A few months ago, I wrote an article and called it “What we should be building in Africa.”
The tone was very pessimistic, and the idea was that intra-African cross-border payments, especially for significant volumes, are hard because every payment has to go through the dollar. Financial institutions have to prefund dollars, and in most markets, these dollars are scarce. The primary problem is that there aren’t enough dollars in the system because the country isn’t earning enough dollars from exporting goods and services.
The diagnostic is somewhat right. Even though that’s true, it doesn’t mean building a cross-border payments company is a dead end. In fact, that reality creates a significant business thesis because dollars are in very high demand, and no matter how broken the system is, there is usually a business case in that broken system. The business exists because the system is broken, and if you can be the alternative to the broken system, then you have a business.
A good way to think about this is the parallel market.
When we talk about the dollar rate against the local currency, there are usually two rates, and one usually tells the truth more than the other: the central bank rate, CBR, and the parallel market rate.
When dollars are scarce in the system, banks will not want to serve everyone who needs dollars. They prioritize the government and big corporates, and if you are an SME and need dollars to import something from China, you will be the last to be served.
Anywhere there is demand, there is a market.
Unofficial channels, usually called parallel markets, emerge where SMEs or anyone the bank won’t serve buy dollars at a premium.
The parallel rate is usually a little higher than the CBR, and the difference between the CBR and the parallel rate shows how scarce dollars are in the system. The higher the difference, the greater the scarcity of dollars. Scarcity means demand, and demand means a market for anyone who can meet that demand.
The moat in African cross-border trade is therefore being able to access dollars cheaply.
The winners will be those who can access dollars cheaply.
Now back to East Africa. When we look at the payment landscape here, we have telcos. They have done a really good job of fixing consumer cross-border payments. Sending a few dollars across borders from Tanzania to Kenya is fast using M-Pesa or Airtel.
But if we move the volume up to thousands of dollars, which is the kind of money SMEs need to settle across borders, they have to use correspondent banks. That means sending money from Kenya to Tanzania requires the funds to go through an account in New York — i.e. a dollar detour.
The Kenyan bank will instruct its correspondent bank in New York or London to pay out of its dollar account there, called a nostro account, which has to be prefunded with dollars. The bank will credit the Tanzanian supplier’s Tanzanian bank dollar account. The Tanzanian bank will then convert the dollars to Tanzanian shillings and credit the recipient.
As you can see, banks have to keep nostro accounts prefunded with dollars. If a lot of people want to make international payments, banks need to have enough dollars prefunded. It’s not scalable.
If demand for international payments is greater than the dollars banks have prefunded, banks have to make a choice about who to serve first and who to serve last.
That is the dollar liquidity problem.
Good technology, including blockchain rails, doesn’t really remove the liquidity problem. Liquidity cannot be deleted. It can only be moved up to higher balance sheets that are capable of handling it.
Cross-border payments technology on blockchain rails still has the same issue. We need LPs, and the LPs usually have to have dollar liquidity too, but in their case, USDC/USDT liquidity.
If I want to send money from Kenya to Tanzania using blockchain rails, I need a provider who accepts my Kenyan shillings and converts them to USDC/USDT. That provider needs USDC or USDT sitting somewhere. They then send the stablecoin dollars to a provider in Tanzania, who needs to sell the dollar stablecoins, get the Tanzanian shillings, and credit the recipient.
So the liquidity constraint is still there.
The reason I think blockchains and stablecoins are significant in this evolution of money is that we can have all these platforms for cross-border payments. Stablecoin dollars are at least easier to access than real USD, although they are pegged one-to-one. You don’t need a bank account or heavy compliance to access USDC, which makes it a very accessible form of dollars for many people.
And the good thing is that you can redeem 1:1 for dollars, although some issuers have a minimum redemption amount. For Tether, it’s $100k USD.
So if cross-border payments companies can actually get a cheap source of dollars, I think there is a real business here. However, getting cheap access to dollars requires significant skin in the game.
I think stablecoin companies are actually good for banks. Banks don’t need to build the stablecoin rails themselves. If they want to serve an SME that wants to make a cross-border payment, it makes sense to route that transaction to stablecoin companies using blockchain rails and have them provide the liquidity, rather than using the prefunded dollars in their nostro accounts.
Coming back to East Africa, a lot is happening right now. We have moved from importing energy to producing it. We’ve got refineries being built, railways, e.g. the TAZARA, etc. I have been following the money over the past few months, and it’s following physical infrastructure.
All of this physical infrastructure being built will need fuel and machinery over the next few months, and this fuel and machinery will need dollars since it has to be imported. There’s going to be real demand for dollars and cross-border settlement tools in East Africa.
DFIs are also financing intra-African trade heavily. Afreximbank is pushing towards $40 billion in intra-African trade. But the money still settles the old way: through slow and expensive correspondent banks.
Now the opportunity is there, and earlier I said that the winners will be those who can access dollars cheaply.
The way to do this is to get dollars from people with unused dollars on their balance sheets who are looking for yield. Hedge funds, credit lines, etc.
Two things will determine the success: how you structure the offer to make it less risky for these institutions with dollars, and the yield they get.
The former is the real hard one. For people to give you their dollars, you need to structure the offer as an asset that is less risky for them. That requires trust and skin in the game.
The latter is doable. Payments, especially cross-border payments, happen in intervals of, say, two weeks. If, in every two-week cycle, you charge a certain percentage, say 0.6%, and your LPs get half of it, i.e. 0.3%, and payments happen in intervals of two weeks, then at the end of the year the returns are around 8% APY. So it’s possible to beat US T-bills. This assumes that transactions can happen in those intervals, but it’s definitely doable.
I think getting dollars cheaply comes down to how you structure the offer to make it less risky and the yield you can provide.
If you can do that, the rest is not as hard, and you have a real shot at success.
With blockchain tech, though, there are other ways to ease the dollar liquidity crunch. That's a topic for another day, though!