One aspect of the Havven network that is often questioned is the inclusion of transaction fees. One of the core assumptions of the Havven model is that transaction fees are required for a payment network to scale. We believe that concerns over transaction fees reveal a fundamental misunderstanding of two-sided markets. We have previously failed to provide a clear explanation of why fees are not just more efficient, but in fact necessary. This post will explore why the argument that transaction fees are an impediment to adoption is unfounded and that some form of fee is always required for a two-sided market to scale.
Fees are everywhere
Critics of fees are almost certainly already paying fees for other transactions. All VISA/Mastercard/Amex transactions impose a fee on merchants. Merchants, therefore, either adjust their prices to accommodate these fees or apply a small surcharge (often 1–2%) to transactions via credit card networks. In either case, customers and merchants pay millions of dollars in fees every day, whether they realise it or not.
While critics of fees rightly point out they create friction, this does not prevent them from being the optimal strategy for building incentive mechanisms. For example, fees haven’t proven to be a problem for Bitcoin, which uses transaction fees. Bitcoin’s transaction fees reward miners for processing transactions. These fees also ensure the system is protected from spam transactions. Bitcoin even has a second, less direct fee, in the form of the block reward. This is a fee creates inflation, as the bitcoin supply inflates with every block mined. The block reward also ensures the network continues to function regardless of demand for transactions. Bitcoin’s fees are critical aspects of the incentive mechanism and are therefore a necessity even though they create friction for users. Clearly without fees the Bitcoin network would be non-viable.
Fees are an integral part of any ecosystem that generates utility. As utility almost always requires work, this work must be rewarded. Even services that appear fee-less, such as advertising-driven models like Facebook, are typically driven by fees that are hidden from users. Advertising funded platforms scale because there is almost infinite demand for consumer attention. But if you are providing a free service to one side of the market you must have demand on the other side to consume the product of the marketplace. In the case of Facebook, the platform produces an almost infinite supply of consumer attention, this is the product Facebook sells. This model is successful because there are always advertisers willing to pay fees to consume the attention from the Facebook user base. If advertisers were unwilling to pay for consumer attention the Facebook platform would be unsustainable.
One might think of Wikipedia as being fee-less, but it actually relies on donations for funding, a kind of fee that is not effectively enforced. Even Telegram, which is currently fee-less, is working towards an ICO to fund their system and we can only assume that the decentralised version of their service will rely upon a fee structure to maintain consensus.
Fees allow people on one side of a market to ‘purchase’ utility from people on the other side of the market who provide utility. Two-sided markets can be efficient, elegant, and effective. But if the supply on either is imbalanced scaling is not possible. This is critical to understanding cryptoeconomic systems, it is perfectly fine to give something away as long as the other side of the market is generating enough utility and paying enough fees to subsidise demand for the free side of the market.
This concept is critical to the creation of stability within cryptoeconomic systems. You cannot simply provide free utility in the form of a stablecoin without someone paying for it. If users don’t pay then there must be someone else within the system who does. One system that does not appear to sufficiently balance supply and demand to scale effectively is MakerDAO’s Dai stablecoin. In this system Dai is actually a byproduct of another process, the creation of collateralised debt positions. This is a critical aspect of the system which must be understood, even though Dai is free, it is constrained by demand for the issuance of CDP’s within the system. So Dai may be in high demand, but supply of Dai has not been significant since it launched in December. This is likely because there are insufficient incentives for the creation of Dai, because there is less utility for users creating CDP’s than users derive from Dai. This means that although demand for Dai will likely remain high, Dai issuance will never scale to meet that demand. Given that demand for a stablecoin is almost infinite, the demand side is the optimal place to create a small amount of friction from fees to ensure the scalability of the network. If you choose the correct side to charge for utility, then the system should scale. This thesis has been repeatedly demonstrated by centralised payment networks such as Amex and PayPal.
Fees are beneficial
Even though fees reduce demand for a stablecoin, they substantially increase supply. For a stablecoin, supply is critical. The Havven network has a positive feedback loop, where increased usage and demand for the network increases the value of the network. This increases fee rewards for the havven holders, who collateralise the network thus providing scalability, as demand for stablecoins directly increases supply.
Havven uses the paradigm of a centralised payment network but replaces the central authority with distributed collateral holders. These collateral holders stake havvens to provide stability and security for the users of the network. As the network grows and matures, the value of participating as a collateral holder increases. This scalability is why the Amex network is worth almost $100b, because the total network value increases as more users enter the network creating strong network effects.
A scalable stablecoin will be a major benefit to the entire cryptocurrency ecosystem. While the path to a stable cryptocurrency is currently unclear, more innovation, experimentation and honest debate is needed. We believe fees are necessary for scalability. Based on this assumption Havven is taking the proven paradigm of a centralised payment network and decentralising it.