If Nu were to be done again? A better system?

The Plaza
#1

Hi everyone.

I was thinking about Nu earlier and I was thinking about how the model could be improved. I can’t say my thoughts have any merit but I’ll provide them anyway.

I’ve read that Nu is supposed to be a zero reserve system. I think this is a wrong way to look at things. In recent times I think people have come to think of Nubits as being backed by NuShares, so Nushares acts as a Nubits reserve. With this in mind, perhaps the design of Nu has missed the point all along. The way it has been designed made it difficult to protect the peg with NuShares. If it were to be done again perhaps the focus should be on the relationship between NuBits and NuShares.

Also it appears to me that park rates have had little success. All they have done is provide a source of Nubits inflation that can exacerbate problems further down the line. The marketing efforts to promote park rates appeared too ineffective.

Let me suggest a new system

For the purpose of the post I’ll give the name UltraShares (US) and UltraCoins (UC).

An exchange mechanism can be created between UC and US, so that UC is backed up by US. Shareholders would provide or delegate agents to provide the network with the current price of US. UC can be exchanged for US at the US price and vice versa. When UC are exchanged for US, the UC is destroyed and vice versa. This would ensure that 1 dollar worth (or whatever the currency prices are denominated in) of US can be exchanged with exactly 1 UC, less any fee, thus maintaining a peg. Multiple coins could be created denominated in different currencies.

This is maintainable as long as the value of US is maintained. To simplify this, the US/UC reserve ratio (in terms of market cap) can be monitored. If the value of US goes down then the reserve ratio goes down, and if it goes up, the reserve ratio goes up. To protect against market volatility, the reserve ratio should be high to start with, and may be adjusted if market conditions allow for it but should always be high enough to protect against sudden loss of reserves.

If multiple coins denominated in different currencies are issued, then the total value of all those coins should be used to determine the reserve ratio.

Maintaining Reserve Ratios

The task of maintaining the peg can be simplified to maintaining the reserve ratio. A threshold could be put in place for when the ratio is considered high enough (for instance a 4x ratio). When the ratio becomes high enough it can be reduced by issuing new UC, the proceeds of which can be used for share buybacks. New UC should only be issued when reserves are high enough. This could be enforced at the protocol level by preventing issuance when the reserves are below the agreed threshold.

When the ratio becomes too low things can be trickier. I do not suggest using park rates in attempt to promote demand, as this may not bring about enough demand to overcome the guaranteed inflation that it imposes, and such demand may be too temporary for a long term solution. There is no profit opportunity to shareholders with park rates either. Park rates should remain out of the protocol all together.

By making the reserve ratio high enough it should be possible to weather the reduced reserve ratio long enough for the situation to reverse. In the long term fees will act as a deflationary force, taking UCs out of circulation. As this is not a short term solution, reserve ratios should have a high enough tolerance. Raising fees too high would reduce demand for UC and may cause long-term reputation issues, so fees should be balanced.

By slowly reducing UC in circulation with fees, this provides profits to the shareholders by creating eventual opportunity to issue new UC, by increasing the reserve ratio. Therefore the business model would involve taking as many UC out of circulation via fees as possible. This would also help protect against falling reserve ratios. Once again, this does not mean high fees should be used as that will hurt demand. But it would likely be wise to implement proportional fees. A minimum fee could be calculated according to the size of transactions, but otherwise a percentage fee is applied. I understand this was being planned for Nu 3.0, though it hasn’t been included yet? The expression could be:

fee = max(feePerKb * bytes / 1000, amount * proportionalFee)

Where amount could be calculated as:

amount = inputAmount - outputAmountSentToInputAddresses

So this way a sender can prove change by simply sending change back to an input address, and change can avoid being charged a fee. This is already established with avatar mode. Anyone that wants higher privacy by making it harder to determine change outputs would need to pay a higher fee.

A percentage fee would provide profits that are more proportional to transaction volumes. UC would need to be marketed as a tool of exchange and not as a store of value. Another reason to avoid park rates.

Profit to shareholders in the system would be obtainable through higher demand and through fees. Sustainable profits would come through fees whereas demand is unpredictable and cannot be guaranteed to always rise like in a ponzi scheme.

In Summary

  • Remove park rates
  • Provide exchange mechanism between shares and pegged coins
  • Keep a high reserve ratio
  • Focus on using fees to reduce coins in circulation
  • Balance fees to ensure demand remains strong for the coins
  • Promote trade in coins
  • Discourage savings and storage of coins
  • Only issue new coins when reserve is high enough

Are there any other views on this? The reserve ratio would need to be well above 100% in my opinion, and ideally several multiples higher until a suitable reserve level could be determined. NuShares did have a higher market cap than NuBits and could have been used to support the peg more effectively, but the system does not include a direct exchange mechanism which would have helped. Also NuShares did once have a market cap over twice NuBits. Maybe however the ratio should be higher than that.

4 Likes
#2

We already have made most of the changes you suggest. I gather you like the direction things are headed then?

#3

I’ve not paid attention to all the changes completed and that are in development. It’s hard to track everything that goes on if I’m honest. Following everything in the forums is not easy. Perhaps another suggestion is to communicate all the changes more clearly in a blog or in a particular forum thread in a clear and concise manner for those who find things hard to follow.

Do you mind pointing out which of my suggestions are in the pipeline? I know the proportional fees are, though they aren’t in Nu 3.0.

It does not appear possible to regain a high reserve ratio at this point unless some miracle occurs. That is the most critical part of my whole suggestion. Nu’s reputation is shot regardless of anything else. It’s not possible to issue new currencies such as Chinese Nubits without making the ratio even worse. Without a strong NuShares market cap there is no hope.

It’s not really meant to be a suggestion for Nu, because I don’t think it is obtainable. It would make more sense to wipe the board and start from scratch. I think most people would not be interested in starting a new system, but I thought I’d give my view regardless.

#4

I understand the peg failed due to a willing abandonment and inability to secure the peg with NuShare sales. This ultimately comes down to the reliance on human delegates to provide liquidity and maintain the peg. By providing an exchange mechanism within the protocol itself it ensures the peg is maintained without needing delegation to human operators. As long as the exchange rate is set correctly, it would work as intended, and third party exchanges would not be needed to support the peg. Exchanges would only be needed to determine the market price of the shares, and to provide the ability for people to trade into other assets.

I do not believe that park rates helped or would have helped and are completely unnecessary where it is possible to provide support from NuShares.

#5

I too think the NuShare backing mechanism may function (much better) if implemented in such a way that it’s automatically reactive to the market. There’s some discussion about this in the following thread.

2 Likes
#6

I did it over a year ago without any protocol changes and open sourced the automation program. We even had a suggestion from ben for how to do it even better with a hard fork using only 1 extra bit per block. That all fell hard on deaf ears because it couldnt be manipulated for insider trading. I dont know why i still comment here…

#7

Many of the suggestions in that thread are quite over-complex however. It is far simpler to implement an exchange rate mechanism that fixes the exchange rate according to the NuShares price. If people could exchange NuBits for NuShares and vice versa according to a exchange rate within the protocol using an exchange transaction on the blockchain, then that is all which is needed and would work OK if sufficient NuShares reserves could be maintained. However sufficient NuShares reserves do not exist, which is one reason why I say it would be best to start over.

Maybe enough people think it is a good idea that they would be willing to create a new network with this mechanism.

It’s absolutely important that is is done at the protocol level to remove any human-trust issues, and so that it works smoothly and consistently, and that it costs nothing because it’s all builtin to the system.

#8

Finding a static exchange rate is a fool’s game. You either have to do it as seeded auctions (Some Suggestions To Eliminate Peg Maintenance Expenditures) or as bts did it, using decentralized counterparties and debt calls, or something no one has mentioned yet. We’ve been discussing this for 2 years, but have never been able to get lead devs or majority shareholders involved in anything that wasnt highly manipulable by insiders (i.e. lead devs and majority shareholders).

#9

I’m not suggesting either of those things. I’ve already mentioned it in my OP, so please read through it fully.

I’m suggesting that there is a new type of transaction that allows for exchange between the hypothetical UltraCoins and UltraShares at the recorded exchange rate. So if UltraShares were $500 a piece and UltraCoins were pegged to $1, you could exchange 1,000 UltraCoins plus fees for 2 UltraShares. The UltraCoins are destroyed and the UltraShares are created. Likewise you could exchange 2 UltraShares plus fees for 1,000 UltraCoins, destroying the UltraShares and creating the UltraCoins.

#10

Regaining a high reserve ratio is very much within our grasp. Today we have tier 1 through 4 reserves (this excludes park rates and NSR sales) of $17,000 on an outstanding currency supply of $136,000. To reach our target of 48% reserves only requires acquiring about $48,000 additional funds. These can be acquired either by currency or NSR sales. We are selling $480 of NSR every day to restore that reserve. All it takes to restore a high reserve ratio is for buyers to continue showing up for small daily NSR sales for a few months. A high reserve ratio can also be restored with currency sales of less than $100,000. So long as NSR remains liquid, restoring a high reserve ratio seems almost certain with a number of months.

Nu has worked openly, publicly and transparently for three years to solve the hard problem of providing a decentralized and stable blockchain currency. We have accomplished some amazing things. The liquidity model works to support the peg, even under extreme duress. We had a problem with ignorant people we allowed to handle our short term funds not employing the model and violating agreements they had made with shareholders. So we fixed those. Now funds are reliably available for liquidity operations and they follow shareholder directives. We even have plans for more improvement in the way funds are handled.

We keep building our unprecedented network and improving it as quickly as possible. How else could something like this be built? It is something we can be proud of. We shouldn’t be shirking our brand. Nu is the world’s greatest laboratory for stable currency currently selling for a mere $450,000 (that’s the market cap). That’s an incredible value, especially when you consider there is a chance NuShare holders will benefit from share buybacks powered by very large quantities of currency sales. It can certainly happen. There isn’t really a competitor in our space out competing us.

Why would it makes more sense to start from scratch? That is what Augeas said, and that was completely unsuccessful. The energy for innovation in stable currency is all right here. The infrastructure is all right here. The expertise and experience are all right here. You can’t build a system like a stable currency by tearing it down every time an adjustment is needed and building it from scratch. Businesses never do that. Rather, such systems iteratively develop. When Amazon improves, it doesn’t shut down its operations to start new. It just makes adjustments. The call to start over is simply non-nonsensical. It seem to be motivated by a desire to have a bigger role in the network, as people who have been frustrated by their own lack of influence within Nu sometimes think they will have more influence in a new system that starts over, like Augeas. That didn’t work out. The problems have to be solved in Nu.

#11

With my system the value of the shares needs to be far in excess of 100% the value of the pegged currencies. Only after a stable system is in place for some time can suitable tolerance levels be determined (by looking at volatility, worst case scenarios and deflation) for reducing reserves. Maybe it’s a bit of a misnomer using the term “reserves” in my system, as the existing shares are not used to back-up the pegged currencies, but rather the market value is. As long as the market for the shares is big enough to support the pegged currencies, then it would work.

#12

Consistent with a shareholder directive to not create any additional NuBits until the peg was restored, the peg was restored from July to September without the use of park rates. NSR sales have clearly emerged as the much more effective tool, although it may not be the cheapest. While I don’t advocate the removal of park rates, it is a fact they haven’t been very important to the network recently, so Nu is already mostly aligned with what you are suggesting here.

This is basically what we do in liquidity operations. We use quite a bit of automation (such as NuBot). We strive to add automation, of course. This has to be done iteratively and it has to be funded.

It is important to point out we restored and initially maintained the peg in 2016 with absolutely zero tier 1 to 4 reserves. All we used was NSR. However, a high reserve does add extra peg security (so long as the reserve isn’t lost, which is a concern). A high reserve also permits liquidity operations a much higher margin for error. We are biulding a 48% tier 1 to tier 4 reserve, which when combined with park rates and NSR sales, is quite high.

The fee for sending NuBits was increased in 2016 five fold, from $0.01 to $0.05. Shareholders have passed a motion to implement a more complex and profitable fee system that can be based on the quantity of funds being sent, as opposed to the size in bytes of the transaction message. All that is needed to implement is the funding to develop it.

I think the rise in fee from 0.01 to 0.05 does this.

Indeed, @jooize and I are having a variety of conversations with exchanges on how to promote trading.

There was a prohibition on new NuBits at the time the peg was abandoned. Only issuing new NuBits when the reserve is met doesn’t make sense, because the best way to acquire the reserve is to sell NuBits.

#13

The difficulty is in shareholders maintaining control of the reserve. We have had considerable problems with that, mixed with some success. Truly, the only way to give NuShare holders perfect and complete control over a reserve is to place it directly in the NuShare market cap with a NuShare buyback.

#14

I’m suggesting that the mechanism is built directly into the protocol and that the exchange takes place on the blockchain. This removes the need for liquidity providers completely.

Parks rates discourages trading and encourages savings. Another reason to bin park rates.

That eats into buy-side liquidity. Did you mean NuShares? There’s no need to sell NuShares with a protocol level exchange system. Shares are created as necessary when people want to trade-in their pegged currency for shares.

With an exchange between the pegged currency and the shares, outside reserves are not needed. The value of the shares themselves act as reserves for the currency. The Shareholders do not need to control this, because it’s enforced at the protocol level. The only thing needed is a record of exchange rates so the exchanges are done at the correct market price, perhaps to be provided by special transactions made by delegates to update the price information.

Maybe the word “exchange” is confusing people. Perhaps saying “conversion” is better. Replace all uses of “exchange” with “conversion” or “convert”.

#15

You either use oracles for this (like bts) or you find it with a market mechanism (like auctions). Im not sure what your secret option number 3 is here.

#16

So this sounds exactly like bts, except using a slightly different PoS (peercoin’s instead of nxt’s), restricting coin creation and thereby eliminating the DEX that bts has, and having a much more naive share-coin conversion that can be easily manipulated.

Maybe i should explain why bts uses the debt-call system it uses. Let’s say we use your method, and the delegates say the price is 1 coin per share. Then someone sells 100 shares. Now the delegates lower the price. That same person then buys those shares back cheaper. Rinse and repeat. The debt call system of bts ensures this cant happen by having a distributed counterparty and debt calls.

#17

Although I agree with @MatthewLM and @Phoenix about many opinions, I found your weakness:

Yes, shares MUST backup coins, but no need to actually trade shares and coins! And the fee is a bad idea to control the coins demands because many coins holders deposited their coins at exchanges and don’t use blockchain network at all! They just trade coins and BTC on Poloniex --off chain!!!

The Nobel winner -F A Hayek’s solution is much simpler and more elegant than any of your mechanisms. Hayek’s way for liquidity providing is decentralized, robust, and effective. SHORT-TERM LENDING BUSINESS.

This is Hayek version of Nu

  1. We issue Hayek share (HYS) and get it listed on an exchange(B&C is better), after a while, the share price becomes relative stable, e.g. ranging from $2.5 to $4/share, usually this price influenced by ICO(such as BKS’s $4,but BKS price destroyed due to Nu crisis, if not, it will be $4-6 today.)

  2. Shareholders continuously vote a pledge ratio according to share price and money supply policy. The backing ratio should be controlled around 200%-400%. Only human beings can do such work because only human beings can perceive the share price in real world and free market movement.E.g. one day the ratio is voted as 1:1, which means potential LP can park 1 share to borrow 1 Hayek coin(HYK) from protocol. This is short-term lending (one month , fortnight, or 2 months). Shareholders may not bother to change their votes for a long time as long as the share price remain relative stable.

  3. LP then build HYK pairs on various exchange with original peg price $1, their incentive is to earn spread trading fee, which is paid by customers. In this way, Official community don’t pay anything to LP, The spread(1%? 0.5%? 3%?) is totally determined by free market, in the end, perhaps 1% at FIAT pair, 3% at BTC pair. 4% at LTC pair, this is also dynamic, both LP and customers can accept it. Spread depends on the BTC/LTC volatile, exchange risk level, etc.

  4. Shareholders also vote an inflation rate(positive or negative, I guess positive in most times), the official pegging price is continuously changing. Eg. Each month/day/block, to shift the pegging price.

5)One year later, the pegging price increases from original $1.0 to $1.06, but our share price still remaining $2.5 to $4, We don’t need to worry about the 1.06$ because it is hided behind the white noise of share price fluctuation and turnover rate. Even after 5 years later, when HYK becomes $1.3 and share price still $2.5–$4 level and we have to shrink our HYK supply a little, we can still maintain the peg. Because if we “close the valve completely”—shareholders vote a 1:0 ratio, i e stop lending business, then all the lending contract will end within one month(HYK returning), otherwise the LP’s pledged share will be confiscated by protocol(grace period suggested here). I guess in most situation, we can “close the valve a little”, no need to completely shut off it.
The turnover rate is critical, even in 5th year we have 90% HYK supply(in circulation) of the 1st year, if the turnover rate is 10% more, the trade volume can be same.
(I think you know what’s turnover rate, we trade one dollar 1000 times, the trade volume is $1000, turn over rate is 1000.)

  1. In 6th year, we find a good revenue(Youtube accept HYK as payment, customers can pay youtube video service on basis of each mouse click!) Now the price of HYK is $1.3(selling price), and $1.25(buying price), but who care the spread!? The HYK protects customer from inflation at all! The inflation itself is a relative fuzzy concept, perhaps there are 1000 inflation rate in 1000 men’s brain. This is fuzzy math, our brains works in this way. 1000 men believe Britney is beautiful, although the real scores in their heart slightly vary. Most our customers will admit HYK is anti-inflation, just like most of them admit Britney is a beauty.
    Finally, we have more confidence to buyback shares and maintain ever raising pegging price, and pay the software development, marketing etc.
#18

What’s to prevent the delegates/oracles/whatever from reducing the price, and then buying up lots of BTS via the forced settlement mechanism? BitShares can still be manipulated by manipulating prices. The way around this is to use many oracles and take the median price and perhaps punish oracles that provide prices too far outside the correct price by requiring them to put down a deposit.

@Sabreiib What is LP?

1 Like
#19

Your system can be abused with 100% legitimately acting oracles. The forced settlement system cannot.

So back to the example. I sell shares in your system on the blockchain exchange. Either the oracles decrease the price or they dont. If they do, then i buy bacl cheaper. If they don’t, then that means i can get as many shares at that price as i want without moving the price so as long as i have more volume than the non-blockchain markets I can use this fact to manipulate the system.

The bts system makes sure this kind of manipulation cant happen by making the DEX an open market for settlements. In order for you to make a huge trade on the DEX you have to find a counterparty willing to take on the risk.

#20

Liquidity provider.

E.g. I park my shares ro borrow coins from the protocol, and build coin/usd pair on poloniex. I earn money from spread trade ie buy at 0.99, sell at $1.00. I provide liquidity to coins, and my revenue is from the customers. Anyone with shares can do such a thing, so this is completely decentralized liquidity providing.

If on free market the coins demands is low. I’ll find the trade volume of mine decline, then I may borrow less coins from protocol, if the coins demand is high, I’ll borrow more until park all of my shares and borrow as many coins as possible. And my pledged shares value is always higher than my borrowed coins, so when the lending contract will end, I have to buy back all the coins and return to protocol, otherwise, I lose my shares.

In this way, we can easily shrink the coins supply in a short term period, because each lending contract is short term, this is the beauty of Hayek’s mechanism.