Putting a value on gold

As a graduate trainee in the late 1980s, I spent some time discussing how to value gold with my then boss, the CIO of a small asset management business. He was not a fan of gold and argued that a bar of chocolate is more useful. He did, however, offer up two old City sayings: an ounce of gold should be able to buy you a tailor-made suit on Savile Row; and an ounce of silver should be able to buy you a good lunch at the Savoy Grill afterwards.

More recently, efforts at estimating a fair value for gold have thankfully become more sophisticated. Given that gold currently forms a backbone in many Cazenove Capital portfolios, I thought it would be worthwhile setting out my thoughts on the subject.

UK institutional investor Charles Morris recently wrote a research paper for the World Gold Council on this topic. He started by asking a question: “If gold was a bond, what kind of bond would it be?” His thought process was as follows: it would be zero-coupon as it pays no interest; it would have a long duration as it lasts forever; it would be inflation-linked, as gold’s historic ability to preserve its purchasing power (to a greater or lesser extent) has demonstrated; it has zero credit risk (assuming it is held in physical form), and finally – it would be issued by God (although he wisely did not suggest which one). His conclusion was that gold most closely resembles a long-duration, zero coupon US Treasury Inflation-Protected Security (TIPS).

A model which looks at gold on this basis gives a fairly good estimate for the level and direction of the gold price, as can be seen in the below chart, produced by my colleague Ahmet Feridun.

Gold is performing as our model suggests it should

Gold price vs. estimated value based on historic interest rates and inflation

^{Source: Bloomberg}

There are times where the spot price is at a premium or discount to the estimated value. This can be explained by one or more of the following reasons: firstly, gold is a finite resource, so sensitive to supply dynamics which are arguably more rigid (or inelastic) than TIPS; secondly, gold has physical as well as financial demand, most notably for jewellery; thirdly, gold is ultimately a commodity which is priced in dollars, giving it sensitivity to currency swings; and finally gold is held as a store of value by central banks as well as institutional and retail investors, all of which are subject to behavioural biases and changing sentiment.

A model that tries to estimate gold’s value could use variables representing some of these factors. However, we found that the additional informational value was marginal compared to the more basic model illustrated above. Although the gold price appears slightly ahead of where the naïve model would suggest, it is still a good directional indicator.

The father and son team Charles and Louis Gave, founders of Gavekal, are two people whose work I follow closely. They recently tried to address the challenge of valuing gold and took a different approach. Their starting point was that “gold is the ultimate scarce asset, and its value is determined neither by its utility nor an ability to generate cash flows, but because it is rare”.

Valuing this type of asset is tricky as there is nothing obvious to base that valuation on, and prices of these scarce assets are driven by the supply and demand for them at any time. They also observed that gold represents different things to different people. For some, it is money, for others a commodity, a financial asset or a piece of jewellery. They used different proxies to estimate the value of gold for these different groups. If gold is thought of as money, they concluded that it is far from looking overvalued. If gold is viewed as a commodity, it appeared expensive; however, there may be an argument that other commodities are too cheap (particularly oil). If gold is treated as a financial asset, it has scope to rise further relative to US equities and treasuries, though perhaps less so relative to global equities. If gold is considered a consumption good, it looks expensive when compared to the average worker’s income but more affordable in the context of his or her balance sheet. Their overall conclusion is that gold has not yet reached the kind of extreme valuations seen at the top of previous bull markets.

These are two very different approaches to valuing gold, but both come to similar conclusions: there is still gasoline in the tank of this gold bull market. Gold has risen a long way and made a significant positive contribution to the returns of Cazenove Capital portfolios over the last two years. However, the precious metal still has room to run before it regains parity with the cost of a Savile Row suit.