This article is a follow-up to last week’s installment. In the present piece, I’m going to tie up one loose end from last week, and then I am going to respond to George Gammon’s YouTube critique of my Part I from last week. It should go without saying, that if you haven’t yet read my Part I, you absolutely need to go do that first, for the present piece to make any sense.
To refresh your memory, in the prior article I walked through four different scenarios, each having its associated map of the world. In Maps 1 and 2, there was no leverage; people around the world directly held gold bars as a reserve asset, without any paper claims on the gold. In Map 1, there was widespread global demand, corresponding to our (roughly) current gold price of $2,500/oz.
Then in Map 2, I supposed that the demand to hold physical gold disappeared in South America, Africa, and Australia. This drop in demand wouldn’t physically destroy the gold bars, and so if everyone in the southern hemisphere wanted to sell, the gold bars would migrate up into vaults in the north. Map 2 thus showed the same number of gold bars as before, but now they were all stored in North America, Europe, and Asia. The global price of gold had fallen to $2,000/oz., to reflect the fact that it would take a lower price to induce the people in the north to expand their inventories.
So far, so good. In Map 3 I introduced the complication of fractionally-backed paper claims on gold bars, meaning that investors who wanted exposure to gold as a reserve asset didn’t need to hold the physical metal on hand. As in the first scenario, here I assumed that the global price of gold was $2,500/oz. That was the price at which “supply equals demand,” but crucially in this context, the quantities refer to the notional amounts represented by the paper claims on gold. Furthermore, I assumed that the paper notes were backed up by actual gold in the vaults at a 1/3 reserve ratio.
Finally, in Map 4 I once again assumed that the demand to hold gold as an asset disappeared in the southern hemisphere. So all of the gold and paper claims on it migrated north. Again, I assumed the global price would fall to $2,000/oz. in order to induce the northerners to expand their holdings.
My point with the above was to show that the “leverage” in the gold market was a complication that didn’t affect the basic result of the pure commodity scenario: When the global demand to hold gold as a reserve asset drops, that obviously pushes down (other things equal) the global price of gold. By implication, I argued that the same logic flows through to the case of USD reserves (which can only be created by the US authorities) which foreign institutions can use as the base upon which to pyramid legal claims on USD. As with gold, so too with the dollar: If foreigners collectively reduce their willingness to hold USD-denominated assets, that will (other things equal) reduce the price of the USD. The leverage in foreign markets is a complication, but it doesn’t change the direction of the result, the way Brent Johnson, George Gammon, and others have been arguing.
After I posted my article last week, I realized I probably should have made a concession to Brent Johnson that would have clarified my perspective (as well as making my discussion more realistic). Specifically, in any scenario where foreigners reduce their desire to hold USD-denominated assets, we are likely going to see (especially in the medium-run) a reduction in the leverage ratios. This would indeed reduce the total quantity of “US dollars” (broadly measured) in the global financial system, and hence would mitigate the fall in the USD price, compared to a situation where there was no original leverage.
Now I want to be clear: Where Brent and George go wrong (I claim) is that they think the deleveraging of foreign USD credit operations happens simultaneously with a drop in foreign appetite to hold USD assets. This is not correct. Once a foreign credit transaction allows for the creation of new “dollars”—by simultaneously creating an asset and liability, denominated in USD—the mere fact of the asset holder trying to unload it, doesn’t automatically unwind the credit operation. So the immediate impact on the dollar price hits, and only gradually might the credit transactions begin to unwind, shrinking the total quantity of USD on planet earth.
Consequently, the “accordion” nature of the foreign quantity of USD at best acts as a shock absorber. In this light, it would have been better if I had posted the following Map 4B, rather than my original Map 4:
Map 4B: Reduced Demand to Hold Gold Assets, Less Leverage, World Price $2,300/oz.
In Map 4B, the number of gold bars is the same as in the other scenarios. The difference is that the reserve ratio has risen from 1/3 to ½. That is, each gold bar in the vaults now has only two (not three) paper claims on it, which are held by outside investors. Because the total quantity of “gold” (broadly defined, including paper claims on it) has fallen, the fall in the world gold price is arrested. Rather than falling from $2,500 down to $2,000, in this revised scenario it only drops to $2,300/oz.
(Note for purists: If I could do it all over again, I would have fiddled with the starting scenario to make it easier to show an increase in the reserve ratio that still allowed—with the migration north of the physical gold bars—for the total quantity of notes held in the US, Europe, and Asia—to increase, even in Map4B. That’s what you would need, to explain why the world price of gold fell from $2,500 down to $2,300. But because of the particular numbers I used last week in Part I, now when I am trying to show nice round numbers, unfortunately my map shows the total number of notes held in the north dropping.)
As a final note, let me spell out the specific mistake George Gammon made, when he was trying to methodically demonstrate the flaw in my own logic. About halfway through his video, he referred to the following diagram:
The point of George’s demonstration was to illustrate the three possible foreign holders of USD-denominated assets and/or liabilities. The guy on the right held USD assets, the guy on the left held USD liabilities, and the guy in the middle held a mixture. George then wanted to see what would happen to the price of the USD (measured against these people’s currency) when these three people reduced their willingness to hold USD.
George started with the guy on the right. If the guy tried to use his USD assets to buy gold, George agreed that (other things equal) this would weaken the USD.
The guy in the middle was a wash. If he just wanted to get away from the USD, he would use his assets to pay off his liabilities. I could quibble with this, but I don’t need to. I’m happy to stipulate it.
But it’s the guy on the left that is the problem. Here, George matter-of-factly says that if the guy “doesn’t want to deal with the USD ever again” (or words to that effect), then he would sell his Rupee assets (shown by the “R”) to buy USD, in order to pay off his USD liabilities. Yes, that’s true, but that’s not the scenario we are discussing.
Remember, the whole point of this great debate has been: “What will happen if foreigners want to dump their dollar assets?” They are doing this, presumably, because they are less confident about the strength of the dollar going forward. So yes, in this scenario, the guy on the right would want to swap his USD asset for something else.
But if, by assumption, we are saying foreigners have become more skeptical about the prospects for the dollar, then someone who is net short dollars isn’t going to all of a sudden cover the short! No, if you are starting out holding “negative dollars,” and then new information makes you even more bearish on the dollar, you aren’t going to close out your position.
Once you make this correction to George’s discussion, you will see that his conclusion falls apart. In the stipulated scenario of these three foreigners wanting to reduce their USD denominated assets, or if you prefer, their net holdings of USD, then the aggregate result is a fall in the price of the USD.
Dr. Robert P. Murphy is the Chief Economist at infineo, bridging together Whole Life insurance policies and digital blockchain-based issuance.
Twitter: @infineogroup, @BobMurphyEcon
Linkedin: infineo group, Robert Murphy
Youtube: infineo group
To learn more about infineo, please visit the infineo website.