Gold News

How Commodities Lag Gold

Want a stable currency...?
SOMEWHAT unexpectedly, I've been doing a lot of talking about 'commodity basket'-type monetary proposals, writes Nathan Lewis at New World Economics.
The basic gist of these is that a commodity basket is a good proxy for Stable Value. But, the typical (actually only) argument for this is basically the assumption that commodities are a good proxy for Stable Value, by definition.
I think that commodities, individually and also as a group or 'basket', have a lot of price (or value) variation basically due to supply/demand issues of commodities. I think that gold is far better proxy, or approximation, of this Stable Value ideal. And, as it turns out, everybody also agrees with me. The whole monetary history of humanity from prehistory to the present has been the abandonment of all other commodity bases for currencies, except for gold.
If you don't just assume upfront, without argument, that commodities are better than gold, and you actually think about it, you usually conclude that gold is superior, just humans always have for thousands of years. That is why the people that think about it become gold standard fans.
But, let's set that aside and talk about something else.
If a currency either rises or falls in value a lot, this will likely be expressed in commodity prices. Commodity prices go up and down a lot for their own non-monetary issues, but if a currency rises a lot (deflation) commodity prices are nearly certain to fall, as measured in that currency; and if a currency falls a lot (inflation), then commodity prices are nearly certain to rise, as measured in that currency.
However, the rise in commodity prices (vs. a currency) tends to lag the rise in the price of gold vs. that currency. In other words, the commodities price signal has about a year lag, compared to gold. 
It doesn't always have to be exactly a year lag. In a situation where people have become accustomed to declining currency value, the adjustment process becomes shorter and shorter, until it may be more like a one day lag in a hyperinflationary situation. But, where there is an assumption of stable currency value, where the "Money Illusion" is strong, the lag can be about a year.
We saw this recently. Gold moved above $1350 definitively in late 2019, eventually continuing to over $2000. This was a clue that commodity prices would follow.
Commodity prices were flattish in 2016-2019, but instead of going up like gold (reflecting Dollar weakness related to extreme "easy money" from central banks), they crashed lower, reflecting supply/demand conditions for commodities in the middle of the flash-recession caused by Covid. Was the Dollar 'strong' (low commodity prices)? No, it was weak (high gold prices) due to the huge expansion of the Federal Reserve's balance sheet taking place at that time.
Gold was telling you that, not only would commodity prices regain their 2016-2019 plateau when supply/demand conditions for commodities normalized, they would go a lot higher, reflecting the lower value of the Dollar (takes more Dollars to buy an ounce of gold). But, there was a lag. Gold peaked in mid-2020, while commodities peaked in 2022.
This was important, because in September 2020, when commodity prices were barely off the bottom, you could see gold making new highs above $2000/oz. Time to bet big on commodities.
Gold told you the correct story, while commodity prices were not only affected by non-monetary factors (recession), but also, lagged by many months.
What if the Federal Reserve looked at commodity prices, not gold?
They would have to conclude that, even despite their really over-the-top expansionism in 2020, they were not nearly expansionary enough! They would have to expand a whole lot more. This would of course have been followed, with a lag, by not only the substantial CPI 'inflation' and commodity price rises that we saw, but a whole lot more of it! And then what? The Federal Reserve would then have to contract, a whole lot more than they have; and they would also probably have contract way too much, because of the lag in commodity prices, which would have been whiplash both ways. This would have been very unpopular politically, and their commodity-price-targeting system would have been thrown in the junkpile, much like Paul Volcker had to give up Monetarism in 1982 when it was a proven failure.
Already by mid-1981, gold had a huge decline from its levels of late 1980. The 1-year average ended 1980 around $650, and in mid-1981 US Dollar gold was around $400. That raised red flags among the gold guys, who were yelling about "deflation" and upcoming recession even as the CPI was still over 10%, and commodities had barely budged.
Throughout our experience since 1971, a commodities basket has failed not only in magnitude, but also in time. Gold has been a much better indicator of currency value than commodities.

Formerly a chief economist providing advice to institutional investors, Nathan Lewis now runs a private investing partnership in New York state. Published in the Financial Times, Asian Wall Street Journal, Huffington Post, Daily Yomiuri, The Daily Reckoning, Pravda, Forbes magazine, and by Dow Jones Newswires, he is also the author – with Addison Wiggin – of Gold: The Once and Future Money (John Wiley & Sons, 2007), as well as the essays and thoughts at New World Economics.

See the full archive of Nathan Lewis articles.

Please Note: All articles published here are to inform your thinking, not lead it. Only you can decide the best place for your money, and any decision you make will put your money at risk. Information or data included here may have already been overtaken by events – and must be verified elsewhere – should you choose to act on it. Please review our Terms & Conditions for accessing Gold News.

Follow Us

Facebook Youtube Twitter LinkedIn



Market Fundamentals