Short-run and long-run determinants of the price of gold

Abstract

In 1833 the price of gold was $20.65 per ounce, about $415 in 2005 terms, while in 2005 the actual price of gold was $445 - a very small change in the real price of gold over a period of one hundred and seventy two years. Despite this apparent constancy in real terms over the long run, it is also true that, outside of periods when the gold price was fixed through various iterations of the gold standard, it has fluctuated significantly in the shorter term, sometimes for years at a time. Can these two apparently contradictory realities be reconciled? And can one be sure that the long run positive relationship between gold and inflation has persisted beyond the era of Bretton Woods? Indeed, is there any credence to the claim that gold can be used as a long-run hedge against inflation? The results reported in this paper provide some answers to these questions that are so central to the gold market and its many participants around the world. We also address the inflation hedging properties of gold in the currencies of the major gold-consuming countries outside of the USA, taking into account both the domestic exchange rate relative to the dollar and domestic consumer price index movements. Real gold prices denominated in the home currency of investors outside of the USA also deviate in the short-run from their home country inflation hedge price and there is also a long-run tendency for gold prices to revert to the long-run hedge price. The major gold consuming countries outside of the USA, that is, India, China, Turkey, Saudi Arabia and Indonesia were rational to purchase gold in that it proved more than adequate as an inflation hedge. For these countries the actual USA dollar gold price between 1976 and 2005 far exceeded the dollar gold price required to provide an inflation hedge after taking account of exchange rates between the US dollar and the home country and the home country consumer price index movements.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 159 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.