Consumable Digital Commodities
Our team at Collab Crypto spends a lot of time thinking about markets. We concur with Naval Ravikant’s statement that blockchains will replace networks with markets, so we’ve been thinking carefully about what tradable assets will look like in the future and how markets will form around them. Particularly, commoditized assets.
A common meme assigned to Bitcoin is “digital gold,” a commodity that functions primarily as a store of value. However, there are many other types of commodities in the economy, the vast majority of which are consumable inputs. Every day, billions of dollars of commodities like oil, sugar, and wheat are traded between producers and consumers of these goods, and billions of dollars more of purely financial transactions occur in the derivatives markets that exist around these commodities. A thought experiment we’ve been playing with is, “what do consumable commodities look like in the digital age?”
In the 20th century economy, companies used commodities like steel, cotton, and fossil fuels to manufacture products. In the 21st century, many products are digitized and the production function calls for different types of commodities that are based around data. This data needs to be stored, processed, and transmitted. We believe these needs will form the basis of digital age commodities, and markets will form around them.
It used to be the case that the phrase “it’s going to be commoditized” meant the category was unattractive for venture investment. In crypto, now it means the opposite because we can invest in the markets that aggregate commoditized supply.
In many ways, consumable digital commodities aren’t that different from consumable tangible commodities. Consider raw compute. There are producers – those that have invested in GPUs, for example. And there are consumers – those that require compute power for execution; proof of work mining and artificial intelligence applications are examples. And the goods being transferred from producers to consumers are fungible, with consumers placing emphasis on the price of the good rather than the source of the good.
Historically, these digital commodities have been supplied primarily by centralized entities like Dropbox and Amazon Web Services. The rise of protocols that aggregate decentralized supply and push payments to these suppliers is a game-changer because it promises to drop the cost of these goods substantially for the end user by creating a competitive market among suppliers.
What does it mean for a product to become commoditized? Effectively, it means the good is homogeneous and can only be distinguished by price. Commoditized goods induce high levels of price sensitivity in buyers. It’s the reason people go out of their way to save a few cents per gallon on gasoline. It’s also the reason centralized suppliers of homogeneous digital consumables will have trouble competing against a decentralized supply where price is a market outcome. Suppliers become price takers rather than price makers. As markets move towards perfect competition and suppliers no longer have market power, price drops to the marginal cost of production in equilibrium.
But real world markets are rarely, if ever, perfectly competitive. Even if the good is homogeneous, suppliers can band together to generate market power that allows price setting. OPEC is an example within traditional commodities. A more common violation is that goods that might appear to be commodities are heterogeneous. An example from traditional commodities is corn. Not all corn is the same - there’s sweet corn, popcorn, and a variety of others with different characteristics. Within these categories you have further differentiation like organic sweet corn, and some consumers are willing to pay a premium for these characteristics. However, the type of corn that is used in industry is called dent corn, and it accounts for the vast majority of all corn production. This is the commoditized good, and differentiation is mitigated by a set of standards published by the USDA.
Producers have an incentive to create product differentiation in the minds of consumers, to allow price setting and the accumulation of economic rents. Branding is one method of differentiation and it’s one of the reasons that trusted brands often constitute great investments. For many digital goods, the importance of trust cannot be understated. Amazon is winning “commodity-level” compute with AWS not only because of low cost but also based on the reputation of the Amazon brand. Their customers trust that Amazon is executing the code faithfully and trust that they won’t arbitrarily lock the client out or cheat them. Similarly, file storage on Dropbox or Google Drive is currently differentiated because we trust that the data we store there will be available on demand without fail when we want to retrieve it.
Trust and standardization will take different forms in crypto-incentivized markets. As Muneeb Ali and Chris Dixon have articulated, we’re moving from “Don’t be evil” to “Can’t be evil.” Protocols creating markets for data storage will need to find ways to ensure the ability to retrieve either cryptographically via proofs, or through other mechanisms to deter cheating on the supply side. The US government created the standards for corn, but standardization in digital commodities will be enforced by the protocol.
One of our recent investments in this category is Nebulous, the parent company behind Sia, which is the leading decentralized storage platform. Cloud storage is a commodity that should be homogeneous but is currently dominated by a few existing players who have created product differentiation based on security and UI/UX. Against these centralized providers, Sia stands out in price, privacy, and reliability. Coming back to the price sensitivity of digital commodities, Sia is over 10x cheaper than AWS S3. To store 1Tb of data, Sia’s current cost is around $1-2 per month, whereas AWS S3 costs about $23. Moreover, by leveraging blockchain technology Sia offers users true privacy and control over their data, without relying on centralized providers. Since its launch in 2015 Sia has stored over 4 Pb of data across 495k smart contracts, and currently has 2 Pb of available storage capacity on their network.
As the UI/UX continues to improve and users begin to view data storage as a homogeneous good, we believe they will increasingly trust markets like Sia to procure this resource. We’re excited to support Nebulous alongside other investors like Bain Capital Ventures, Bessemer, A.Capital, and Dragonfly. Further, we’re enthusiastic about markets for consumable digital commodities generally and expect more investments along this thesis as opportunities present themselves.
Bryan Chang and Matt Lucas also contributed to this article.