Friday thought experiment: Mac-PESA

I”m very wary of promulgating the “political correctness gone mad” meme, as it is so often a lazy reactionary knee-jerk response to changing times, but I could not resist tweeting about the news that a British police force launched an investigation after a man claimed he had been the victim of a “hate crime” when… a branch of the Post Office refused to accept his Scottish banknote. This incident has now indeed entered our official statistics as a hate crime.

Frankly, this is mental. Scottish banknotes are not legal tender, even in Scotland, as I have explained before. The Post Office is no more obliged to accept a Scottish Fiver than it is to accept Euros, gold or cowrie shells. The story did, however, cause me to reflect on what will happen when, post-Brexit, Scotland votes to leave the UK. Will Scotland then join the euro or create its own currency?

As supporters of Scottish independence insist, once Scotland becomes an independent country, it will be responsible for managing its currency in the same way that every other country that has its own currency is responsible for managing. But how should the Scots go about creating this currency? Surely messing around with notes and coins, other than for post-functional symbolic purposes, is a total waste of time and money.

A much better idea would be to go straight to the modern age and create Mac-PESA, which would be a digital money system rather like Kenya’s M-PESA with with a few crucial enhancements to take advantage of new technology. M-PESA, as a post on the Harvard Business School blog says, is “the protagonist in a tale of global prosperity to which we all can look for lessons on the impact of market-creating innovations”, going on to say that its “roots are far more humble”. They are indeed, and if you are interested in learning more about them, I wrote a detailed post about the origins of M-PESA (and Consult Hyperion’s role in the shaping of this amazing scheme) and the success factors.

The most important of these was the role of regulator: the Central Bank of Kenya (CBK) didn’t ban it. Conversely, one of the reason for the slow take-up of mobile payments (and the related slow improvement in financial inclusion) in other countries was the regulators’ insistence that banks be involved in the development and delivery of mobile payment schemes. The results were predictable. (Here’s a post from a few years ago looking at the situation in India, for example).

Anyway, back to M-PESA. It is an amazing success. But it is not perfect. In recent times it has gone down, leaving millions of customers unable to receive or send money. These failures cost the economy significant sums (billions of shillings), which not not surprising when you remember that M-PESA moves around 16 billion Kenyan shillings per day. So when it drops out, it leaves customers hanging, it leaves agents losing revenue and it leaves the banks unable to transact.

It is now vital national infrastructure, just as Mac-PESA would be.

So what if there were no system in the middle to go down any more? What if the telco, regulator and banks were to co-operate on a Enterprise Shared Ledger (ESL) solution where the nodes all have a copy of the ledger and take part in a consensus process to commit transactions to that ledger?

Do the math, as our American cousins say. Suppose there are 10,000 agents across Scotland with 100 “super agents” (network aggregators) managing 100 agents each. Suppose there are 10m customers (there are currently around 20m in Kenya, which has ten times the population of Scotland). Suppose a customer’s Mac-PESA balance and associated flags/status are 100 bytes.

So that’s 10^2 bytes * 10^6 customers, which is 10^8 bytes, or 10^5 kilobytes or 10^2 megabytes. In computer terms, this is nothing. 100Mbytes? My phone can store multiples of this, no problem.

In other other words, you could imagine a distributed Mac-PESA where every agent could store every balance. You could even imagine, thanks to the miracles of homomorphic encryption, that every agent’s node could store every customers’ balance without actually being able to read those balances. So when Alice sends Bob 10 Thistles (the currency of the independent Scotland), Alice connect to any agent node (the phone would have a random list of agents – if it can’t connect to one, it just connects to another) which then decrements her encrypted balance by 10 and increments Bob’s encrypted balance by 10, then sends the transaction off into the network so that everyone’s ledger gets updated.

You can have a 24/7 365 scheme without having a Mac-PESA system in the middle. When you make a transaction with your handset, it gets routed to a superagent who decrements your balance, increments your payee’s balance, and then transmits the new balances (all digitally-signed of course) to the other superagents.


It would be a bit like making an ATM network where every ATM knows the balance of every debit card. No switch or authorisation server to go down. And if an ATM goes down, so what? When it comes back up, it can resynch itself.

So please, someone challenge me on this. As a thought experiment, why not have Scotland grab a world-leading position by shifting to a Central Bank Digital Currency (CBDC) based on a shared ledger. I very much agree with the Bank of England’s view of such a thing, which is that the real innovation might come from the programmability of such a currency. This would be money with apps and an API, and I would hope that innovators across Scotland and beyond would use it create great new products and services.

Central banks, tokens and privacy

Christine Lagarde, the Managing Director of the International Monetary Fund (IMF) and therefore to a first approximation the person in charge of money, gave a speech in Singapore on 14th November 2018 in which she asked…

Should central banks issue a new digital form of money? A state-backed token, or perhaps an account held directly at the central bank, available to people and firms for retail payments?

This is a question that, of course, interests me greatly. The IMF Staff Discussion Note (18/08) on which her speech is based sets out these two options clearly:

  1. Token-based CBDC—with payments that involve the transfer of an object (namely, a digital token)—could extend some of the attributes of cash to the digital world. CBDC could provide varying degrees of anonymity and immediate settlement. It could thus curtail the development of private forms of anonymous payment but could increase risks to financial integrity. Design features such as size limits on payments in, and holdings of, CBDC would reduce but not eliminate these concerns.

  2. Account-based CBDC—with payments through the transfer of claims recorded on an account— could increase risks to financial intermediation. It would raise funding costs for deposit-taking institutions and facilitate bank runs during periods of distress. Again, careful design and accompanying policies should reduce, but not eliminate, these risks. 

 Or, as I said a few years ago, should the Bank of England create BritCoin or BritPESA?

I’ve written before about the advantages and disadvantages of moving to digital currencies and don’t want to go over these arguments again here. Ms. Lagarde has also spoken about them before, specifically noting that digital currencies “could be issued one-for-one for dollars, or a stable basket of currencies”. Why her new speech was reported in some outlets as being somewhat supportive of cryptocurrencies is puzzling, especially since in this speech she specifically said she remained unconvinced about the “trust = technology” (“code is law”) view of cryptocurrencies. But the key point of her speech is that the IMF is taking digital currency seriously and treating it as something that might actually happen.

(Note that the IMF position seems different to the position of European Central Bank, where President Mario Draghi recently said that they have “no plan to issue a digital currency because the underlying technology is still fragile and the use of physical cash still high in the euro zone”.)

The reason for this comment on her speech is to re-iterate my view on the BritCoin approach. I think Ms. Lagarde is right to mention a state-backed token as an option. The idea of using token technology to implement cryptoassets of any kind, which I have labelled digital bearer instruments, is feasible and deserves detailed exploration. What we might call “digital fiat”* is simply a particular kind of cryptoasset, as shown in the diagram below, a particular kind that happens to be create digital money based on an institutional binding (where the institution is central bank) to national currency.

Cryptomarket Model


Now, nothing in this formulation makes the use of cryptoassets (rather than a central database) inevitable. There are, however, other arguments in favour of using there newer and potentially more radical technologies to implement digital money. One of them is privacy.

(As The Economist noted on this topic, people might well be “uncomfortable with accounts that give governments detailed information about transactions, particularly if they hasten the decline of good old anonymous cash”.)

In her speech, Ms. Lagarde said that…

Central banks might design digital currency so that users’ identities would be authenticated through customer due diligence procedures and transactions recorded. But identities would not be disclosed to third parties or governments unless required by law.

As a fan of practical pseudonymity as a means to raise the bar on both privacy and security, I am very much in favour of exploring this line of thinking. Technology gives us ways to deliver appropriate levels of privacy into this kind of transactional system and to do it securely and efficiently within a democratic framework. In particular, new cryptographic technology gives us the apparently paradoxical ability to keep private data on a public leader, which I think will form the basis on new financial institutions (the “glass bank” that I am fond of using as the key image) that work in new kinds of markets.

* I happened to sit in on the panel discussion on digital fiat at Money2020 China. The discussion was chaired by Carolyn MacMahon from the San Francisco-based Digital Fiat Institute, which I must confess I’d not heard of until today, but intend to visit next time I’m over on the West Coast. In the Q&A I was going to ask about the anonymity issue but go sidetracked with the impact on commercial banks. Next time.