Sovereign Gold Bonds in 2019: Really?

Mark Weidemaier and Mitu Gulati

For a while now, we have been meaning to write about “sovereign gold bonds,” or “SGBs,” which the Indian government has been marketing under domestic law to residents of the country since November 2015. Gold bonds were supposed to have been a thing of the past. We’ve written previously about the U.S. government’s abrogation of gold clauses in both public and private debt in the 1930s. Last seen (to our knowledge) in government and corporate debt around that time, these clauses obliged the borrower to repay in either gold or currency at the option of the holder. (For detailed treatments, see here, here and here.) The point was to protect investors against currency devaluation. Thus, the famous case of Perry v. United States concerned U.S. government bonds that provided for payment of principal and interest “in United States gold coin of the present standard of value.” As the U.S. Supreme Court recognized, the promise sought “to assure one who lent his money to the government and took its bond that he would not suffer loss through depreciation in the medium of payment.” (An investor also would not benefit from an appreciation in the value of the currency, for payment was tied to gold coin of the “present standard of value.”)

The bonds in Perry were “Liberty” bonds issued to finance the 1st World War. The government therefore marketed the bonds as patriotic investments, although then, as now, marketers favored subtlety over heavy-handed appeals to emotion.

Regrettably for investors, it also turned out to be their patriotic duty to accept less than full payment.

As part of efforts to address the deflationary spiral of the Depression, the U.S. government devalued the currency. Because of the widespread use of gold clauses, this would have bankrupted many borrowers. So Congress passed a law in 1933 abrogating gold clauses in both private and public debt. Effectively, this imposed a significant haircut on debt obligations. In a series of cases, bondholders challenged the abrogation of gold clauses and lost. In Perry, the U.S. Supreme Court ruled that, although the government had acted unconstitutionally, bondholders were not entitled to any remedy. Never mind the logic—or illogic—of the result. The lesson is clear and has been repeated in various contexts over the years: When a government’s debt is subject to its own law, it has a great deal of freedom to restructure.

Against this historical backdrop, the SGBs issued by the Indian government (technically, the Reserve Bank of India on the government’s behalf) seem like an anomaly. But these are somewhat different investments.

In the Liberty Bonds, the point of the gold clauses was to protect the real value of the payment stream represented by the bond. The SGBs, by contrast, are an investment in gold. To quote the Indian Central Bank (here), the “SGBs are government securities denominated in grams of gold. They are substitutes for holding physical gold.” The basic idea is that investors (who mostly must reside in India) buy bonds denominated in grams of gold, which mature in 8 years (although there is a possibility of early redemption after five years). The per-gram purchase price is equal to the average 3-day price of gold published by the India Bullion and Jewellers Association (IBJA). The bonds pay 2.5% annual interest. At maturity, bonds are redeemed according to the same pricing formula. If the price of gold has gone up, the investor wins. If it has gone down, the government wins. (We’re ignoring the 2.5% annual interest.)

At first glance, this is a bit weird. For one thing, it is as if the Indian government is systematically shorting gold. That’s … not typical? But one can see a logic to the program. There may be a significant demand for gold, perhaps because residents of India think it is a better store of value than the Rupee. That’s a significant source of financing that the government might like to tap. And the pricing benefits to the government seem clear. For Q1 of 2020, the Indian government reported (pp 9-10) a weighted average yield of 7.21% for new debt issuances with a weighted average maturity of 15.9 years. The government’s outstanding debt had, on average, shorter maturities (10.49 years) and slightly higher coupons (7.81%). If the government can persuade investors that SGBs are functionally identical to gold, it can borrow at 2.5%. That’s a smoking deal, even if we assume the government’s actual costs are a bit higher (say, because it decides to hedge against an increase in gold prices). And for investors, that 2.5% is a “guaranteed” return not available to owners of gold itself. Finally, the SGBs might also generate some cost savings that can be shared between the parties (e.g., by eliminating the cost and risk associated with storing physical gold). So there’s a way to make this into a story of market efficiency.

But there’s also a way to tell the story in which the Indian government is exploiting two attributes of its resident investor base: (a) a strong demand for gold and (b) a mistaken belief that SGBs are in fact substitutes for physical ownership of gold. SGBs are governed by Indian law. And although the IBJA determines gold prices—presumably based on actual transaction data—governments tend to develop a funny relationship with gold in times of crisis, especially when gold prices affect the real value of debt obligations. Imagine a scenario in which the government imposed price controls on gold and other precious metals. What would be the prevailing price of gold published by the IBJA? The price set by the government. At the end of the day, isn’t a local law bond a local law bond? Why would investors accept such low returns for SGBs relative to other local-law bonds?

Perhaps the logic is that the government is uniquely unlikely to default on SGBs. Even if the price of gold were to rise precipitously, and the government had not hedged against this risk, the overall amount of gold-denominated obligations is relatively small. And given the relatively short maturities, the government might prefer simply to change or end the program than to change the law to minimize its obligations. Perhaps, too, there is a belief that the government will hesitate to turn the screws on investors (again, these are residents, with exceptions that appear to be quite minor). Maybe investors also believe that in the “bad” scenario—where the government steps in to manipulate gold prices—they would be no better off with physical gold. After all, another lesson of the 1930s is that the government can ban ownership of gold and force holders to exchange physical gold at prices set by the government. That historical example seems less relevant in a world where currencies are not tied to gold, but perhaps it’s not entirely irrelevant.

In any event, we are fascinated and a bit puzzled by these SGBs. Is this a story of market efficiency, or a funding scheme that will work right up until the point that it doesn’t? If that day comes, it will be when the government is in a deep financial crisis. At that point, all bets are off.

Aside: Mitu’s mom lives in India and asked whether she should invest in SGBs. Mitu’s initial reaction was, “absolutely not; these look like a scam.” She promptly invested in them, explaining that she has learned, through long experience, that the best way to make money in financial markets is to do the opposite of whatever Mitu advises.