Slouching towards Britcoin, part 2: Quids Game

Welcome back, digiPound enthusiasts.

In part one, we looked at the Bank of England’s difficulties establishing the “need” for a digital pound (“Britcoin”). Here we’re moving on from the cheap thrills of theorisation to the fibrous goodness of pragmatics.

Platform choose

The Bank of England and Treasury are happy to frame their ideas as preliminary. Here’s the Technological Working Paper:

The illustrative model set out in this paper is not an end-state CBDC architecture. Rather, it reflects one possible approach to the architecture for a UK CBDC. There are a number of viable approaches for a UK CBDC which will be evaluated during the design phase.

The BoE calls this concept the “platform model”, which the Britcoin consultation says “would be a public-private partnership that reflects the comparative advantages of each sector” (this model was initially outlined in 2020).

Anyway, the platform model works thusly:

— The Bank of England develops and maintains a “core ledger”, which provides functionality for the digitalpoundsphere
— It creates an application programming interface (API) that allows third parties to plug into the core ledger
— Regulated private companies produce consumer-facing software
— End users do their thang

The companies involved would fall into two broad categories:

1) Payment Interface Providers (PIPs)
2) External Service Interface Providers (ESIPs)

And this is how the BoE sums up their roles:

PIPs function as gateways to the CBDC ecosystem, offering users digital ‘pass-through’ wallets to interact with, and manage, their CBDC holdings. ESIPs might provide nonpayment, value-add services, such as business analytics, budgeting tools and fraud monitoring.

The idea is that PIPs and ESIPs (who we imagine would cross over a lot) would compete to innovate and therefore make wonderfully fantastic products for consumers. 🥰

The BoE says the management of these PIPS/ESIPS (we’re just gonna call them PIPESIPs) would “determine the appropriate revenue models”, but have set out some ideas:

Anyway here’s the overall thing, in graphic form:

The consultation paper also lays out some alternative models (relegated to a box p.61 of document, p.63 of PDF), albeit dismissively (it says they “do not, at this time, appear as suited to our policy objectives”). They are:

— A ‘delegated model’: Rather than everything operating through the core ledger, PIPs would maintain their own ledgers. Transactions between users of the same PIP would be contained to their ledgers, but inter-PIP payments would hit the core ledger. Bank sez no because they’d lose oversight and it might make PIPs’ jobs harder, but notes it would reduce risks on the core ledger as a point of failure.
— A ‘bearer instrument model’: Digital pound ownership is recorded on individual user device, and transactions occur between users without interaction with the BoE. So, kinda like cash, except vulnerable to hacking. Banks sez no because of this risk, along with dangers of double spending. But it notes this could work for offline payments.
— They also note that a reserve-backed stablecoin is not a CBDC, so don’t even think about bringing that up.

Question time

So that’s the basics! We’re now going to have a brief interlude for questions. Here are some of ours:

— Why does the BoE assume the public sector has a comparative advantage in developing the core ledger technology?
— Why does the BoE assume the private sector has a comparative advantage in developing the public-facing technology?
— If one of the few tangible advantages of a CBDC that the BoE is able to imagine is lower fees, how does that interact with suggesting PIPESIPs levy fees?
— If a large PIPESIP fails, the core ledger should prevent the Britcoins being lost to their owners. But would that potentially leave the owners unable to use said Britcoins?

Our hypothesis would be that this public inside/private outside model would not work by itself, because (given the BoE wants Britcoin accounts to be used as primary current/salary accounts), the risks of system failure are very high.

Leaving aside the dangers of the core ledger itself conking out, it would surely make sense for the Bank of England to create its own front-end, effectively becoming a PIPESIP itself. This should provide a PIPESIP of last resort in the event of systemic failure. Alternatively, the BoE or Treasury would have to be prepared to bail out or nationalise critical PIPESIPs.

These are complicated questions that should (hopefully) be better answered along the seemingly inevitable road to Britcoin.

Current affairs

We’ve touched upon the back end and the front end, but what about those damned unreasonable fleshbags that might actually use Britcoin?

The BoE says it would place limits on how many digital pounds an individual could hold, proposing an upper limit of £10,000 to £20,000 (while welcoming views on a lower limit “such as £5,000”). The consultation paper says:

That limit would, however, be set in a way that supports wide usability of the digital pound. For example, users may want to use their digital pound wallet to receive their salary, which may vary over time and may include bonuses or overtime payments. Any limit would also need to consider the roll-over of any balances from the previous month.

The £20,000 cap would be very inclusive (nb the typical size of a UK current account deposit is £800, or £1,800 for longstanding customers):

Once again, though, we find ourselves asking “why?” What practical use would consumers get from receiving their salary in Britcoin and then transferring it to a current account, versus simply having it be paid directly into their current account? The BoE white-heat-of-technology line would presumably be “Oh well those Britcoins would be able to interact with all kind of innovative services”, to which the obvious response is “You know what can already do that? Any current account.”

The clearest reason may be that the UK has, per Bank of America, some £520bn of retail and small business deposits above £85,000, the cap that is covered by the UK’s Financial Services Compensation Scheme. Having this money effectively stored via a CBDC through the platform model would be an improvement because even if the intermediary collapsed, Britcoin holders would maintain their claim. Which is good and all, but firstly suggests holding limits might have to be set substantially higher, and secondly indicates just how inert people tend to be with their savings.

So, three(?) paths:

1. These PIPESIPs are launched, and because they’re operating within an already competitive field and their creators have to make money somehow, they offer little meaningful benefit for customers so adoption is low (possibly low enough to make Britcoin be considered a failure).

2. The PIPESIPs are somehow attractive enough that people suddenly want to use them as primary current account (or as an intermediary to a further account), adoption is high, the current account base at existing banks is eroded (success, but at what cost? Read on to maybe find out . . .)

3. sOmEwHeRe iNbEtWeEn 🤷‍♀️

In the case of (1), that’s an obviously undesirable outcome (waste of money, companies set up to rent-seek fail, blah blah blah . . .) but it’s swings and roundabouts, and would make little practical difference to the world.

Number (2) is decidedly more complicated. BofA, which has examined the potential impacts in the context of its UK banks coverage, reckons this could mess up £11bn of high-quality income that is “fundamental to banks’ business models and ability to remunerate their £175bn in UK equity”.

Its analysts say (across two notes):

The BOE seeks the digital pound to ameliorate the risk of a shift in financial significance, but could facilitate it. We see a digital pound as currently sketched risking large disintermediation of commercial banks by the central bank, for limited practical gain . ..

[AND]

. . . Coming at the banks from both sides, a combination of the two would be really difficult for incumbent banks, if the playing field were unlevelled. Bank capital is expressed relative to assets but is fundamentally demanded to back deposits, a liability. A new architecture in which bank capital requirements stay at current high levels, while competitors offering Britcoin wallets attract little would naturally be problematic for banks, we think.

So, finally, (3). Something easy to overlook is that the relationship between high street banks and PIPESIPs might look a little bit like this:

After all, who has:

— The experience in creating these types of applications?
— The cash to throw at a project that may flop?
— The capacity and incentive to be on both sides of this dynamic?
— An incentive (whisper it) to muscle out/buy out insurgent PIPESIPs to ensure their conventional products remain competitive?

It’s worth noting that (3) is how the BoE expects this to play out. Sir Jon Cunliffe previously told the House of Lords Economic Affairs Committee “[it] is a pretty prudent assumption that 20% of household and corporate transactional deposits move to CBDC.”

He added:

We looked to see what the implication would be for credit spreads and for credit availability in that eventuality. We do not know what the demand will be for a central bank digital currency, but the modelling and the sensitivity analysis around it, which we published, suggests that the impact on banks’ credit spreads might be 20 basis points. Banks would have to adjust. They would have to fund themselves more with long-term wholesale debt. They would lose a revenue stream from payments, which at the moment is quite a reliable and reasonably substantial revenue stream.

If you go back to 1960—I am just about old enough to remember that—nearly 70% of people in this country did not have bank accounts, and 30% of the money held for transaction was in cash. It is not as if we are talking about a banking model that has been there from time immemorial. Banks have adjusted to changing circumstances before. If they are healthy and competitive, one would expect them to adjust in the future, but it would be an adjustment. Whether or not that 20% figure is the right one, it was the most prudent, sensible and plausible assumption we could make. It could be much less than that.

It could be less, sure, or it could be much more. It could also be nothing.