Share
Numerous research papers on gold begin by establishing that of all the commodities gold is a distinct asset. Anecdotal themes include the idea that human civilization has long been enamored with gold going back to ancient times, as well as the fact that, unlike most other commodities, bullion can be stored almost indefinitely and requires very little maintenance. These attributes are given as the reason gold has throughout history been used as a medium of exchange and maintains its purchasing power. In Research Study No. 22, Gold as a Store of Value, published by the World Gold Council in 1998, author Stephen Harmston relates the oft-heard story that "an ounce of gold bought 350 loaves of bread in the time of Nebuchadnezzar… still buys approximately 350 loaves of bread today." Notwithstanding that we have been unable to identify an archeological or biblical source for this tale,[1] let us assume that this bit of trivia is true, along with other stories involving an ounce of gold buying a nice suit, armor or otherwise.
Assuming gold does have the unique ability to maintain consistent purchasing power over an extended period of time, such asset would seem to fit the definition of a zero-beta asset. This is the conclusion of a paper by McCown and Zimmerman (2006) titled, Is Gold a Zero-Beta Asset? Analysis of the Investment Potential of Precious Metals in which they analyze gold returns using three different frameworks: the Sharp (1964) and Lintner (1965) capital asset pricing model (CAPM); the arbitrage pricing theory (APT) of Ross (1976); and cointegration using a test methodology developed by Elliott, Rothenberg and Stock (1996), and another methodology by Kwiatkowski, Phillips, Schmidt and Shin (1992).
Running a regression analysis from 1970 to 2003 where the "risk-free" rate is the yield on the US T-Bill, McCown and Zimmerman use three different proxies for the market portfolio: MSCI US Index, MSCI World denominated in local currency, and MSCI World denominated in US dollar. Acknowledging that the CAPM is "limited in usefulness as a tool for investment analysis," the authors observe that "all of the estimated beta coefficients for gold are statistically indifferent from zero," and therefore gold does not reflect "much, if any, systematic risk". In their CAPM study the conclusion is that gold effectively behaves as a zero-beta asset, yet it also "has a positive risk premium". The APT methodology and cointegration tests also validate the authors' hypothesis.
The paper concludes that gold shows "the characteristics of zero-beta asset," and that the "mean return is just slightly higher than that of T-Bills". Further, "estimates of CAPM and the APT show that both gold and silver bear virtually no market risk, and their estimated betas are statistically indifferent from zero". All three frameworks showed evidence of inflation-hedging ability when added to an investor’s portfolio. Obviously, McCown and Zimmerman's research doesn’t take into account the rise of gold since 2003, which brings up that "old saw" about how correlation does not imply causation. Given that gold is priced today much higher than in 2003, yet inflation has generally been tame since the millennium, raises a question concerning the underlying perspective of this study. Perhaps a better framework for studying gold is that is acts like a currency proxy?
Over the course of our lifetime we are trained to value assets from a US dollar-centric perspective, and it is hard for US-based investors to think differently. In Europe one views prices from a euro-centric perspective. However, it is also possible to revalue the economy and markets from the perspective of preserving purchasing power. When assets such as stocks are revalued in gold as a unit of account, we gain a completely different outlook on whether or not such assets have risen or fallen, and to what extent. For example, if we were to value the Dow Jones Industrial Average, not in US dollars but in ounces of gold, it would have cost around 30 ounces of gold to buy the Average at 10,000 when the index first crossed over that line in 1998. When it crossed over that line again back in October 2009, it would have cost less than 10 ounces of gold.
In a March 30, 2010 Financial Times article titled, "Will negative swap spreads be our coal mine canaries?" Gillian Tett forces us to rethink what constitutes a "risk-free" rate. In finance classes the rate based on the three-month US T-Bill is almost always assumed as the risk-free proxy for input into the CAPM. But what if, as Tett suggests in her article, that there is no risk-free rate? Black's (1972) zero-beta CAPM provides a solution by supposing a riskless asset does not exist, and that investors hold different risky portfolios all existing on the efficent frontier. In order to calculate the zero-beta CAPM, one simply replaces the "risk-free" rate with a "zero-beta" rate. This is necessary to calculate the mean-variance of the market portfolio.[2]
One of the more interesting aspects of McCown and Zimmerman's research was that gold still reflected characteristics of a "zero-beta" asset when using the 'MSCI World denominated in local currency' as a proxy for the market portfolio. When employing the CAPM, analysts traditionally measure portfolio variance denominated in either the US dollar or in the local currency in which the portfolio's assets are actually denominated. What we are suggesting is that zero-beta also implies the idea of relative valuation when measuring variance based on some unit of account. In other words, a better method for measuring variance when using the zero-beta CAPM necessitates an exchange rate constant which consistently maintains purchasing power over an extremely long time horizon, rather than a floating exchange rate variable which reflexively alters the value of the assets it is suppose to measure.
Footnotes:
[1] Dr. Claude Mariottini, Professor of Old Testament, Northern Baptist Seminary, states that no where in the Old Testament does it say that "in the days of Nebuchadnezzar an ounce of gold bought 350 loaves of bread." Further, Dr. Mariottini logically points out that "one must assume that the ounce, a unit of weight in the avoirdupois system, once used in the United Kingdom and still used in the United States, was also used in Babylon. Since the Babylonians did not use imperial units, this statement is false." Source: Gold and Bread http://tinyurl.com/yjxpn6d
[2] Zhang, Lu. "The Capital Asset Pricing Model" Stephen M. Ross School of Business, University of Michigan (2007).
Is Gold a Zero-Beta Asset? Investment Potential of Precious Metals
References:
Harmston, Stephen. "Gold as a Store of Value" Research Study No. 22, World Gold Council (November 1998).
Mariottini, Claude. "Gold and Bread" http://doctor.claudemariottini.com/ (June 10, 2008).
McCown, James Ross and Zimmerman, John R. "Is Gold a Zero-Beta Asset? Analysis of the Investment Potential of Precious Metals" Meinders School of business, Oklahoma City University (July 24, 2006).
Tett, Gillian. "Will negative swap spreads be our coal mine canaries?" Financial Times (March 30 2010).
Zhang, Lu. "The Capital Asset Pricing Model" Stephen M. Ross School of Business, University of Michigan (2007).