5 min read

Tokens as Commodities, Not Money

Tokens as Commodities, Not Money
Photo by Shubham Dhage / Unsplash

A while ago I was in a group discussion about how to make gold a decentralised currency on the internet. Ultimately we couldn’t find a structure which accommodated decentralised issuance while remaining a fungible currency.

Gold itself might be uniform, but the various forms it takes are merely comparable. If nothing else, a 400oz bar in America will have a fractionally different price to one in South Africa. Why else would it ever be shipped? All ‘loco London’ good delivery bars [“good delivery is a system of accreditation and standards for the gold bullion market set by the London Bullion Market Association”] on the other hand do have the same price. The cost of moving gold between good delivery vaults _within London_ is incidental compared to its value.


On delivery of FCOJ under Exchange contract, the maximum amount of bottom (sinking) pulp shall be twelve percent (12%) on the initial test and the percentage of recoverable oil shall not be less than .005% nor more than .020%.

So the global trading system has been dealing with these issues for a while. Commodities aren’t commodities by their nature, but through social conventions. Certain differences are ignored, some are placed into objective scales, others are smoothed over by customary adjustments to price or quantity.

For example, every good delivery bar has a known purity. Providing the purity, or fineness, is over 99.99% it is acceptable, however its price will be set in terms of its fine gold content. A slightly less pure bar is worth slightly less, but in a completely predictable way.

A similar system exists for Frozen Concentrated Orange Juice where the value is adjusted by sugar content. And although you might believe oil is a uniform commodity, its value is graded and adjusted significantly according to another definition of “sweetness” [lack of sulphur]. The Indian NDEX has a contract for Turmeric.

So, I would hold, the way forward lies not with inventing elaborate schemes to make physical assets identical. It is in their very nature that they are not. But in inventing interlocking schema to make them comparable. So token are fine and good, but to represent actual goods they need to be parameterised.

Returning to our gold example, we might choose to define a bar of gold as something like…

Gold<Locus -> London, Bar -> Weight, Fineness, Origin>
Spec { Fineness < 0.9999, Weight * Fineness > 400oz, LBMA Refiners ∋ Origin }

We might equally say…

GoldCoin<Location -> Birmingham Jewelry Quarter, Coin -> Krugerand, 0.95>

We have computers, so tracking a few million gold coins isn’t expensive any more. Perhaps I could simply buy the 300 cheapest coins located in any EU member state. If I wish to get a physical coin, I’ll simply swap one of my 300 coins for one of equivalent value at the nearest coin shop I happen to be close to.

It doesn’t have to be an exact swap. A tiny payment can happen to make good the difference in price between the street I’m on in Canada and the coin I hold in Prague. At a micro-scale, this difference will be barely observable, in fact my smart phone will probably settle it without me. At a macro-scale, there should be a sufficiently liquid market in these differences that my phone can accept prevailing prices as fair.

(Interestingly, the reason that I would do this in a open economy isn’t that I would want flexibility, but that I wished to be exposed to the value of gold coin in preference to bank money.)

It is in showing and defining more of these differences, rather than homogenising them, that the internet/blockchain/IoT/smart phones/link:EDI etc can make a difference to trade.

Suppose we have a bushel of wheat. Here’s what it means to different people.

Farmer Bushel of wheat.
Trucker Loose dry good weighting 20kg for collection at Stockport Farm [a 5 mile drive from me].
Wheat trader Instrument trading at an average $0.65 discount to an in-country 90-day milled wheat future.

Modern economies rely on specialisation; all of these perspectives are equally valid. To make them comparable, we need to introduce standardised dimensions to how we look at the bushel.

For example :-

  • Location [GPS co-ordinates]
  • Weight [kg]
  • Bulk [cm^2^/kg]
  • Handling types [dry goods, loose, containerised, etc…]

And in this light, the trucker’s service can be seen as the transformation of a commodity along the location dimension, which can itself be priced as a commodity. So we might consider the material economy to be a huge collection of unspent physical resources and potential transformations.

We might choose to input one truck, some diesel and a bushel of wheat in Scotland at 7am, and output both a truck and a bushel of wheat at 2pm in Derbyshire. You might assume that someone drove the bushel in a truck burning some diesel along the way. In practice though, no-one would care if the bushel was burnt on the spot, the diesel siphoned off and sold, and another bushel bought for delivery. It is all the same to the many faced God of market capitalism 🙂

(There is a UK Feed Wheat Future(link is external) already.)

Supposing we treat the entire economy in this fashion. As a factory manager, I could simply advertise my desired inputs and let others bid on them. Most of those bidders would probably be brokers, who would then arrange the production and movement of goods by others. I might sell my production in a similar fashion. This might seem like Uber-for-Everything, and on a short term basis it would be.

On a long-term basis, things are much more interesting. If I require a shipment of steel in Newcastle in 100 days, there will be a price for this. (There is a price for everything in market economy.) There will probably also be a break clause allowing my counter-party to skip delivery and pay compensation instead. Under these circumstances, I’m essentially buying insurance against those goods not being available. The efficient price givers will be those _most able to predict the price_, not necessarily an actual supplier.

We could imagine a few primitives for this sort of tokenised economy :-

Package(Shipping Container, Collection<Things>) -> Shipping Container*
Move(Thing[NYC], SF) -> Thing[SF]
Process(Oil) -> [Kerosin, Petrol]

And, of course, a resource that transforms commodities might be used for all sort of similar transformations, but it can only be consumed in one transition at a time. There is one truck, or one oil refinery, so it is itself “consumed” with every transaction. (Interestingly this is how bitcoin transactions work. You spend the entire balance each time, and just send the change back to an address you control.)

Building up from that a little more we could imagine our specification incorporating options.

// Use my truck to move the bushel, charge me more and leave my truck free. Crash the truck and pay me $74k.
(Bushel<Scotland>, Truck, $100) -> ( (Bushel<Derby>, Truck) OR (Bushel<Derby>, $74k) ) OR ( (Bushel<Scotland>, $300) -> 

We could even see that as utility function with an indifference curve, e.g. this formula tell you how much I value different outcomes, and it is possible to draw lines through that solution space where many different outcomes are equally desirable. That would enable the use of financing techniques from capital markets to be applied further down the chain…


But in summary!