Regarding reserves and fractional reserve


Various members of our community and other communities have at times spoken about reserves and fractional reserve in relation to our network. Some seem to think that 100% reserves are desirable and equate to solvency, while others (including myself) wish to avoid the use of reserves altogether to avoid counterparty risk. Some view reserves as the backing for NuBits, while others see NuShares as the only possible backing for NuBits:

Benjamin contends that reserves cannot be reliably used to back NuBits. I agree. This view implies that solvency and fractional reserve cannot be assessed by the size of any reserve. If there is no reserve, the notion of “fractional reserve” doesn’t even make sense in the context of our network. However, a related concept is the ratio of assets to liabilities as Benjamin pointed out, where NuShares are network assets and currencies issued are liabilities. If the NuShare market cap is 8 million NBT and there are 1 million NBT in circulation, the asset to liability ratio is 8, not counting any reserves that exist. This is a highly solvent state that doesn’t resemble anything like fractional reserve.

It is clear the original design of Nu as articulated in my whitepaper is a network that avoids the use of reserves entirely to eliminate counterparty risk. When a LPC provides their own funding for liquidity operations, they are implementing this zero reserve model. Shortly before our September release it became clear that NuBot would need to support primarily NBT/crypto pairs instead of just NBT/USD. This was difficult to implement quickly, which means it had some bugs initially and this approach introduces some extra risks which were not well understood and could not be accurately measured at the time. As a result, KTm and Jamie began using shareholder funds to provide liquidity, which may be regarded as reserves in some contexts. It is has been my intention all along that this would be a temporary situation as we would transition back to the original design of LPCs providing their own funds consistent with the zero reserve and zero counterparty risk model. Indeed all custodians besides Jamie and KTm are providing their funds and the transition is well underway.

NuShares are being traded actively at Cryptopia
Possible solution to custodian loss of value due to BTC exposure
Someone in bitcointalk says NuNet is insolvent
[Draft] motion to increase the percentage of reserves
[Passed] Temporarily cap NBT supply with full reserve
History of the Nu Network (Feedback Please)

This is what @Benjamin sumarizes in the abstract of his paper:

In a dual cryptocurrency, on the other hand, shareholder may choose to repay depositors, but are not bound by any legal obligations to dso. I show that, under the assimption that shareholders behave rationally, this implies pegged cryptocurrencies cannot be backed by tangible assets

On page two @Benjamin explains,

In the absence of third-party legal enforcement, solvency is not a very meaningful condition. Since shareholders are not bound be law to repurchase nubits, shareholders could allow the nubits peg to break even when they have the financial capacity to support it. In my view, we should only expect nubits repurchase to occur when repurchase align with shareholders’ economic interests. … I we allow shareholders to vote on the use of reserve funds as shareholder and assume that they vote acording to self-interest, the argument that a 100% reserve offers supoort for the peg falls apart.

To put it straight: he thinks that when the law is not watching, even they had 100% reserve, the shareholders would just pocket the reserve fund themselves even when they could save the peg. They won’t have the will to save the peg when things get tough (even not tough)

This would be a terrible assessment of Nu shareholders. I for one am interested in Nu not only because of neat technologies (POS, voting, distributed computing, and all that) but also because I see potential for massive adoption of nubits by normal businesses. Knowing that the enterprise is an experiment that could fail I still went ahead to support Nu. However most businesses and individuals Nu cators for operate under normal legal framework.These businesses, and the government, would be very concerned, to put it mildly, if shareholders agree to be painted like that.

Practically, as I said in comments on the paper, if we insist that econimic interest is the only thing shareholders care, then things will likely play out such that 1) Nu adoption will be hindered. Nay sayers, doubters, and competitors will have a lot of fuel against Nu. 2) The government will come down on Nu to add that missing “third-party legal enforcement” when Nu gets big enough. 3) If Nu grows in size and has a lot of business depending on it, and it fails in a way incompatible with moral and legal norms, people WILL hold shareholders responsible and get even one way or another. Economic interest does not equal to self-interest.

In short I think Nu shareholders should stay away @Benjamin’s thinking of how shareholders should behave, and by extention his extreme view on reserve, for shareholders’ own good if not for good PR.


Some shareholders have ideological reasons for supporting Nu; it’s true that economic interest is not the only motivator. However, I contend that economic interest is sufficient to satisfy rational investors. I also believe that many investors would not be as comfortable relying upon ideological motivations, particularly since many nu shareholders are anonymous to varying degrees. As long as shareholders believe that future expected returns are greater than the cost to maintain promises made by the network, they will tend to maintain those promises. I think there are good reasons to believe that future expected returns are much higher than the cost to keep Nu promises. With proper management, expected returns could remain greater than the costs of promises for a very long time. Ideological motivations could mean that promises will be maintained in cases where costs are believed to be higher than future expected returns, but this is difficult to predict and ultimately unreliable. If Nu can convince people that they are economically motivated to keep their promises, those people will tend to trust Nu.

That being said, I agree that it is beneficial for shareholders to maintain a positive image.


Fully agreed. Any system that relies on altruism as a substitute for economic incentive will fail. We need to keep working towards solutions that make NuShares a more reliable backing for NuBits, without making them a simple repackaged exposure as in the case of BitShares. It’s a very tricky problem but I have confidence in Jordan Lee’s vision and the collective problem-solving ability we are gradually building in our community.

Or worded differently, once NuShareholders believe that aggregate demand has definitively reached its all-time peak, they have no incentive to support the price of NuBits at demand levels below that peak, unless other Nu products exist that would suffer from reputation damage should one currency fail. I could see voluntary burning being used quite often in periods of modest demand declines (similar to how Parking Rates will be effective in that scenario), but not used at all in a complete collapse. I’m not even sure how this would look in a future where NBT have many billions in circulation; it would be very difficult to tell when a permanent demand decline has set in if there hasn’t been a major protocol failure or regulatory ban.

In a way, permanent demand declines become much less likely the larger our network grows due to the size of the collective belief in the value of NuBits. Perhaps the signal for a permanent demand decline will be when the network starts to see the emergence of “speculators willing to take great risk for great reward, having long signaled that it was in difficulty”, as Jordan states on the website.


Forgive me if I’m retreading on old topics here (Still working my way through the forum history), but what is the incentive for people to become an LPC? To make this work, the risk needs to be spread between thousands of LPC’s. Having one large LPC is dangerous, because the system is dependent on that person. They could be hacked or their liquidity funds could be seized by the government. Distributing that risk between thousands of LPC’s makes more sense. The question remains though, how do we incentivize thousands of people around the world to provide liquidity with their own money? What are they getting out of it? Are the shareholders supposed to pay every single LPC for their services? If so, how? Also, how do we manage to continuously vote for and judge thousands of LPC proposals? Again, sorry if this has already been discussed elsewhere. I should hopefully be caught up on Nu info by the end of the week.


Great question Sentinelrv!

I think there are a few good reasons to believe that the cost of providing liquidity could decrease substantially in the future. Currently Nu needs to offer rather large returns to LPCs because it still needs to build trust. In the future, if wildly successful, Nu would be well trusted by many and therefore a significantly lower return could be attractive (lower perceived risk -> lower reward necessary to be attractive).

Additionally, a mature NBT market that is extremely large would have significant privately provided liquidity. Exchanges would compete with one another for NBT volume, and would offer lower fees and various other incentives. With extremely small fees, private NBT investors looking to profit from NBT trade might be sufficient to satisfy much of the demand with very small profit margins due to the large volume. Thus, in a successful future, a significantly smaller percentage of NBT liquidity may be required of LPCs.

There have also been some discussions about alternate methods to provide liquidity in the future. I agree that the current LPC system is imperfect, and that it will need to evolve in order to be optimal.


This is a fascinating scenario to consider! I think that if more investors are convinced that Nu based products are valuable, the ratio of NSR market cap to NBT Market Cap will increase. This is largely because those investors would realize that future demand for Nu based products could be very high, and they would want to capture some of that growth. Current NBT demand is extremely small compared to what it could reasonably become in the future. As NBT demand grows relative to future expected NBT demand, this NSR : NBT market cap ratio may tend to decrease as investors would believe that less growth is possible. This might not happen for quite some time as it would require Nu to be well known and understood by the global market.

Anyhow, if the NSR : NBT ratio decreases as NBT demand approaches what shareholders believe to be peak demand, NSR market cap may ultimately align itself with expected dividends. Similar to how a new public corporation with a great idea tends to have an extreme P/E ratio that decreases as the corporation grows relative to investor expectations. The corporation then attracts a different sort of investor that is more interested in reliable returns with less risk. I think Nu could follow a similar pattern. In an extreme success scenario, Nu assets may be expected to approach global GDP growth rates. In this case, they might not be perceived to peak until humanity is perceived to peak.

If the P/E ratio of NSR becomes lower as it matures, NuShareholders would have more reason to support Nu products after the price of NSR decreases to match the lower expected dividends. Expected volatility of NSR would be much lower, and Nu products would be perceived to be backed by a more stable and reliable organization.


@GreatLogic, if you were replying my post, my post was 10% about ideological reason and 90% about practical reasons.

I believe the anonymity that shareholders, developers, and custodians have is relative. When there is enough motivation anonymity will be broken. It is prudent to keep some anonymity to fend off casual harrasment but to assume you are not anonumous against government and resourceful private investigators.

Realizing this, one has to care about reputation and legitimacy of things one is seen significantly involved in with cryptocurrencies. Besides economical interest, personal interest also concerns family, friends, professional career, hobby and fun, reputation, sleeping peacefully at night.


@GreatLogic Instead of relying on individuals providing their own funds, why not relying on a big fund owned solely by NuNet and managed by a bot (software) in a trustless decentralized fashion on an exchange owned by NuNet? That way, we would not be exposed to the risks involved with individuals. I admit. I have no idea how such a thing could be implemented.


So what matters in Nu is not to back NBT or collateralize NBT as in a bank, but to provide deep liquidity on the buy side.
As long as deep liquidity is ensured, the peg is maintained.
This clarification from Jordan Lee has changed my view of Nu.
Before I saw it more like a decentralized central bank (Nu has certainly some aspects of it) which issues tokens backed by collateral (reserve or shares); Now I see it more like a decentralized IT company whose value proposition is to offer a pegged crypto by ensuring that there is always a deep liquidity on the buy side.
The question is now how Nu will be able to expand drastically that liquidity.
Currently we have only a few custodians.
But in order to increase drastically the liquidity, I imagine that the first step would be to increase drastically the number of custodians, which would imply a lot of voting and reporting in particular.
In that scenario, I see Nu being in fact a manager of custodians, constantly looking for new custodians and offering incentives for that.


If you have M nubits in circulation, having enough liquidity for all of them on the buy side is having 100% reserve. I think that doesn’t necessarily mean Nu always have M nubits “in cash”, however.


But do you need necessarily to be able to buy back every NBT in circulation at any time ?
I think this is an important question that should be asked to @JordanLee .
I feel we only need to show that we could do it potentially not strictly.
Since we want to be 0% reserve based, that means we would rely on NSR to buy back NBTs.
But as pointed out by @chronos here, NSR market cap is not an irrefutable proof that Nu can buy back every single NBT in circulation. NSR market cap is just a confidence indicator.

That is why I think it all comes down to the quality of the service of Nu, which consists in providing deep liquidity.
If the quality is high, then the market has a high confidence in the peg, then no need for 100% reserve, imho.


We rely on each of the six tiers of liquidity described in this passed motion:

Note that the liquidity available in tier 5 and tier 6 cannot be measured accurately. Tier 5 (interest rates) is quite vast (perhaps greater than tier 6), but its use has negative implications for the future of the peg. We have never come close to needing to use these last two tiers of liquidity.


My interpretation is that maintaining the peg (1) is a different problem that buying back all NuBits (2).
(1) is done via providing liquidity.
(2) is done via converting the necessary quantity assets into “cash” (if the assets are not “cash”) via exchange rate to match the NuBits in circulation
In my understanding, (2) would be rather done using selling NuShares.

Would it be possible to have a real-time tool that displays how much liquidity we have for each tier (except 5 and 6) at any time? I think it would increase even more the transparency of Nu that has been so far of a high level, and help us acquire users even more rapidly.
Also it would be great to know how many times tiers 2-, 3- and 4- have been used so far and how much has been transferred in case it actually happened.


Yes, this is possible and already planned for implementation. Just as the Nu client displays tier 1 liquidity now, we will add liquidity info about the other tiers. Enhancing data feed features, currently planned protocol changes, currency burning (if it passes) and parametric order books for NuBot are all higher priorities, however.