Decentralized Liquidity Without Counterparty Risk Not Yet Implemented
On January 24th, 2015 Jordan Lee issued a stern warning to shareholders about their responsibility in working to decentralize liquidity operations and the consequences that would result if they didn't. At this point in time, there were several LPCs besides jmiller and KTm who were providing their own liquidity funds for a fee paid by shareholders. The problem was that shareholders were being too picky in what they wanted to pay LPCs. Jordan saw this pickiness as dangerous to the health of network, especially when potential LPCs able to provide their own liquidity funds were in such limited supply. Jordan wanted to develop the LPC market, but shareholders often refused to approve LPCs because of the high fees they were asking for. Here is a quote from the original whitepaper:
"First someone who wishes to fulfill this role must seek shareholder approval via the custodial grant mechanism. Say a particular liquidity provider or LP custodian has 10 million USD he wishes to use to provide NuBit liquidity. He would expect compensation for lost opportunity cost (he could otherwise invest those funds in rental property, stocks or bonds) and for the risk of loss via an exchange default, such as we have seen with Mt. Gox and others. While the market will continually reprice this, let's say in our case the prospective LP custodian decides a 5% return every six months is fair compensation for lost opportunity cost and risk of exchange default. So, he promises shareholders to provide 10 million USD/NBT worth of liquidity for one year in exchange for 500,000 NBT. Shareholders approve this using the grant mechanism and he is granted 500,000 NBT. Now he must provide 10 million in liquidity constantly over the next year. Alternatively, once shareholders have sufficient credibility, LPCs may be compensated after providing liquidity services."
Now here is a direct quote of Jordan Lee's warning to shareholders:
"What Jamie Miller and Kiara Tamm are doing differs from what is described above in a number of important ways. Most importantly, they are not using their own funds, but shareholder funds. This simple fact introduces tremendous counterparty risk and is a source of centralization. It is a serious deviation from my design with potentially disastrous consequences. With data feeds now in beta, decentralized LPCs presenting zero counterparty risk remains the only major feature described in the whitepaper which has not been implemented. This feature is by no means of peripheral importance. In fact, I believe Nu will not succeed unless it is implemented. Time that we spend waiting to implement it is time that we put the network at considerable risk of failure, including catastrophic and total failure.
The software needed to implement this is already distributed. What is missing are LPCs presenting proposals and shareholders willing to approve their proposals. The situation is comparable to a healthy person who reasons they will not purchase medical insurance to save money (in our metaphorical scenario there is no government provided health care). After all, they have better things to spend their money on that improves their quality of life now. Being healthy, not having medical insurance is not a problem, until suddenly it is a very big problem because they develop cancer. Similarly, it is clear to me that shareholders are choosing an option that is cheaper in most circumstances, but is vastly more expensive in certain failure modes.
Consider what would happen if KTm lost control of the approximately one million NuBits in her possession and they were brought to the open market suddenly for sale. At a minimum, very high interests rates would need to be offered to stimulate enough demand for NuBits. At worst, the young system would be unable to stimulate sufficient demand in the wake of such a scandal. With a plummeting market cap, currency burning may not be effective enough and the peg may be permanently lost. However, the liquidity KTm provides is very cheap compared to the proper alternative.
Now consider what occurs if muchogusto or someone with a similar proposal providing liquidity lost control of the funds they were using. In order to have NuBits in the first place, muchogusto must buy them, so it would only represent the reversal of that initial purchase, not a brand new liability like with KTm. So even if those lost NuBits were immediately sold (and there would be much less incentive to so than with KTm), the sudden sale of 25,000 stolen NuBits carries no risk to the system. The loss is borne by muchogusto, not the network or shareholders.
The network had begun a transition to offering decentralized, counterparty risk free liquidity but in recent days it has become clear that the network is actually turning back toward increased reliance on KTm and Jamie. KTm and Jamie understand that their role is supposed to be transitional, and they are hesitating to expand their operations beyond Bter and CCEDK to provide shareholders incentive to do the right thing. So in fact what is occurring is that the network is failing to expand liquidity provision to newer exchanges while increasing reliance on centralized liquidity that brings risk of systemic failure. It is important this development not escape the attention of shareholders, because shareholders are certainly in a position to change the course of events.
So, as a shareholder, if you find yourself saying that a specific decentralized and counterparty risk free liquidity proposal such as the one proposed by muchogusto is too expensive, it certainly is more expensive than the liquidity being offered by KTm, just as not buying medical insurance is cheaper than buying it. My advice would be that if you think it is too expensive, present a competing proposal that is less expensive. If you are unable or unwilling to do so, you cannot demonstrate that it is too expensive.
It was my expectation that liquidity provision would be comparable in some ways to Bitcoin mining. It is a relatively simple way (all that you need to do is run some software and deposit some funds on an exchange) that people involved in the project can use modest skills to support the network while being rewarded financially for their contribution.
When Nu first began liquidity providers received no transaction fee discounts on exchanges. We felt lucky just to get support on exchanges at all. Now, the profitability of having NuBits on your exchange (with liquidity support) is undeniable, and we have begun to move toward zero transaction fees for liquidity providers. But when the market was immature, we had to pay a higher price for liquidity. Shareholders face a similar dynamic with decentralized counterparty risk free liquidity provision. That market is extremely immature, and the only path to lower prices is the creation of a market for those services. This market will provide innovation for hedging, ways to reduce exchange default risk, etc, but if there is no market then those innovations will not occur. Therefore, it is of paramount importance that shareholders pay whatever price is necessary to establish a market in decentralized liquidity provision. If they do so I'm certain innovation and competition will bring pricing down. If they don't, the Nu network will continue to be exposed to centralization, counterparty risk and serious lack of scalability in its liquidity operations."
Unfortunately, shareholders didn't have much time to act on Jordan's warning. In the month of February, his fears were realized as the Nu Network suffered from several back to back hackings of its most used exchanges. [1, 2, 3, 4] Together, these exchanges hosted 99% of open market NuBit trading. Liquidity was wiped out in a matter of weeks and it was discovered that a certain amount of shareholder funds had been stolen across all the exchanges. Thankfully, Jordan had the foresight to implement his 6 tiers of liquidity. This greatly limited the financial damage to Nu, since most liquidity was stored in external wallets in tier 3. Regardless, damage was done to Nu's liquidity operation and it needed to be dealt with quickly.
NuShare Buybacks & Ending Centralized Liquidity Operations
In response to multiple exchange hackings, on February 14th, 2015 Jordan Lee proposed a motion to cease the centralized liquidity operations of KTm and jmiller and allow NuShare buybacks:
EDIT: I have been notified by Jordan that it is incorrect that this motion was in response to the hackings, because he told shareholders in this post before the hackings even took place that he was going to advance it. So he had already planned to introduce this motion regardless of the hackings that took place later. This is just a quick edit to point out my error. I'll be revising this section soon.
"When our network began, we had a way to increase the NuBit supply at the will of shareholders. We then decided to add a way to decrease the NuBit supply at the expense of increasing the NuShare supply (due in the 2.0 release). It should not be surprising then, that a mechanism for reducing the NuShare supply would also be introduced, to complete our flexible supply of both shares and currency.
Specifically, I am proposing that in most cases, instead of distributing dividends, proceeds from the sale of NuBits be used to purchase NuShares in the open market, which are then burned. This is commonly done with equities and is known as a share buyback. Due to low liquidity in the NuShare market, we can expect such activity will have a marked positive effect on the NuShare price. This high NuShare price will enhance our ability to sell NuShares later when we are in the opposite part of our economic cycle and need to support the NuBit price. Therefore, we will have two complimentary mechanisms which are the exact inverse of one another:
1. When NuBit demand is low, NuShares will be created and sold while NuBits will be purchased with the proceeds and burned. NuShare supply increases as NuBit supply decreases. This depresses the NuShare price as it supports the NuBit price to the pegged level.
2. When NuBit demand is high, NuBits will be created and sold while NuShares will be purchased with the proceeds and burned. NuBit supply increases as NuShare supply decreases. This inflates the NuShare price as it suppresses the NuBit price to the pegged level.
Fortunately, no protocol or even client changes are needed to do this beyond what is already approved by shareholders and scheduled for our 2.0 release. Whether dividends are distributed in addition to share buybacks being done will be up to shareholders to decide on an ongoing basis. Shareholders can also decide what proportion of NuBit sale proceeds go to dividends and what proportion go to share buyback by how they pass custodial grants.
In order to move to the model described above, which is decentralized and eliminates counterparty risk to NuShare and NuBit holders, the LPC operations of KTm and Jamie must be wound down. Their service was of great value at the launch of NuBits, but it is time for Nu to mature into the decentralized, counterparty risk free system it was originally envisioned as. Similarly, it is important that the undistributed NuShares I possess (currently nearly 300 million) either be distributed or burned. Once these actions are taken we can say that Nu is free of any single points of failure or catastrophe: decentralized, free of counterparty risk, robust and resilient.
Ending the LPC service of KTm and Jamie is also the key to unleashing share buybacks and dividends. Right now they are holding the proceeds of their NuBit sales as reserves. Transitioning to our zero reserve model is synonymous with ending the LPC operations of KTm and Jamie. When this transition is complete, proceeds of NuBit sales will not be held at all. Rather, NuShare holders will be immediately rewarded when those proceeds are used for NuShare buyback and burn or dividend distribution. In order to unlock the full benefits of share buyback and dividends, shareholders need to make a full commitment to paying a transparent and upfront cost for liquidity from LPCs providing their own funds. As I've said before, this feels kind of like paying health insurance premiums, but allows you to accurately predict the costs that will be incurred. While KTm and Jamie continue operations, we only know that our cost for their service will be between their modest 2% fee and 2 million NBT, if shareholders lost control of the funds somehow. Nu cannot afford a 2 million NBT cost at this point in its development. We can and ought to eliminate the possibility of this potential catastrophe quickly by purchasing insurance, so to speak."
The contents of Jordan's motion which passed shortly after mandated that KTm and jmiller cease their liquidity operations within 90 days and burn all remaining NuBits in their possession. Within 60 days after the protocol is altered to permit custodial grants of NuShares, Jordan must burn all undistributed NuShares and within 30 days after the passage of this motion, Jordan must burn all his NuBits held on behalf of shareholders in excess of 250,000 NuBits. He continued:
"Good preparations will need to be made to see the network in good health after this motion is fully executed. New LPCs will need to be funded and cultivated. The implementation of a liquidity pool is one particularly inclusive way to do this. Being an LPC requires ownership of significant liquid funds that can be put at high risk, it requires the skills and time to run NuBot, as well as a good understanding of trading and the nature of Nu liquidity operations. A liquidity pool would separate these requirements: A pool operator needs the skill and time to run NuBot and the understanding of liquidity operations, but they do not need to have any funds. Individual users of the pool only need a tiny sum of liquidity they can put at high risk, because users' funds are pooled together to create a significant quantity in aggregate. Individual pool users don't need any special knowledge and it won't require any of their time on an ongoing basis. The potential profits for pool operators and individual pool users are huge at the present time. Based on recently passed LPC proposals, we can see that 10% return per month can be had. We shouldn't wait for pools to emerge and fail to elect standard LPCs. As the Nu network matures, I expect we will see a mixture of LPC pools and LPCs providing their own funding. I would rather see the network supported by 100,000 NBT of liquidity by LPCs providing their own funds than 400,000 NBT of liquidity by LPCs using shareholder funds like KTm and Jamie.
Once the transition is complete, we will have two main types of LPCs, as mentioned in the whitepaper: single side (called sell side in the whitepaper, but this is being modified to include granted NSR that are used to apply buy side pressure on NuBits) and dual side LPCs. Single side LPCs come in two sub-types: (1) those that have been granted NuBits to apply sell side pressure and NSR buyback (and burn) or dividends, and (2) those that have been granted NuShares which they can sell to apply buy side pressure and NuBit burn. Dual side LPCs are the more common type that liquidity pools will be which provide both buy and sell side liquidity in order to create a buffer zone of liquidity for the peg. The purpose of single side LPCs, on the other hand, is to balance the buffer zone created by the dual side LPCs.
Proper preparations for the burning of undistributed NuShares will be the granting of a small percentage of outstanding NuShares (perhaps two or three percent) to a multisig NuShare address. This will allow the signers to agree to sell the NSR quickly to purchase NBT to support the peg when buy side liquidity is low. The idea is to have no more NSR than could be possibly needed for sale in a two week period. If additional NuShares are needed for sale beyond a two week period, they can be created in a new NSR custodial grant. Rather than creating a large fund for future development as originally planned, development will be funded with NuBits on an ongoing basis. If the emission of these NuBits creates too much selling pressure, the selling pressure will be counteracted by a NSR grant and a NBT burn. Our development savings are held in our market cap.
Interest rates may still be used as a peg support mechanism, however it is possible that NSR grants combined with NBT burning will become the more important mechanism.
If this plan is implemented NuShares will be even more volatile than they are now. It makes sense that they would be, because we are basically diverting the volatility the market naturally wants to impose on NuBits to NuShares. NuShares will be a speculator's delight. They should never approach zero so long as the network purpose of a pegged currency that is used continues to be served."
NuShare Auction for Peg Support
On February 20th, 2015 Jordan Lee submitted a proposal to hold an auction for 100 million NuShares that remained undistributed and in his possession. The purpose of this was to use the proceeds of the auction to help support the peg, which had become strained after multiple exchange failures. Jordan explains below:
"I designed Nu to maintain stable currency pricing as currency passed through cycles of increased and reduced demand. Cyclical changes in the level of demand are inevitable and the system can gracefully handle those changes while maintaining the peg. We can see that we are in the midst of a time of reduced demand as indicated by the recent decline in buy side liquidity. This means it is time to use interest rates as well as issuing new NSR where the proceeds will be used to buy NuBits or increase buy side liquidity. Shareholders won't have the ability to create NSR by themselves using the protocol until the next protocol change is introduced. However, I have the ability to simulate this with the undistributed NSR I control.
It is true that much of the reason why this is needed is the losses that have occurred as a result of poor shareholder speculations. Shareholders have lost value in the BTC held by myself, KTm and Jamie. They have also taken large losses with the shareholder funds lost at BTER and Excoin. It only makes sense that shareholders would experience a dilution of their shares as we take steps to regain this lost value. Some people have the perspective that shareholders have been unlucky by having experienced these losses. While it is possible there is some validity to that perspective, it is not helpful. A much more helpful and empowering perspective is that shareholders made poor choices that increased risk far beyond what it needed to be. We are essentially a subset of Bitcoiners, who can be characterized as aggressive risk takers. So shareholders have taken risks on the value of BTC, on the reliability of exchanges and on the reliability of myself, KTm and Jamie. The risk taken on myself, KTm and Jamie have worked out so far. The speculation on BTC price and exchange default risk went quite badly, on the other hand. The notion of taking large unnecessary risks with network assets is very much at odds with trying to maintain a sure thing like a pegged currency. It won't create trust in the stable value of NuBits. It is a serious mistake. It is contrary to the design I first revealed last April and have reiterated loudly ever since.
Here is a quote from my opening paragraph in Decentralized liquidity without counterparty risk not yet implemented, globally pinned for weeks, so that it has been the most prominent message on our forum. I believe it needs even more attention than it has received:
What Jamie Miller and Kiara Tamm are doing differs from what is described above in a number of important ways. Most importantly, they are not using their own funds, but shareholder funds. This simple fact introduces tremendous counterparty risk and is a source of centralization. It is a serious deviation from my design with potentially disastrous consequences. With data feeds now in beta, decentralized LPCs presenting zero counterparty risk remains the only major feature described in the whitepaper which has not been implemented. This feature is by no means of peripheral importance. In fact, I believe Nu will not succeed unless it is implemented. Time that we spend waiting to implement it is time that we put the network at considerable risk of failure, including catastrophic and total failure.
The urgency and prominence I assigned to this message has proved very appropriate. If we want the Nu network to succeed, we must send a strong signal that shareholders are committed to containing risk by implementing the design in full, and not a centralized, deformed variant laden with counterparty risk as shareholders have decided to do to date. I certainly wouldn't bid on NuShares unless I saw shareholders make this commitment to quit their gambling ways. To demonstrate this commitment, shareholders need to pass this motion and they also need to quickly pass additional (existing) LPC proposals that provide liquidity with zero systemic risk by utilizing private funds. If this is done, our solution will have the credibility needed to have a successful NSR auction.
So, I'm proposing that we hold a blind auction for 100 million NSR in two weeks time. Bids would be accepted via Bitmessage or an email address dedicated to the purpose. Bids will only be accepted from business partners, which means you are or will be active in creating value for the network outside the scope of being a shareholder. The top price, lowest price and total funds raised will be revealed publicly at the conclusion of the auction. NBT, BTC or PPC will be accepted as payment, with a 2% surcharge added to bids when PPC or BTC is used. The exchange rate used for PPC and BTC will be the price at the time bids are due. Multiple bids from the same individual will not be accepted."
Jordan subsequently posted a draft of the auction rules. The draft also laid out the priorities for using the funds raised from the auction:
1. The first priority for funds raised will be to ensure the buy side liquidity is equal to the sell side liquidity as displayed in the Nu client. PPC and BTC will be used to purchase NuBits to accomplish this if necessary.
2. The second priority for funds raised will be to provide tier 4 liquidity. Suitable persons to hold these funds using multi sig addresses will be sought out. Hedging will not be employed due to counterparty risk. Tier 4 liquidity must be highly liquid, and BTC is the only instrument that meets this need at this time, although I hope that liquid NSR markets will permit NSR to be used as tier 4 liquidity eventually. The quantity of NSR for sale is as low as it is to limit the effect of BTC volatility.
3. The third priority for funds raised will be the payment of contractors engaged in development. NuBits are generally used for this purpose, and in that case all that is needed is to ensure there is good buy side liquidity on the open market.
There apparently had been some controversy or confusion over why this auction was necessary and what the funds would be used for, so Jordan took the time to explain it in further detail and had shareholders pass a motion authorizing the auction.
"There has been a modest amount of controversy regarding the planned auction. The passing of this motion will give it the legitimacy it needs and make clear that the majority of shareholders support the auction. I proceeded to announce the auction with the understanding that the time needed to have extensive community discussions and a full voting period was too great for our scenario. Still, there is no reason to not advance a motion such as this after the auction countdown is in motion. I expect it to pass and prevent a minority of shareholders from feeling that something unfair was done.
While the context that makes the auction a desirable course of action is clear to me, I may have made the mistake of not fully presenting that context in connection with my auction proposal. Here is that context:
The original plan was to distribute 1 billion NSR at ever increasing prices. At the time that commitment was made, Nu was designed as a network without systemic or counterparty risk (from the perspective of shareholders and NuBit holders) in its liquidity operations. The use of shareholder funds by LPCs like KTm and Jamie had not even been contemplated, so far as I am aware. The idea to do that first emerged in early September. Prior to that, it was anticipated that LPCs would only operate on NBT/USD pairs, and that is all NuBot was originally built to support. As our release date of September 23rd approached, there was no exchange supporting NBT/USD, but we did have NBT/BTC support worked out. So, we pivoted quickly to providing liquidity on BTC pairs as a matter of necessity. This meant NuBot was hastily modified to support that. Using a new piece of software with major last minute modifications that also exposed LPCs to complex scenarios regarding arbitrage and BTC volatility made the risks of operating NuBot very difficult to assess. So we turned to shareholder funded operations with the expectation that in a matter of weeks the level of risk would become known via the experience of KTm and we could switch entirely to self funded LPCs quickly. That did not occur for variety of reasons, chief among them being shareholder rejection of LPC proposals. I have consistently and earnestly pointed out the importance of self funding LPCs, beginning with the internal release of the whitepaper last April. Two days before release, I was focusing on recruiting self funded LPCs by offering bonuses to start the role. A month ago a summary of the risks became the only globally pinned topic on this forum, giving it the primary attention I understood it needed. Then came our wave of exchange defaults. This was combined with outstanding NuBits declining by about 60% over the last two months to create significant stress on the network. It is possible to overcome this by offering very high interest rates on NuBits. However, consistent with a shareholder motion to burn currency with the proceeds of NSR sold, shareholders have indicated they view the NSR market cap as a more credible way to ensure NBT value -- one that demonstrates willingness to sacrifice NSR value to maintain NBT if necessary. It is a good signal to send to the market as well as a prudent way to manage the peg. That is what this auction does.
So to summarize, I said long ago that NSR would be sold for increasing prices over time and that liquidity would be provided without counterparty risk or systemic risk to holders of NuBits and NuShares. The latter has not yet occurred, against my own will and wishes. Current liquidity operations in the form chosen by shareholders have resulted in substantial defaults and systemic losses. This important change made by shareholders made it impractical to only sell NSR at increasing prices. Combined with shareholders' decision to sell NSR on an as needed basis perpetually (not necessarily at ever increasing prices), this auction is a very reasonable action attuned to the evolving will of shareholders.
Finally, while what I have said may paint a picture of the network as strained, in recent days I have been elated about how recent stresses prove the resilience and robustness of the network. We had about 60% of all NuBits in circulation two months ago exit circulation and sustained exchange defaults that are likely to equal 30% to 50% of all NuBits outstanding at the peak of demand a couple months ago. This is the equivalent of having all NuBits sold back to LPCs and we kept the peg perfectly! It is clear we can repair our financial standing with additional NSR sales. If we also eliminate systemic risk in liquidity operations as shareholders are currently voting to do with 81% consensus (in the last 100 blocks), it becomes difficult to imagine what could break the peg in the future. Our design has been successfully stress tested to the extreme.
I should probably articulate what the alternatives to having this NSR auction are:
1. Have a similar auction with somewhat different terms.
2. Raise parking rates to double digits to stimulate purchase and parking of NuBits.
3. Cut network expenses by terminating compensation agreements with most contractors building our network infrastructure. Currently, such network expenses are around 60,000 NBT per month."