Monday, August 22, 2016

End of a stablecoin

Bitcoin has had another white knuckle year, rising from a low of $350 in January to a high of $780 in June. As I've said many times before, if crytpocurrency is going to make a genuine dent in the world monetary order, it'll only happen when its price isn't so volatile. And one of the best ways to smooth things out is by creating a stablecoin, specifically by choosing to peg a cryptocurrency to a popular existing medium of exchange like the U.S. dollar.

Stablecoin isn't just theoretical, a number of these tokens exist. They fail too. It's worth investigating the recent collapse of one the fledgling stablecoins, NuBits, to look for clues about the dangers and pitfalls of maintaining a stablecoin peg.

Back in September 2014 a developer going by the pseudonym JordanLee set out on a brave attempt at pegging a cryptocurrency, NuBits, to the U.S. dollar. The $1 peg lasted for around twenty months before falling in late May to 95 cents and outright failing on June 8. See the chart below.

Source: Coinmarketcap

My understanding is that at its peak, NuBits was registering a healthy $3-4 million in trading volumes every day. While it never grew to be a genuine medium of exchange--you couldn't buy a coffee with it--the tokens were a popular way to hedge against bitcoin volatility. Had the peg held, who knows what might have happened?

In late May, a single large seller of NuBits emerged, offloading around 10% of the entire NuBits supply in a day or two (you can see the volume spike in the chart above). To keep the price of NuBits at $1, the NuBits team that was tasked with maintaining the peg had to use up a large quantity of its reserves. In the course of events the team decided to widen the peg to $0.95-$1.01 in order to slow the reserve bleed. However, the next round of selling broke the peg for good and the now-floating NuBits price quickly plummeted below 50 cents.

Did NuBits fail because it was well-thought out but poorly implemented? Or was it unsound from the get-go? I lean towards the latter.

To understand how the NuBits peg works, we need to back up and investigate the architecture of NuBits a bit more. NuBits are digital tokens issued by the overlying "Nu network," which is sort of like a digital corporation with no physical hub and no real legal structure. In addition to NuBits, the Nu network also issues shares which give shareholders the right to earn profits from the network and participate in governance. These shares are called NuShares, and as I'll show later on these NuShares are a key part of maintaining the NuBits peg.

Protecting the peg when there is a large increase in demand for NuBits is easy; just create more NuBits. Protecting it when there is an excess supply of NuBits is trickier. In this respect, NuBits is no different from any currency issuer maintaining a peg, say like the People's Bank of China which must have enough resources to keep the yuan anchored in place when a run into dollars begins.

A small quantity of Bitcoin reserves is just one of the Nu network's bulwarks. Once used up, the peg's next line of defence is referred to as parking. Think of this as paying an interest rate on term deposits. When NuBits sellers are pressuring the peg, the Nu network relieves that pressure by offering, say, 10% interest to anyone who is willing to park, or freeze, their NuBits for a period of time. Instead of selling on the open market, the idea is that people will decide to stay put.

Unfortunately, I think that parking may have actually contributed to a failure of the peg. The problem is that the Nu network's parking rates are paid with new NuBits rather than existing NuBits and thus they increase the supply. If a peg is being threatened by an excess supply of money, it makes little sense to relieve said pressure by creating even more NuBits. Instead, the Nu network could have avoided parking-related supply bloat by paying interest using existing NuBits. And it should have earned these NuBits out of the regular course of its operations, say by charging a small user fee, or cutting down on costs. Whatever the case, it means that the Nu network needs to be profitable.

In addition to parking, the Nu network has another mechanism for maintaining the peg. It can issue new NuShares (by engaging in a stock split) and use these to buy back NuBits. The actual route taken seems to have been more lengthy, specifically selling new NuShares for bitcoin, and then using bitcoin to buy back enough NuBits so the $1 peg holds.

But using NuShares to 'back' the NuBits side of the network seems to me like a ticking time bomb. Consider the following dynamic: the moment the peg is challenged by NuBits sellers, speculators will push down the price of NuShares in anticipation of potentially dilutive issue of new NuShares. Which means that the Nu network will have to issue even more NuShares to protect the peg, which only makes the peg that much more costly to maintain and therefore more vulnerable. This leads to more speculation against the peg, and yet another round of NuShares panic as speculators unload in anticipation of dilution. It's a vicious circle that ends in a price of zero for NuShares, and for NuBits too.

There's also an inherent conflict of interest. If maintaining the peg means that shareholders must bear a collapse in the price of their NuShares, at some point is it worth it for them to maintain the peg? Perhaps it's better to pull a Richard Nixon and close the conversion window. Scan the NuBits discussion boards and there seems to have been a bit of this self-serving thinking during the breaking of the peg episode.

Better to use a stable asset like gold or dollars in a vault to back a NuBits-style scheme than volatile share tokens. Of course, this isn't as elegant a solution, since it requires a lifeline to a real world safety deposit box. Nevertheless, it would have helped keep the peg in place.

In sum, NuShares is a good attempt, but it didn't quite succeed. That it lasted so long is a testament to the value of the community that emerged around it. As a tight-knit group it was probably able to self-support the peg at least for a while, insulated from its own inherent short comings by its own momentum. That being said, the Nu network is still around, and according to some rumours, trying to re-peg themselves at $1. I wish them luck!

Related links:

A Report on Peg Abandonment and How to Proceed From Here [link]
The Search for a Stable Cryptocurrency [link]
NuBits - The price stable currency... until it’s NOT!! [link] [update]
My Interpretation of Jordan Lee’s Liquidity Engine Model & Why its First Attempt at Pegging Failed [link]
Theoretical Fedcoin, Meet Operational NuBits [link]
Hayek-Style Cybercurrency [link]
Some Thoughts on Cryptocurrencies and the Block Chain [link]
Robert Sams: Bitcoin, Volatility and the Search for a Stable Cryptocurrency [link]

Wednesday, August 10, 2016

Central banks deposits for you and me

The Bank of England recently announced that it will end a 300-year tradition of allowing employees to keep chequing accounts at the Bank. You can see an example of a cheque above, which is marked with the sort code 10-00-00. Traditionally, folks like you and me have only been able to get a piece of the central bank's balance sheet by holding banknotes. Central bank deposit accounts, which are far more convenient, have been limited to banks and other financial institutions. But the BoE provides a rare example of regular people, specifically employees, being permitted to directly own fully transferable central bank deposits, at least until recently.

The BoE's termination of this seemingly archaic practice is especially interesting in the context of growing efforts to crack open central bank balance sheets to those who have traditionally been hived off from them. A concrete step in this direction is the Federal Reserve's overnight reverse repurchase facility, which allows money market mutual funds to hold overnight balances at the Fed. More ambitious (but less concrete) is the Bank of England's Ben Broadbent, who describes the idea of a central bank adding more counterparties-
"perhaps a wide range of non-bank financial companies, say.  It might mean something more dramatic:  in the limiting case, everyone – including individuals – would be able to hold such balances."   
Echoing Broadbent, BoE Deputy Governor Minouche Shafik has spoken of the need to rethink "about to whom we give access to the advantages of central bank money with its unique qualities of finality of settlement." The idea here is to allow non-banks involved in fintech direct access to the Bank's real time gross settlement system, as Mark Carney goes on to illustrate here.

Turning to the blogosphere, John Cochrane has recently written about having all money backed by the government in order to end bank runs. And in the same vein, here is David Andolfatto's idea of allowing TreasuryDirect balances to be tradeable, thus providing individuals and firms with a safe place to keep cash other than the banking or shadow banking system.

This democratization of central banking sounds like a novel idea. Nosing deeper into the Bank of England's history, however, we learn that it was not just employees who could hold Bank of England deposits; most people could. Some of them were quite famous. An interesting anecdote from the 1700s has Sarah, Duchess of Marlborough, asking her bankers at the Bank of England to provide her with a freebie, namely some pens because she 'could get none that were good.' [1]

Moving forward a century, we learn that upon opening for business in 1855 the Bank of England's Western branch tended to attract small businessmen and private individuals "of modest means" as clients, including the accounts of the University of London, a Regent Street hatter, and a Sackville Street clothier—all on its first day of operations. Later that year the household of Queen Victoria left its private bank in order to do business with the Western branch. By all accounts, the Bank of England seems to have had excellent service:
"Staff were expected to recognize customers on sight and have a good, clear hand for writing up passbooks. Two porters were always on duty at the main entrance, clad in the pink livery of the Bank of England with silk top hats, and even the cashiers wore top hats when serving at the counter." [2]
The Bank of England kept up a retail presence well into the 1900s. But as public service came to be regarded as the Bank's main role, the aggressiveness of its commercial and retail businesses was reduced and it transitioned into a purely bankers' bank. The Western branch would be sold in 1930 to the Royal Bank of Scotland. [3]

By 1963 the Bank of England's services to the public were limited to a small number of accounts for existing customers. Presumably as they died, these accounts were closed. The most recent example of the Bank of England allowing non-banks to set up accounts comes from 2003, when Bank officials decided to provide Huntingdon Live Sciences, a drug company facing threats from animal rights activists, with a BoE account because commercial banks refused to offer their services.

So while it may seem that the Bank of England's Broadbent and Shafik are introducing a modern approach to central banking, the practice of allowing private individuals and non-financial businesses to directly hold central bank balance sheet space (in a non-cash form) is actually an old one. What lessons can we take from this historical example?

Many people believe that an open central bank balance sheet has the potential to render our traditional banking system extinct. Banks are special. They provide the world's most popular exhange media—deposits—as an offshoot of their primary business, lending. But as Robert Sams points out, if individuals are allowed to own safe central bank deposits directly there may no  longer be a reason for them to hold risky bank accounts. Demand for bank deposits falling to zero, banks will have to fund their loans to the public with regular bonds or equity, even as payments are re-routed through the books of the central bank. Fractional reserve banking as we know it is dead.

Depending on who you ask, the replacement of fractional reserve banking with so-called narrow banking, or 100% reserve banking, can be either good or bad. I'll leave that discussion for another day. Whatever the case, Bank of England history illustrates that private bank deposits can coexist with an open central bank balance sheet. Given the choice between keeping accounts with the safe Western branch or a risky private bank, individuals did not collectively flock to the former.

No doubt this was partly due to the two things, the Bank of England's policy of charging for servicing unprofitable accounts and of not paying interest on deposits. Its competitors, on the other hand, did pay interest. In essence, private banks had to retain customers by offering better services.

Which brings us back to the Bank of England's recent decision to close employee accounts. Given the Bank's stated intention of opening up its balance sheet, wouldn't it have been an opportune time to open up its retail banking business to all of England rather than shutting it down? The Bank maintains that it had been having trouble competing with services like online banking being offered by private banks. But as history shows, since a central bank offers something private banks can't —safety—it needn't be as competitive in the services it provides. Alternatively, it could be that the Bank will be following a different strategy of opening itself up, say through a distributed ledger or something like PositiveMoney's Digital Cash Accounts. Whatever the case, don't be fooled by the technological terminology; if the Bank were to open itself up, this would be more of a returning to the fold than a bold new future.

[1] Bank of England: first report, session 1969-70 (link)
[2] Western Branch of The Royal Bank of Scotland - The Story of a Bank and its Building (pdf)
[3] Branches of the Bank of England, 1963 (pdf)