(、ᐛ)ヘ_/ Nadim's Blog

Blockchains for digital payments

The nature of money is, and has always been, a topic of contention. It has certainly evolved over time. If we are to assume that money is credit (central bank and commercial bank debt), then it would be fair to say that money is primarily tracked digitally today. Cash is merely an IOU document that can't be redeemed for anything except itself, it solely represents this debt relationship. The difference between money at the bank and cash is simply the degree of separation with the central bank, as you are holding commercial bank debt, which itself is derived from central bank debt.

Payments are perhaps the most fundamental economic activity. Money's primary function is to be a medium of exchange. Bartering is what preceded payments: beads, shells and metals were commonly used. The emergence of credit changed everything: it introduced new intermediaries, immediate liquidity and a new relationship with the underlying store of value. Banks would safeguard various commodities and issue notes and cheques. It became easier for people to exchange these notes directly instead of redeeming the more impractical underlying store of value. These notes came to be recognised as money, and banks realised that they could issue more notes than they had reserves for, simply because a bank run became unlikely now that everyone was accepting the notes as money and recognising them as such. It became paper money. As time passed, the link between these notes (cash) and the underlying store of value (primarily gold at this point in time) was abolished. The limited gold supply was restricting the issuance of new money, which was limiting growth, and so money became solely backed by trust.

Consumer payments: from paper to digital

The shift from paper payments to digital payments started with the invention of credit cards. The first example of this was the Diners Club Card in 1950. American Express introduced the digital stripe in 1971 and credit cards kept growing in popularity throughout the 20th century.

Then came smartphones and the internet, reducing the friction for retail payments even further. Digital payment platforms became giants. The developed world went cashless and emerging economies are quickly catching up. Note that various different payment systems are emerging in different parts of the world. It is a mix of state-led initiatives and incumbent banks adopting technology. The trend is clear, however: the world at large is going cashless, and digital payments are quickly spreading to emerging economies.

◞(、ᐛ)ヘ_/ Nadim's blog

There is a contrast between the bank/card model in the Western World and the closed fintech apps of China. Emerging economies sit somewhere in-between. The rich world is getting there with instant bank-to-bank payments schemes appearing in the US (FedNow) and in Europe.

Multilateral linkages can further the adoption of these schemes. UPI wallets (India) can be used across Asia. Alipay is accepted across the world. Grab-Pay (Malaysia) is dominant in South East Asia. The systems are arguably becoming more open: UPI provides a much less siloed alternative when compared to Alipay, as it doesn't lock users into a single interface (can be used with Google Pay, PhonePe etc.)

AliPay and WeChat Pay currently process 90% of Chinese digital payments. UPI accounts for 52% of all digital transactions. Pix accounts for 30% of Brazil's electronic payments. M-Pesa is used by over 90% of Kenyan households.

◞(、ᐛ)ヘ_/ Nadim's blog

The various digital wallets that exist across the world are far from limited to P2P payments. They often include programmable payments, financial services and lifestyle services, mimicking both banks and credit cards.

◞(、ᐛ)ヘ_/ Nadim's blog

Resilience of the credit card model

No one likes credit card fees. Everyone loves credit card rewards. These two elements balance each other in a well orchestrated loop with a number of intermediaries.

The simplified payment flow outlined below goes through the lifecycle of a representative (Visa Signature, US) credit card transaction, with some intermediaries excluded. In this process, it's important to point out that the payment is not settled instantly: the merchant gets a payment guarantee from the acquirer and receives money 1-2 days later. Payments are settled in batches but credit card networks advertise their throughput as if payments were final.

Depending on who you ask, this has the advantage (or disadvantage) of making chargebacks possible.

◞(、ᐛ)ヘ_/ Nadim's blog

It would be easy to assume that this system is ripe for disruption. The Western bank/card model was created before the age of the internet and has mostly remained the same for half a century. Its economics are predatory and hurt smaller merchants with no negotiating power, while bribing customers (who are bearing higher prices as a result) with rewards and benefits.

This also means that people with limited or no access to credit cards are paying higher prices to cover others' rewards. On top of that, banks are incentivised to issue cards with networks that are charging higher fees, as they get paid more in the process. In countries with no caps on credit card fees (mainly the US), credit card networks are competing to raise fees. Their profit margins are as high as 40%-50%. They have enough consumer adoption to force retailers into opting-in.

The case for blockchain

Existing payment networks and platforms around the world are, for the most part, siloed. While there have been efforts to make them more interoperable and allow them to communicate more seamlessly, operators are incentivised (economically and politically) to keep their networks segregated or tie them to specific platforms.

Blockchains have the potential to introduce better interoperability, as well as clearer standards for privacy & data transparency and instant finality for payments (if desirable).

Blockchains also offer various programmable layers, making them extremely flexible:

Tokens can flow freely across a network (defined as a set of validators). Programmability can be implemented at the token level (restricting token flows but fragmenting liquidity and breaking composability), protocol-level, blockchain-level or network level.

Heterogenous blockchain networks, such as Avalanche, go a step further by enabling interoperability between separate networks, which is achieved by introducing the concept of a primary network that groups all validators together. This has the potential to allow money to flow between fundamentally different networks. This primary network can also reduce fragmentation by serving as a liquidity hub for various currencies.

◞(、ᐛ)ヘ_/ Nadim's blog

I could dive much deeper into this, but my goal is simply to illustrate how blockchains enable programmability at each level without sacrificing composability and interoperability. This cannot be said of existing systems: it means that each country or payment processor could design their own proprietary network that complies with local regulation without sacrificing the ability to send money to another network.

You can imagine a design where a system could route payments through various networks and protocols to balance the load or optimise for cost or speed. In theory, existing payment networks and schemes such as UPI, M-Pesa, SWIFT, CIPS, FedNow, Visa and others could be adapted to fit such an architecture without sacrificing their uniqueness and sovereignty.

The need for fiat ramps

Most of the benefits derived from using blockchains for payments (ie. non-custodial, instant finality, traceability, interoperability) are only applicable if the system is not reliant on traditional payment rails (ramps to fiat). On-ramps and off-ramps introduce intermediaries, delays and fees that are often less competitive than those of established P2P payment platforms.

This is why I that blockchains can provide the most value in payments today by focusing on emerging countries with a lack of banking infrastructure and high reliance on cash, echoing the example of M-Pesa in Kenya.

M-Pesa relies on agents that are authorized to offer cash and redemption services to top-up and withdraw from M-Pesa accounts and are required to perform KYC and register with local regulators. M-Pesa is owned by Vodafone and Safaricom, allowing the network to be seamlessly tied to sim-cards that function as accounts.

A similar system could be replicated and improved upon in the context of a blockchain payment network. KYC checks could be done once by a user and become interoperable across agents. Agents would sign up with stablecoin issuers (which could be commercial banks or central banks) to be able to mint and redeem stablecoins on-chain. You could even imagine a world where each agent is trusted to custody cash and provides accounting of their reserves (with some external auditing). Spreading the risk of having a central custodian (which is how stablecoin issers are set up today) might be advantageous, as each failure would be limited in scope.

This is only crucial in the beginning: as such a network gets more adopted, the need for fiat ramps would progressively disappear. If a merchant can buy goods & services with stablecoins held on-chain (or even pay taxes & salaries), they would no longer need to rely on off-ramps. The same applies to individuals that could receive their salary directly in their wallet and set up automated payments for rent, savings, remittances or taxes.

Each participating country could have their own network customised to fit their needs, with fees in their local currency and rules on who can validate the network, issue assets or deploy contracts. The beauty of blockchains is that these networks would remain interoperable: any other blockchain could send assets to this network, as long as it allows them to do so. This would be the same as having a user in China sending money through AliPay that a user in Kenya can redeem in cash through M-Pesa. Each would have to adhere to the rules of the network through which they're issuing the transaction (including rules for on-ramping and off-ramping, KYC, limits on transactions etc.)

Dealing with multiple currencies

Cross-border payments often involve multiple currencies. While it is true that many emerging (or failing) economies are dollarized and could rely on USD-pegged stablecoins, there will still be situations where local currencies are preferred.

Currency swaps could either happen on-chain or at the time of on-ramping/off-ramping (where agents would facilitate the transaction). The latter option is mostly required in countries where the local currency's value is extremely volatile, making it much harder to track its price in real-time. It's worth pointing out that decentralized exchanges are capable of price discovery, and with enough participation from local actors, could act as a source of truth for the price of currencies that are mainly traded in illiquid black markets.

It will be hard for on-chain exchanges to compete with their off-chain counterparts if liquidity is too fragmented, this is why a specific FX network could be designed as a public good for the larger ecosystem. There are many competing architectures for decentralised exchanges: automated market markers (AMM), concentrated liquidity AMMs, fully on-chain central limit order books, stableswaps and stable pools with oracle pricing etc. I will leave it at this but might cover each of these options in-depth in a future post. My point is mainly that acceptable FX pricing can be achieved on-chain as long as a central liquidity hub exists.

Competing for adoption

It is worth considering how a blockchain payment networks could learn from the credit card model and use its playbook to incentivise adoption. If you compare the credit card payment flow with a typical blockchain payment flow, you'll find the blockchain payment to be much simpler. If we use the same example of a USD 100 purchase between a consumer and a merchant: the consumer would pay USD 100.1 to cover the network fee and the merchant would receive USD 100. In most cases, the fees levied by the blockchain are redistributed to validators.

If on-ramping and off-ramping is necessary, additional (variable) fees will be incurred by both the consumer and the merchant. Blockchain fee structures are completely customisable (can be used to pay validators or be burned, can be fixed or dynamic etc.) but are generally dependent on the complexity of the transaction rather than the amount. This means that moving larger amounts of money on-chain is generally much cheaper than doing it through traditional payment rails.

On fees alone, modern blockchains can be very competitive and introduce a much healthier model. Combining this with a robust loyalty program could be the recipe for success. What would this look like? I don't know. I'll write more ideas on this specific topic in the future.

My goal here was just to pique your curiosity and cover a few of the reasons why blockchains should be considered as a better technological solution to enable electronic payments. If you're building anything related to this, please reach out!

Enjoyed this blog post? Consider subscribing: