The policy reversal of the Federal Reserve Open Markets Committee (FOMC), which forecasted a 50 basis point (bps) increase in its lending rate in January only to lower the same figure by 75bps since that time, has turned a headwind into a tailwind for gold prices that reached $1,546/oz. on September 4th, a 21% increase from the beginning of the year. With Central Banks purchasing bullion at the fastest pace since 1967 and gold inventories held by electronic traded fund (ETF) at record levels, we believe a new ‘Bull-Market’ has formed for the yellow-metal, which has pulled-back 6% in price (to $1,462/oz.) in the last several weeks. Moving forward, we expect elevated Central Bank demand to support current price levels, and believe the heightened risk environment – as it relates to (i) negative interest rates, (ii) geopolitics, and (iii) the possibility of recession – will put upward pressure on gold prices through 2020, when prices could very well challenge their September 2011 high-point of $1,895/oz., a 23% increase from today. Should headline risks be realized, prior precedents suggest gold prices will double in value to $3,000/oz. in the following 2-3 years. Given the risk/reward and the benefits to diversification, we also recommend a gold allocation upwards of 10%.
Background
Gold is a Unique Asset. “For more than two millennia, gold has had virtually unquestioned acceptance as payment to discharge an obligation,” writes Alan Greenspan. “Gold has always been a means of payment that does not require a credit guarantee of a third party.” The only comparable asset is silver, which is not only subject to tarnish and more malleable than gold, but also found in 10x the quantity. For these reasons, gold continues to play a central role in the International Monetary Fund (IMF) system that was established in 1944 at Bretton Woods on the principal that its members would benchmark their currencies against the U.S. dollar (USD), because it was the only currency (then) freely-convertible into gold. Despite the abandonment of the gold-standard in 1973, U.S. gold reserves have notably remained unchanged at 8,134 tons, equal to 77% of total reserves. “The fact that central banks choose to hold any gold, which could otherwise be invested in interest earning fiat money securities and not incur the costs of storage and security, is evidence of gold’s unique status,” argues the former FOMC chairman.[1] “Gold is money,” noted J.P. Morgan. “Everything else is credit.”
Bullion supply is extremely well-understood. In contrast to most assets, the world’s supply of gold is naturally-constrained and small enough to track, such that the above-ground stock (of 193k tons) can fit into two Olympic-size swimming pools. Per Exhibit 1, jewelry represents 48% of above-ground supply, with the remainder held in bars/coins (20%), Central Bank reserves (17%), Technology (14%), and ETFs (1%). By comparison, below-ground gold reserves total 54k tons, an amount equal to 28% of above-ground supply. New mining supply is expected to grow 2% per annum (p.a.) through 2020, with production declines “as early as 2021,” according to Standard & Poor’s, because output at existing mines is falling faster than expected supply gains at new mines. The pattern is further consistent with prior periods, where there has been a strong relationship between expansionary phases in GDP and under-investment in gold mine capacity, as fixed investment is generally steered to base metals in growth periods. As a result, absent a new technology that improves exploration success, any large increase in supply should be limited to recycled gold.
Investment Demand a Fraction of Gold Market. Gold demand can broadly be broken-down into three groups, of which the largest is (i) non-financial buyers (see Exhibit 2) in the jewelry (47% of last twelve months demand) and Technology (7%) end-markets, where demand is generally stable, but highly elastic, such that buyers turn to recycled supply, when gold prices suddenly rise. The price-sensitivity of these buyers explains why gold prices have, historically, kept pace with inflation and why, for over 100-years, an ounce of gold has afforded a high-end suit on Saville Row. By comparison, (ii) investment demand, in the form of coins/bars (20%) and ETFs (11%), is more volatile and closely-aligned with the health of the overall U.S. economy. Lastly, (iii) Central Bank gold demand (15%) has undergone a massive shift (see Appendix A) in the last 15-years. Having been the largest source of supply for over 30-years, public institutions only became net buyers of bullion in 2010, when central banks in the European Union (EU) stopped systematically selling-down reserves.
Central Bank Gold Demand highest since Gold-Standard abandoned. In the last-twelve-months (LTM), Central Banks have bought 713 tons of gold, a 9% rise from 2018, when total purchases of 656 tons represented the largest annual increase since the abandonment of the gold-standard. The sizable increase is, oddly, attributable to demand (see Exhibit 4) in a small number of countries in the Emerging Markets (EM), such that the top-five purchasers, together, account for 85% of net Central Bank gold demand since 2014. The biggest gold buyer in the period is Russia, where gold reserves grew 105% (from 1,041 tons to 2,230 tons), alongside a corresponding 92% decline (from $132bn to $10bn) in U.S. Treasuries (UST), in response to U.S. imposed sanctions in 2014. Similar diversification activities have occurred in China, where the People’s Bank of China (PBOC) has reduced UST by 14% (from $1.3tn to $1.1tn) since their peak in 2014 and bolstered gold reserves by 84% (from 1,055 tons to 1,942 tons). More recently, in July, the central bank of Poland announced its gold reserves had increased by 100 tons (or 78%), on account of “strategic safeguarding” and “financial security,” alongside an 8% decline in UST holdings since the beginning of 2019.
ETFs increasingly dictating Gold Prices. The 16% increase in gold prices during the first three fiscal quarters of 2019 tracks (see Appendix B) a near-equal gain in ETF inventories that reached a new record high-point of 2,808 tons at the end of Q3. This is not a coincidence. While only 1% of above-ground supply, ETFs have had an outsized influence on gold prices, since their introduction in 2003. Having been previously limited to buying gold in bars/coins only, ETFs unlocked new demand (of 1,231 tons) between 2004 and 2008, which helped offset Central Bank gold sales (of 2,226 tons) and drive annual prices gains of 20%. In the following three-years, ETF demand then surged (to 1,327 tons) pushing-up prices 48% p.a. to an all-time high of $1,895/oz. in September 2011. Gold prices then collapsed in 2013, when ETFs unloaded 913 tons, and trended downwards until 2016, when concerns over Brexit and negative interest rates led to a resurgence in EU-based ETF demand. Taken together, gold prices have recorded gains (see Exhibit 3) in every year ETF inventories have increased and fallen whenever ETF demand has been in decline.
Investment Thesis
Negative Yielding Debt will drive Gold Demand Higher. While gold has no yield, it is a well-known store of value that, unsurprisingly, has seen its price move upwards (see Appendix C) in lock-step with the amount of negative yielding debt outstanding. The rotation is evident in economies issuing negative yielding debt that guarantee a loss at maturity, most notably the EU, where ETF inventories (see Exhibit 5) are now 44% of the global total and 31% higher than their prior record high-point in 2012. With 16 Central Banks having elected to raise interest rates in Q3 and an additional two-dozen expected to follow suit by year end, according to J.P. Morgan, the forward rate curve should, if anything, trend lower into 2020. Given the size of the respective markets, even a small rotation (out of bonds) into gold would have a meaningful price-impact on the yellow-metal. For context, ETF gold inventories at Q3-end had a value of $134 billion, less than 1% of the negative-yielding debt outstanding, so a single percentage point (pp) shift in the latter equates to a doubling of the former.
Demand from Central Banks should support Gold Prices in Forecast Period. The upswell in Central Bank gold demand is a new and underappreciated phenomenon with sizable scale: LTM demand of 713 tons is bigger than that of ETFs in every year on record. Having been dormant for most of the post-Bretton Woods era, Central Bank gold demand is likely to persist at increasing rates, given the deliberate ‘de-dollarizing’ efforts of Russia and China, which both (see Appendix E) have large state-owned domestic mining concerns that offer a convenient and sustainable source of gold in the 5-year forecast period. The breadth of demand is also expected to expand, as evidenced in Poland and a recent survey of Central Banks that showed 54% of respondents planning to add gold to reserves by July 2020.[2]
Gold will remain the obvious ‘Safe-Haven’ for ETF investors. The rate of return on gold is the inverse of that in the overall U.S. economy, according to data from Larry Summers, so it is a natural hedge to recession. Pattern data also shows an inverse correlation between gold prices and the Dow Jones Index, during the last three downturns and the 1929 crash.[3] Gold prices have also performed well (see Exhibit F), when the S&P 500 Index is in decline; a relationship that has become tighter since the introduction of ETFs in 2003. When a recession does materialize, we expect ‘hot money’ to pour into gold ETFs, causing sizable price-gains. After all, gold ETFs launched the passive ETF category, which has grown assets by 5x in the last decade, so when investors return, we suspect they’ll behave like the ‘Sand People’ and be back in “greater numbers.”
Growing Demand for Physical Gold in Asia. Wealth is the primary driver of jewelry demand that occurs at a higher income per capita (see Appendix G) than any other good, as evidenced in China and India, where the combined share of the jewelry market has risen from 25% to 57% since 2000; a figure that is expected to rise higher, given OECD forecasts projecting a 970 million increase in the middle class population by 2035. The argument for forward growth is particularly strong in China, where per capita GDP is forecasted to grow from $8k in 2018 to $30k in 2035, according to Goldman Sachs, where analysts (see Appendix H) have found that peak gold accumulation occurs, when economies have a per capita GDP between $20k-$30k, such that the data exhibits a “hump-back” formation. Based on the historical data, China’s gold demand will grow 5x by 2035.
Markets are heavily-discounting Future Inflation. Given their historic relationship, gold represents an excellent hedge against consumer price inflation (CPI) that has been heavily-discounted by market participants, in our view. September data shows headline CPI increasing 1.8% from last year, with core CPI rising by 2.3%, a level not seen since 2008. Inflation levels are also broadly forecasted to rise higher in 2020 by Wall Street economists, on account of upward pricing pressure from trade tariffs. Meanwhile, Federal Reserve data shows average inflation expectations of 1.7% during the next 5-years, so a modest upward revision (of 50bps) could trigger a meaningful loss in bond prices.
FOMC Balance Sheet Expansion will drive Gold Prices Higher. Historically, gold prices have benefited, when Central Banks debase their currencies with rapid balance sheet expansions. This was demonstrated during the last U.S. recession, when the FOMC introduced $3.5 trillion of new Quantitative Easing (QE) stimulus and gold prices doubled in value in the two-years after the S&P 500 Index bottomed-out in March 2009. Having lowered interest rates (see Appendix I) an average of 6pp to stimulate growth in prior recessions, the FOMC is constrained by interest rate levels of just 1.5% today, so policy-makers are expected to pursue unorthodox monetary policies to combat the next downturn. That could include QE stimulus or ‘helicopter money,’ the policy idea introduced by Milton Friedman, where stimulus is distributed directly to the consumer-level by crediting bank balances or tax rebates. What’s more, the process appears to have already been set in motion by the FOMC, which (i) lent $200 billion to securities dealers and (ii) boosted monthly UST purchases by $60 billion, in the last three months. While the details remain unclear, the adjustments imply an increase of $400 billion (by March 2020) to the FOMC balance sheet that was actively lowered from $4.4 trillion to $3.7 trillion since 2017. This suggests a weakening of the USD and, in theory, a strengthening in gold prices.
Gold likely Beneficiary of US-China Decoupling. Given steps taken by the PBOC and the ‘decoupling’ process already underway with the US, it is easy to envision a scenario, where China’s gold reserves begin to challenge the dominant position of the US in the next decade. A similar idea was put forth by Greenspan, who argued China’s currency, “could take on unexpected strength,” if the PBOC rotated $300 billion of its foreign reserves into bullion to, “displace the U.S. from its position as the world’s largest holder of monetary gold.” On a pro forma basis, gold would represent just 10% of total PBOC reserves, which (see Exhibit 6) compares to 77% for the U.S., so there is an underlying logic to a rotation into bullion. The PBOC could leverage its larger gold holdings, for example, to argue for a higher allocation of the Special Drawing Rights unit of the IMF, which added the renminbi at an 11% weighting in 2016. Perhaps most importantly, “the penalty for being wrong, in terms of lost interest and cost of storage, would be modest,” writes the former FOMC chairman.[4]
Silk Road Nations may reinforce Gold’s Currency Role. If the FOMC introduces large stimulus in the next recession and ‘Silk Road’ nations continue to add bullion at accelerating rates, gold’s significance as a currency reserve component will increase. As an October statement from Denmark’s central bank noted: “If the system collapses, the gold stock can serve as a basis to build it up again.” Alan Greenspan has echoed a similar sentiment, saying: “The broader issue – a return to the gold standard in any form– is nowhere on anybody’s horizon.” While seemingly far-fetched, the notion gains weight when set against the historical precedent of Bretton Woods in 1944, when Great Britain reluctantly agreed to readopt the gold-standard for the third time, as a result of pressure from the U.S. that eagerly performed the same feat for the second time, because the FOMC (then) controlled 60% of the global supply.
Risks
ETF Sell-Off. For gold, the biggest risk is the possibility that a sudden reversal in economic sentiment triggers an ETF-driven sell-off. Ultimately, the gold market consists of a small number of participants, most of whom are trading on the London Metals Exchange, so a sudden influx of ETF sell orders has the capacity to cause rapid price declines. This occurred in April 2013, when an off-hand remark from a European Central Bank representative, who suggested Cypress could satisfy its external debts by selling its gold holdings, triggered an outsized ETF order of 400 tons that led to a 16% price-drop in two-days.[5] Today, we believe the risk of a sell-off is mitigated by existing uncertainties related to negative rates, geopolitics, and the possibility of economic recession that, in our opinion, should compel investors to maintain current gold allocations. Also, while global ETF inventories are at new record levels, recent demand has been largely attributable to EU investors, such that US holdings remain 20% below their prior peak in 2012, which suggests ample runway for growth. All else equal, total ETF gold inventories would increase 31%, if US holdings rose from 51% to 62% of the total.
Central Bank Gold Demand Highly-Concentrated. Since the beginning of 2014, the combined gold demand from five central banks has represented more than 85% of the total. More specifically, Russia and China have, together, accounted for 66% of total net Central Bank demand in the last 5-years, so the forecast is highly-dependent on a select handful of players. While concentration risk is a factor, we believe U.S. foreign policy will cause Russia and China to add to their gold reserves, which represent just 21% and 3%, respectively, of their central bank holdings. By comparison, gold represents 77% of central bank reserves in the U.S. and 73% in Germany. What’s more, China and Russia are large gold producers, so their central banks have an incentive to purchase bullion domestically. This is particularly true in Russia, where fiscal rules require its government to save excess oil reserves in the central bank.
Sudden expansion in supply. While a sudden increase supply has the capacity to hurt gold prices, it is unlikely to occur in the forecast period, because (i) no new large mines are scheduled to come online during the next 5-years and (ii) it takes 3-years to bring a new mine into production. Gold is also naturally constrained on earth, where native deposits sank to the planetary core (along with other precious metals) when it was formed. Contrary to common perception, the gold mined on the earth’s crust actually originated in an asteroid bombardment estimated to have occurred 3.8 billion years ago, according to 2011 research. As a result, precious metals are much rarer on earth than most other elements and they are rarer on earth than in the solar system, where gold is 100x more common. By contrast, diamonds are a particular arrangement of atoms formed from a common element (i.e., carbon), so its supply is constrained by technology, rather than physical availability.
Recommendation
Well-respected financiers, most notably Warren Buffett, have been highly-critical of gold as an investment, so three points, in our view, are worth reiterating. Firstly, we believe the traditional argument against gold – that stresses the yellow-metal earns no yield and is only worth what others are willing to pay for it – does not apply in instances, when ~25% of all corporate and government debt is trading at negative yields, which guarantee a loss at maturity. This idea is supported by (i) the positive relationship between gold prices and the amount of negative yielding debt outstanding and (ii) the rotation into bullion already evident in EU-based ETF inventories that are now 31% higher than the prior 2012 record. We expect this relationship to persist as well.
Secondly, the supply/demand balance in the gold market today suggests an extremely favorable risk/return ratio, in our view. When compared to prior periods, it is important to recognize the current demand picture is unique, because ETFs and Central Banks are both adding gold at elevated levels. After all, Central Banks were the primary source of ETF supply until 2010 and only began ramping-up purchases after ETF inventories had peaked, so we are in the midst of a paradigm shift. And with both groups buying at near-record levels, we believe the upward pressure on gold prices will persist thru the November 2020 U.S. presidential election. Should interest rates and recession concerns remain in place, we expect gold prices to surpass the September 2011 high-point of $1,895/oz. by year-end 2020, implying a 23% increase from prices today. Having already reached all-time highs in eight of the G-10 currencies, we are simply expecting the same scenario to emerge next year for bullion priced in the Swiss franc and USD. Of course, gold prices will rise much higher, in the event of recession, with prior precedents suggesting a doubling in prices to $3,000/oz.
Lastly, gold has proven to be an excellent hedge against falling equity prices that, today, are at valuation levels unseen outside of the Dotcom Bubble. As a reminder, the S&P 500 Index fell 49% over 40-months (from peak-to-trough) between 2000-2002 and 56% over 17-months between 2007-2009, during which the gold price increased by 10% and 24%, respectively. The latter figure includes the outsized price-impact of ETF demand, which was notably absent in the former period, making it representative of future gains in any downturn scenario. Gold prices also doubled in the two-years after the S&P 500 Index bottomed-out in March 2009 and recorded an increase of 31% in the 15-months between the inversion of the yield-curve in 2006 and the beginning of the last recession.[6] A strong argument can be made that we are witnessing a similar episode at present too, given the yield-curve inversion in March and the New York Federal Reserve’s recession probability model that has climbed to 35%, the highest reading since the financial crises, so it’s only prudent for investors to both take and protect profits.
Taken together, we believe investors should acknowledge the risks at-hand and rotate 10% of their portfolio into gold. Ray Dalio of Bridgewater Associates has made a similar recommendation, saying: “It would be both risk-reducing and return-enhancing to consider adding gold to one’s portfolio.” We agree. If you are still trying to reconcile the view put forth here with Mr. Buffet’s sharp criticism, we suggest you focus on the $128 billion cash position of Berkshire Hathaway (BRK), a highly-defensive position for a company with an equity market value of $521 billion. Instead, we are simply favoring gold over cash, because the former represents a clear hedge against falling equity prices. From our vantage, the difference of opinion stems from the size of BRK’s cash holdings that are comparable in scale to the Swedish Riksbank and, therefore, too large to be successfully unwound outside of the UST market. Fortunately, the same restriction does not apply to the vast majority of investors. Any other reservation regarding gold, in our view, can be answered by Joseph Schumpeter’s observation that: “The modern mind dislikes gold because it blurts out unpleasant truths.” This is one point all investors can agree upon.
[1] Greenspan, Alan. The Map and the Territory 2.0. New York: Penguin Books, 2013.
[2] Central Bank Gold Survey, World Gold Council, July 2019.
[3] This relationship was notably interrupted for four decades, beginning in April 1933, when the FDR Administration enacted Executive Order 6,102 that barred U.S. citizens from holding gold, and lasting until its repeal in December 1974.
[4] Greenspan, Alan. “Golden Rule: Why Beijing is Buying,” Foreign Affairs, September 19, 2014.
[5] In its first 8-years of trading, GLD rose 380% in a nearly uninterrupted streak until September 2012, when prices failed to test the prior high-point and then drifted down -10% over 6-months until April 2013. All told, gold prices dropped 44% between 2013-15.
[6] The yield-curve has inverted roughly one year before the last three U.S. economic recessions. The pre-recession inversions also reversed themselves before the subsequent downturn. Both of these events have taken place since March 2019.