Hi there, I have spent many years working as a business analyst in financial technology centered around trading systems functionality including accounting. I am also a libertarian believe cryptocurrencies are the future. As a result, I joined the Nubits community this week to offer some thoughts. They are just thoughts and so feedback (even if extremely negative) is very welcome. Stimulating a debate to find a solution is one of the things I find rewarding.
I would say that the unique selling proposition of Nubits as “The World’s First Stable Digital Currency” certainly has/had potential for success. It offers those who want the privacy advantages of owning a cryptocurrency without the volatility risk. As it is not possible to have derivative contracts (or forwards) in cryptocurrencies at the moment then this would have been one of the few options available for eliminating such volatility.
The technology used to achieve the peg from a functional perspective seemed well thought through and somewhat ingenious.
My thoughts on the business model
I think ConfusedObserver nailed the retrospective my opinion. I will try an elaborate with my thoughts on why Nubits is in its current predicament:
a) Offering a product without a sufficient revenue model
- Nubits offered what is effectively a hedging mechanism (mainly vs BTC) seemingly for free.
b) Reserves held BTC
- The reserve asset (BTC) is not the asset being pegged leaving reserves exposed to market risk and/or a BTC hack.
c) BTC is a volatile asset and therefore the Nubits peg was costly to maintain
- The amount paid to incentive the liquidity providers needed to be high to cover the risks associated with them having to hold a high volatility asset (BTC) in addition to being forced contrarian traders i.e. They would effectively have to sell BTC for Nubits at a time when everyone else is going the other way.
- Meaning (as Sentinelrv says) Nu didn’t have the required amount of transaction fees by users of the network in order to cancel out the NuBits that needed to be printed in order to incentivize liquidity providers.
d) Unsustainable model
- With no income to cover these expenses the incentive being paid in Nubits increased liabilities to an unsustainable point.
- Should the liquidity providers only have been happy to receive BTC as a reward then the result would be depleted reserves.
e) No real cost accounting
- Easy trap to fall into but deadly, especially without a good revenue model.
f) Nubits are more popular when the market is crypto risk adverse but it’s also a crypto
- When the market is crypto currency enthusiastic more users will sell Nubits for a volatile crypto looking for price gains. When sentiment is negative (due to say a hack or tech failure) crypto users may be reluctant to hold any crypto at all, including Nubits.
With the benefit of hindsight, the project should have been pursued once there was clear evidence that Nubit holders were prepared to pay the costs of providing the liquidity to maintain the peg. Further by not holding reserves or allowing custodians to hold their liquidity in the instrument being pegged (USD) the model became exposed to the volatility risk that the customer base was trying to avoid.
I understand that using USD for this would have increased costs but at least the costs would have been quantifiable in eliminating the volatility risk to BTC. If the investors/liquidity providers were not happy with the lack of anonymity in using USD and this caused the adoption of BTC then I understand. However, in my opinion this change in the risk/reward equation would have made the project less viable. The reason being that the fees charged by the custodian would need to be much higher and thereby increasing Nubit liabilities/draining reserves much sooner and as mentioned BTC volatility leaves the reserves exposed.
Zero Reserve Model
My own feeling is that the zero reserve model without charging fees and using BTC as reserves (instead of USD) is unfortunately akin to a Ponzi scheme i.e. works well only while there is demand for Nubits but would fail otherwise. The reason for this is that the more active the custodian the more liabilities/less reserves there are and as a result constant increase in Nubit demand by customers would be required to stave off an attack on the wall.
As an aside you could never allow Nubit lending as an unethical player could be to borrow Nubits and then sell them to a custodian for BTC (or USD) and keep doing this until there are no reserves to back the peg at which point the peg breaks and the Nubit loan is repaid with the BTC (or USD) for a large profit.
Full Reserve Model
If you have a full reserve in that 1 USD is given and held for each Nubit you end up with the risk of a US regulator calling an amount “illegally gained” and just withdrawing it from Nubits cash account. Effectively then this means Nubits is left with the risk of ill-gotten gains while the criminal has eliminated that risk by holding a cryptocurrency.
Tether is using this model but instead of issuing a cryptocurrency, it issues a token tethered to the amount deposited. As Bitreview points out though:
Tether’s central reserve model reintroduces an inherent weaknesses that bitcoin meant to replace. And this is not simply an ideological exercise. Tether’s reserve account represents a single point of failure which, whether by accident, incompetence, evil design or force majeure, could undermine their entire currency.
Also if you have a full USD reserve, you don’t need custodians for the liquidity but then you will be acting as a central bank/issuer and fall under various regulatory authorities. It seems that this problem is absolved if custodians are used for liquidity but then they would need to be and payed for the privilege when you don’t in fact need them.
Lastly (but not finally I am sure) there would need to be an interest rate parity. Should USD interest rates go negative then Nu would lose money on reserves held and this would need to be passed on as a charge to the Nubit holder.
Assuming a continuation of the current model but with fees
I think what most people want to know when evaluating a service up front is the answer to the following 3 questions. 1) What are you offering? 2) How much will it cost? and 3) Is it value for money given the risks. If the answer 1) is great! and the answer to 3) is yes! then they will be happy to pay and you have a business. In fact, I think the two best revenue models for Nu are like what happens in the retail FX market now as they are globally accepted already i.e. either a) a transaction charge, b) a spread or c) a combination of both. Personally I like the idea of spread with no transaction charge as it’s an easier sell. People feel as if they have avoided a charge. It’s just psychological but it works. Everybody expects a spread in the FX market and I believe that if there isn’t one it’s a reason for suspicion. I totally understand Nubits need a firm price at the walls that generates the require revenue but if the reserve currency is USD this spread could be a lot less than using BTC, maybe .98 or .99 vs 1.01 or 1.02. Maybe the peg needs to be re-branded as a “band” or something similar.
In fact, if the Custodian model using BTC is to be retained, then the spread for doing so should probably be more like .965 to 1.035 given what I have read about the true cost of maintain liquidity in BTC previously.
If the custodian is buying at 0.98 and selling at 1.02 against USD or 0.965 vs 1.035 against BTC with x% of this revenue going to Nu to pay dividends and Custodian liquidity running costs, then great. The important questions then are what spread is needed to cover Nu expenses to make the model sustainable and will this spread be acceptable as in the answer to question 3) above. If yes Nubits has a future, if no then maybe not.
Proof of burn
Coingame’s proof of burn concept has some merit although running with the idea and using it in different way may be of use. I know I mentioned that I am a fan of a spread over a fee but if the community view is insistent on a 1 to 1 parity (USD to Nubits) then the following is an example of what may work:
Say the Nubits pricing mechanism against USD shows the following bid/ask rates to the Nubit/USD 0.98/1.02
A customer buys 100 Nubits in exchange for USD 100. Before the transaction can be completed he/she is sent via the API a message “You will incur a transaction charge of y” where y = peg rate (in this case 1) minus the Nubit USD bid rate x number of Nubits or in this case 1 - 0.98 x 100 = 2 Nubits.
Upon acceptance 2 Nubits are sent to a specified Nu address and burnt. The end result is that the customer now has 98 Nubits. Of course for a sale it would be the Nubit USD offer rate – 1 x 100 = 2 USD.
The other alternative based on a spread method would be to grant the customer Nubits at the peg i.e. 1 but behind the scenes send the 2 Nubits to a specified address and burn automatically while adding only 98 Nubits to the customer’s address. All the customer sees is 98 Nubits in his/her account which is what would be expected if they chose to deal at 0.98. The reason for this is that by simply adding 98 to their account accounting event for the fee will not be produced.
A third method would simply be to use a % fee on the amount being transaction.
In all three cases though could be given to using a tiered spread/fee structure so that high volume transactions are charged less in fees in terms of % of transaction amount. On the other hand, small amounts could attract a minimum Nubit charge.
In the above spread example, the deposit transaction and the 2 Nubit burn would need to generate the correct accounting entries (as mentioned Nu really needs accounting!). Assuming that a Custodian is not used the accounting entries in both cased could be:
- DR Nu Assets (for the USD deposited) 100
- CR Nu Liabilities (to reflect the Nubits created) 100
- DR Nu Liabilities (to reflect the burn) 2
- CR Nu Income (to reflect the income generated) 2
- I know that you had some issues with the speed of activating the capital tiers under sale stress. At first I did not see this as the biggest of issues but thinking about it some more this is a major weakness in that someone with prior knowledge of this (and with a large enough holding) could dump their Nubits at such a speed to force a collapse before the capital tiers could be invoked.
- Maybe there is a need to change the name from Nubits to NUUSD etc to cover other fiat possibilities.
- Nubits could be useful for trading existing assets issued in Fiats should a side chain or other method of recording asset ownership be developed i.e. a blockchain asset exchange.