accessible. The rise of the exchange traded fund has made the process of buying gold as easy as buying a stock. These came to the fore in 2004, coinciding with the continued spiking of the metal. While institutional investors can pinpoint their hedges with currency swaps and other sophisticated derivative strategies, mom-and-pop investors must rely on a smaller repertoire featuring physical gold and its ETF proxies.
- Timing the bubble: Investment managers, such as stock mutual fund managers, are compensated and retained in part due to their performance relative to peers. Taking a conservative or contrarian position as a bubble builds results in performance unfavorable to peers. This may cause customers to go elsewhere and can affect the investment manager's own employment or compensation. The typical short-term focus of equity markets exacerbates the risk for investment managers that do not participate during the building phase of a bubble, particularly one that builds over a longer period of time. In attempting to maximize returns for clients and maintain their employment, they may rationally participate in a bubble they believe to be forming, as the risks of not doing so outweigh the benefits that are making the gold bubble stickier.
- Portfolio Insurance: Another aspect is that gold may be increasingly used as portfolio insurance by retail investors. The correlation of gold with most other asset classes over the last 5 year and three year periods is low, not to mention that absolute standard deviation of gold is as good or as bad as most of the other asset classes. Please look at the charts below, showing correlation of different asset classes.
click to enlarge
- Pension funds: Pension funds will increase gold holdings to acquire “financial insurance,” pushing prices higher as currencies drop. Record government debt and interest rates close to zero percent are pushing gold higher for a ninth straight year as investors seek to protect their wealth against the prospect of rising inflation and currency debasement. Teacher Retirement, backed by $95 billion in assets, has launched its first internally managed gold fund, worth $250 million, invested in precious metals, mining stocks and exchange-traded funds. Gold represents only 0.4 percent of total global financial assets, valued at around $102 trillion in 2009.
- Debt monetization: The price of gold could be pushed up if there are expectations that central banks will monetize their countries' debts, but this increases investors' risk aversion and will lead them to sell gold.
- Global AUM: Please look at the chart below with the estimates of global assets under management. If in light of low correlations as shown by the charts above, from a diversification perspective, investments increase their allocation towards gold by 1%. This implies an investment of around US$ 1trillion and marginal demand of 26,500 tons of gold at current market prices, almost equal to central banks' official gold reserves (28,000 metric tons) against an annual supply of 4,000 metric tons from all sources (Jewelry annual demand at 2,000 tons).
- US Dollar: The U.S. dollar has already been slipping, and now there is every indication that the Fed is hungry to start printing even more money. While 2008's gold movement was due to a weaker dollar, in 2009, gold has started to become a bet against all currencies and in anticipation of a major crisis forcing people to shift to Gold and Treasuries. So, in a sense, the gold-dollar relationship of last year has reversed.
- Asset allocation: Any asset allocation strategy suggests that both from a hedging and diversification perspective, investing in gold makes great sense. Our rough calculations show that the total amount of gold ever mined is around 169620 tons, whose value at current prices is $6.5 trillion. That is a huge number, but not all that big compared to total investable capital of $103 trillion, equities $60 trillion, bonds $15 trillion. Bubble territory would be $3,000/oz. Still, those gold commercials and neighborhood gold-buying parties are worrisome.
- Forex reserves: The world's central banks turned net purchasers of gold in the second quarter of this year, having been heavy sellers of gold for decades. This reflected a combination of lower sales by European central banks as well as the emergence of new purchasers, such as China and Russia, as noted above. Again, they seek protection from the dollar. It's not just retail level investors hoarding coins, and buying gold ETFs, either. In the second quarter, China told the world that it had bought 600 tons of gold, and Russia recently admitted it owned 100 tons. Both nations are scared about their amount of exposure to the dollar, and are seeking diversity.
- Supply constraints: One fallacy is that the gold rush is new. As noted at the top of the story, the metal has been appreciating steadily for most of this decade, predating the credit crisis. There are several reasons why. One is that on the supply side, the amount of gold worldwide has been contracting. Mining production has been on the downward slope since 2001, meaning simply, less gold to go around.
- Deflation / Inflation: The Gold price moves in tandem with deflation and has been seen to move inverse to inflation. There is only one example in history (1979-1980) of a big move in gold being a sign of rapidly rising inflation. In 1983, gold rallied nearly 50%, yet inflation fell. From 1985-1987, gold was up nearly 65%, yet inflation was calm. In the early 2000s, gold was up nearly 60%, yet inflation fell. The opposite has happened, too: gold fell more than 10% in 1987, yet inflation jumped between then and the early 1990s. Gold is an effective inflation hedge in the long term, but as with any tradeable commodity, its prices can vary wildly over shorter periods, making it seem uncorrelated even for somewhat long time periods.
- There seems to be a lag of +9 months between the gold price decline and the first interest rate hike. One of the examples is that in the seventies, excess monetary expansion after the U.S. came off the gold standard (August 1971) created massive commodities bubbles. These bubbles only ended when the U.S. Central Bank (Federal Reserve) finally reigned in the excess money, raising federal funds interest rates to over 14%. The commodities bubble popped and prices of oil and gold, for instance, came down to their proper levels. Similarly, low interest rate policies by the U.S. Federal Reserve in 2001-2004 created housing and commodities bubbles. These bubbles once again popped as soon as fed funds rates were raised to natural levels. This is one of the reasons for our expectations of the timing of gold's bubble burst.
Although gold has outrun most of the other asset classes in the last 5 years, current relative value does not appear too rich compared to most of the commodity complex. We have used averages and mean reverting levels of relative ratios and our estimate of forward pricing of these commodities in comparison to crude oil, silver, platinum and copper. The reasons for selecting these commodities for a relative analysis are liquidity, substitution and intrinsic pricing. Based on these measures, the average gold price may be around US $1085/Oz, with an upper bound of US $1700/Oz and lower bound of US $581/Oz, with a downward skew of 34%. This price is meant to provide us with an anchor for a normalized gold price ,and not an absolute price target, and is dependent on reduction in asset class risk / economic risk.
Gold to Oil Ratio: The Gold to oil ratio is a barometer of the relationship between the two most liquid commodities. The gold-oil ratio reached a 10-year high of 28 in the start of 2009 as the global recession eroded demand for energy commodities and investors abandoned monetary assets in favor of the safe haven metal. Decline in the ratio from its March 2009 low augured well for U.S growth in particular, and the world economy in general. The rationale is based on the notion that multi-year highs in the G.O. ratio reflect lingering risk aversion. The pullbacks in the G.O. ratio generally coincided with stabilization in the US economy, while rebounds in the ratio preceded a broadening slowdown. Since 1983, the ratio has averaged 15.7 (15.7 barrels of oil were needed to buy one ounce of gold), but this may be misleading because from 1983 to 2000, the crude oil prices were depressed because of North Sea Oil finds (post 1979-83 oil bubble). We may need to treat that period (labeled as period 1) separately. In period 2, the mean level was at 12.85 and in period 3 at 10.85. If we take an average of period 2 & 3, the mean level from 2000 to date is 11.43. Currently the ratio stands at 16.37, well above the averages seen in the preceding decade. The GO Ratio may decline from the current 16.37 at least to 13.55 (mean reverting level from 1999-2009). To put it in a perspective, if the crude price declines to US $80/barrel, the relative gold price should be US $1043/oz.
Gold Silver Ratio: The gold-silver ratio represents the number of silver ounces it takes to buy a single ounce of gold. The essence of trading the gold-silver ratio is to switch holdings when the ratio swings to historically determined "extremes." During the 44 months leading up to the Great Stock Panic of 2008, silver averaged 1/55th the price of gold (I’ll just use this number rather than the fraction from here on out). This pre-panic 54.9 average was fairly tight too. It wasn’t defined by a few rare extremes, but many years of gentle meandering near the middle of a range between 65 and 45. Despite the countless market-moving gold and silver developments since 2005, this 55 average held nicely. All this leads me to believe that an SGR near 55 is far more normal for this bull market than the levels seen during the stock panic. The reason for the lower extreme was that speculators fled everything, including silver; its price plunged far more than gold’s. This drove the SGR to its lowest levels out of the entire precious-metals bulls. A plunging stock market drove the silver price down to $8.92 by the end of 2008, which was only worth 1/84th of the price of gold at $745. During the entire panic (September to December), the SGR averaged 75.8. I fully expect silver’s historic relationship with gold to normalize. It has existed for so long that a fear bubble in the stock markets lasting a few months shouldn’t be able to sever it forever. Nothing fundamentally changed on the silver or gold mining fronts during the stock panic, so the secular-bull pictures for both metals look similar.
Gold Markets fundamentals
Forecasting Gold from a supply/demand perspective is no easier. For one thing, actual physical demand for Gold is plummeting while investment demand is rising. According to the World Gold Council, change in gold demand and supply for 2009, Q2 2010 & September 2010, the gold market has undergone a change. Jewelry use, the largest demand segment, (40% of total demand) declined sequentially by 20% in 2009, 9% in Q1 2010 and 8% in Q2 2010 (535 tons in Q2 2008, 438 tons in Q2 2009 and 406 tons in Q2 2010). This decline in traditional demand is the argument used by most of the gold detractors as the demand destruction effect of the high gold price, which is expected to further accelerate if the gold price keeps going up. Jewelry usage has reduced drastically in the past 10 years by over a third, from about 3000 tones to 2000 tonnes now.
A decline in Jewelry demand of 100 tons in Q2 2010 (total demand of 405 tons compared to 504 tons in Q1 2010) had been counter balanced by a +200 ton increase in investment demand. Of total gold demand of 1045 tons in the second quarter of 2010, 30% were from ETFs. The Gold ETF, GLD, has become the biggest market mover (current holdings estimated at 1467 tons – the 5th largest in the world) and there has been heavy demand for coins and bars. ETF demand was 92% higher in 2009 compared to 2008. The world’s biggest Gold ETF fund has been seeing a huge increase in their assets (rising from 200 tons to 1467 tons). Holding gold as an ETF is the most cost effective thing, as trading costs are very minimal and it provides assaying and value estimation, security etc.
GLD is the most heavily traded ETF and it has been buying gold in gobbles and now holds more than a 1000 tons of gold (worth about 30 billion USD). GLD’s asset increase has coincided exactly with the increase in the gold price. This is quite evident from the build up of managed money long positions. The Bear / bubble argument is that if this change in consumer/investor choices unwinds, Gold could see a significant downward movement in price in the short to medium term, even at $700/ounce. It remains to be seen how this change in behavior would continue after the end of this crisis (in 3-5 years). It is interesting to note that electronics and industrial usage has come down. Either they have made significant technological progress to reduce the need for gold in their manufacturing process, or it is an indicator for fundamental demand collapse in world markets for those products.
Industrial usage has remained constant failure and investment usage has dramatically increased by over 10 times from 2000 levels. The gold demand in dollar terms has been steadily on an upward slope since 2001. While it is not clear it's a permanent consumer change of heart, or a reflex reaction from collapsing markets, this is bullish for gold in the short term. Another traditional argument for a bubble case is that supply increases with a rise in price. Gold supply increased by 12% in 2009 to 4020 tons despite an 87% decline in central bank sales. This trend continued in 2010 as the Q1 2010 gold supply increased by 18% in Q2 2010. However, we must keep in mind that a large portion of supply increase has come from scrap (sequential increase of 35% in Q2 201) as mine production is almost unchanged. It is pretty likely that people are bringing up junk from their old treasures and cashing in on the bullish markets, but this supply segment is not sustainable and may soon start feeling a burnout effect.