Dutch Central bank’s third DNBcoin prototype delves into smart contracts and trusted ownership

The Netherlands has been one of the leading bitcoin economies in the world since the city of Arnhem became known as the first ‘bitcoin city’ over two years ago. The country’s central bank, De Nederlandsche Bank (DNB), was among the first of the monetary authorities to investigate bitcoin and blockchains, and the bank’s head of research, Jakob de Haan, spoke at a blockchain conference called Reinvent Money in September last year.

In March, the central bank mentioned in their annual report that they are specifically studying their own prototype cryptocurrency, dubbed ‘DNBCoin,’ which received worldwide press. The report addressed how blockchain technology could improve their core business model, but it gave no other details about the project, outside of putting it in a section of their report marked "aims for 2016."

It appears that the details about DNBCoin have finally arrived. According to Ron Berndesen, the head of Market Infrastructures Policy Department at DNB, who gave a speech at the Dutch Blockchain Conference last Monday, explained there was always a plan for three tests of blockchain technology at DNB, each to align with one of the three primary tasks of the central bank.

“DNB is interested because blockchain may have implications for the overarching goal of financial stability and her three primary tasks: 1) promote the smooth functioning of the payment system, 2) prudential supervision and oversight and 3) monetary policy.”

• Ron Berndesen, Head of Market Infrastructures Policy Department at DNB

The first test, loosely purposed to see if a cryptocurrency of some kind could help with becoming a smoother payment system, was completed in-house, long ago. “DNBcoin prototype 1 was basically replicating the early days of bitcoin (Jan – Feb 2009) using five laptops in a connected network and our home-made genesis block,” Berndesen explained in his speech, before adding that “the blockchain is a promising technology for financial market infrastructures.”

The second test, completed more recently, ran in a very different environment, although it still centered around a cryptocurrency. “The second DNBcoin prototype takes the other extreme of bitcoin by jumping to the year 2140,” he offered, “the year when the last fraction of the 21 million bitcoins will be issued.”

This test seemed to show the bank something more promising. “Blockchain technology might potentially add a new dimension to the interaction between the supervisor and the supervised institutions,” he said of the results in his speech. To offer an example, he added that “if they share a blockchain the supervisor can automatically see all verified transactions.”

Berndesen made clear that the coins used in these tests, both of which they referred to as “DNBCoin,” were for testing purposes. “The DNBcoin is only developed for internal test purposes, it will not be put into circulation.” For their third test in the series, however, there won’t even be a coin at all. It seems that the bank is giving up on cryptocurrencies altogether, according to the speaker. “Increasingly central banks view blockchain as a potentially promising technology for financial market infrastructures, but much less so for virtual currencies,” he said, when introducing the plan for DNB’s third phase of testing.

“These two prototypes were focused on the blockchain as a vehicle for a virtual currency. But virtual currencies are not the most promising application of the blockchain. So the third DNBcoin prototype will not be about a virtual currency.”

• Ron Berndesen, Head of Market Infrastructures Policy Department at DNB

The third test, he said in his speech, “is necessary to understand the potential implications of blockchain for the money supply and potential seigniorage consequences of virtual currencies.” However, the lack of testing a cryptocurrency for the third test speaks volumes about the direction that DNB is likely heading in the future.

While still attempting to test blockchain technology for monetary policy, running such a test without using a cryptocurrency narrows it down to a very specific type of platform, one that records transactional data but has no unit of valuable currency to mine like all cryptocurrencies do.

The solution that they are investigating is being described as a ‘permissioned blockchain’, just like one of the many commercially available, private, blockchain data services. Blythe Masters’ Digital Asset Holdings,Factom, and Chain are all such services that are able to begin installation today. There are even enterprise-grade solutions available, including one by professional service firm Deloitte.

Private blockchain services like these must rely on a centralized user database and the age-old user permissions scheme to keep value inside it secure. Hackers and identity thieves have a far easier time cracking a password than they could attack a public blockchain like bitcoin’s, due to its’ open and permissionless nature. At best, according to Andreas Antonopoulos in his speech entitled “Bubble boy and the sewer rat,” such networks simply become bloated intranets, constantly in need of repair and patches for security’s sake.

If Antonopoulos is correct, then for a large corporation, it should already be considered risky to consider holding anything of value in such a blockchain environment. For a central bank, however, it is hard to see how even considering such risk can be justified at all.

At least DNB does not appear to be in a hurry. They have not given any indication of an official start of the testing for this new system, and “It will take a number of years before the full potential will become clear,” Berndesen said in his speech. He also mentioned that for his third test, he is focused on creating an environment with the four following properties:

1. The network is about ownership of a certain digital asset.
2. There is some degree of trust among participants in the network.
3. Resilience of the network is crucial.
4. Adding intelligence to initiate or trigger transactions is important.

While properties three and four are commendable, number one is vague, and number two is the source of the problem Antonopoulos described in his video. Trusting others when trust is not necessary is precisely what opens the door to hackers, typically in the form of identity theft.

Other central banks have been talking about blockchains for one purpose or another already, but nothing as specific as what DNB is planning. The one exception is Canada’s recently-leaked CAD-coin project, however, their plan is for a cryptocurrency that is tied to their fiat Canadian dollars, or perhaps intended to serve as the online extension of them. Neither option would be likely to come with such problems in security as DNB could face.

The closest thing to a central bank deploying a permissioned blockchain network would be the international payment network SWIFT, which has been the target for a constant flow of hacks lately, ever since a successful cyber-heist of US$81 million in February this year. Swift’s CEO, Gottfried Leibbrandt stated that “we’ll be looking for a number of quick wins to improve things in the near term,” during an interview with Bloomberg. However, “the full rollout, and the full shore up, will be a matter of years,” he added.

This is likely why SWIFT is also looking at a permissioned, private blockchain system as a solution to their problems. If they were to implement one, they could be trading one security problem for another.