Description

The intrinsic value of gold is rarely discussed, even by those value investors who own it.  Almost always when someone pitches gold it’s because inflation is ostensibly coming, the dollar is going lower, interest rates are going lower, central banks are printing money, or the financial system can’t be trusted.  Entire theses can rest on quantitative easing or tightening, even though such flows are very small relative to total credit stocks. These reasons are short-term and flow-oriented, which is peculiar for the oldest financial asset.  Such reasons are deeply dissatisfying because they’re directional rather than foundational and are value agnostic, which is perilous for value-conscious investors. Also these flow-oriented drivers require macroeconomic predictions which are inherently low-probability events because they are individually low-probability, Bayesian, path-dependent events, even when the predictions are approached scientifically, but more often they’re intuitive and emotional.  There’s also perhaps a plurality of investors who think gold has no value. Both ways of thinking about gold are filters that haven’t been very predictive. I prefer a filter that assumes that gold has a value, as it has for thousands of years, and as other currencies have a value, and that that value can be very broadly estimated. A framework grounded in stocks rather than flows may create a better foundation for decision making on a longer time scale, but will admittedly be of limited use on shorter time scales when prices are mostly determined by flows.  

Of all the things that never existed, “macro investor” exists the least, in company with “foreign policy expert”, but I only make a modest, self evident proposal here that if Thing One (in this case gold) is priced in quantities of Thing Two (dollars), then gold’s value is proportional to the quantity of dollars.  Whether gold has value can be argued. And whether dollars have value can be argued. As of today, people have agreed that dollars and gold have value. If one accepts that both gold and dollars have value, it can also be said for art, sports teams, wine, Ferrari 250s, Stradivarii and other frivolous investments that their values are proportional to dollars.  The prices of these assets as a class in fact track annoyingly close to the growth in money.

Since measured inflation is money growth minus productivity gains, businesses are not, in a strict sense, the best inflation hedge since productivity gains are necessarily competed away in aggregate in almost all cases.  The slow march forward in living standards depends on this being true. If today’s panoply of goods and services maintains its share of incomes, then new types of goods and services can’t be purchased. And so the vast majority of firms will be on the receiving end of deflation.  As technologies become more complex with each generation, fewer firms will possess the means to be the best in their industry. There were more sword makers than gun makers, and now there are only a few defense contractors, and in fact only a few nations, capable of developing the most sophisticated weapons.  Innovation is concentrated and deflation is diffuse. It’s unreasonable to expect that you will only own businesses that are immune to deflation.

We can expect that sometimes gold is overvalued and sometimes it's undervalued because short term prices are determined by sentiment and liquidity flows, and are only very loosely, if at all, related to intrinsic value over short time scales.  This is a point that isn’t often acknowledged. Every asset class experiences good decades and bad decades while capital flows to more or less preferred risk assets. The explanations for this process only seem obvious in retrospect and are most likely only narratives even ex post.  While no one has an edge in predicting liquidity flows ex ante, it’s safe to think that money and credit stocks will grow over time and gold’s intrinsic value will grow proportionally. Gold’s price, however, may lag behind its intrinsic value for 10 or 20 years if one overpays for it.

Public and private sector debt versus gold price, 1970 to present

The above chart plots the quantity of public and private debt in the U.S. (in blue) versus the price of gold (in red) from the end of Bretton Woods and gold convertibility in 1971 (both normalized to 100 at 31/12/1969).  I use debt here as a proxy for the supply of dollars. Over the last 50 years debt has grown at an 8.2% annualized rate, and the gold price has increased at a 7.6% annualized rate.

The quantity of corporate profits is interdependent with the quantity of money and credit.  Corporate profits create wages, wages create corporate profits, and they both create money and credit, and money and credit create corporate profits.  And money creates prices. If gold has a value, its value will be created by money, like all other prices. If two things are related to another thing, then those two things are related to each other.  So by inference gold’s value is related to corporate profits. The below chart plots the gold price (in yellow) versus the S&P 500 (in white) from the end of Bretton Woods.

Gold price versus S&P 500, 1970 to present

The price of gold in the short to medium term is weakly, or even negatively, correlated to other asset prices.

Correlation of gold returns to equity and bond returns, 1970 to present:

Daily Monthly 1 Yr 5 Yr

Gov’t bonds .06 .10 .01 -.07

S&P 500 -.02 .01 -.25 -.58

One accepted narrative about gold is that it has a strong positive correlation to bonds.  While this has been less untrue recently, possibly as this narrative has persisted, it’s never been true.  One reason why this may be untrue is that higher rates are necessarily accompanied by higher nominal credit growth and inflation expectations, but this is also just a narrative.

Ratio of gold market cap to the stock of public and private debt, 1969 to present

The above chart plots the gold market cap ratio to public and private sector debt.  Gold has been cheaper sometimes, especially when equities were in extreme favor such as 1999, but it's usually more expensive.  The ratio has averaged .148 over this time, and the median is .104. The current ratio is .079. The ratio is lower in less than 25% of the months observed.

Distribution of the monthly ratio of gold market cap to credit, 1970 to present

Gold has performed well in recessions and in times of volatility.  The below charts plot gold (in yellow) versus the S&P 500 (in white), beginning three months before the onset of recession to three months before the end of recession.

December 2007-June 2009

March 2001-November 2001

July 1990-March 1991

July 1981-November 1982

January 1980-July 1980

November 1973-March 1975

Gold’s mean outperformance from recession onset to recession end was 60% and the median outperformance was 22%.  This is a useful feature for an asset that is also fundamentally long growth, in contrast to bonds and equity short positions.  This low to negative correlation to stocks has the effect of reducing portfolio volatility, and giving an investor more capital to buy stocks after sell-offs.  The below log scale chart illustrates the returns from a 100% equity portfolio (grey), a 100% gold portfolio (gold) and a 50% gold/50% equity portfolio (black).  Volatility is lower and returns are higher in the 50% gold/50% equity portfolio versus 100% gold or 100% equity.

Risks

-Structural government budget surpluses

-Secular deleveraging

-Lower geopolitical tensions

-Amazon

-Strong GDP growth and low inflation

Non-risks

-Accounting fraud

-Leverage

-Management incompetence

-Poor capital allocation

-Competition

-Innovation

-Obsolescence

-Modeling error

-The Internet

Catalyst

Nucleosynthesis