Asymmetrical control in NuBits and the problem it creates

[Digiconomist]

Hi everyone,

I’ve been watching NuBits closely and I do not fully understand how interest rates are supposed to keep demand for NuBits high enough to keep the price at $1. I get the idea, but according to my analysis it cannot be effective.

First of all, let’s take a look on a price above $1. The answer to this is: expand supply and sell at $1. Well, great, so far no problem – this is a hard cap and will work beyond doubt.

But below $1 things get tricky. There is no hard cap, someone buying unlimited NuBits to get the price back up at $1. Instead, there is only a soft floor by offering interest (in NuBits) in order to reduce supply and attract buyers. This should in theory get the price up to $1 and then interest rates will drop. In other words, there is asymmetry around the price of $1 and the solution to falling prices is: expanding the supply through interest. But wait, where did we see that before? Wasn’t the solution to an increasing price already expanding the supply?

So where does this lead? Well, first of all the asymmetry causes downward pressure on the price level. If you sell at $1 you will always have a winning trade. You face unlimited supply if prices move higher, but not unlimited demand if prices move down. You do not have to be a speculator to figure out you should sell at $1, it’s the only rational thing to do. It’s like buying a stock at $0, you cannot lose. A little below $1 this can still be an interesting investment, because demand will fluctuate and cause prices to go below $1.

So okay, there will be periods of less demand and this will just enforce falling prices during this period (it could also trigger the imbalance), it’s not the end of the world – that’s what interest rates are for, right? To get demand and the price back up. Well, that’s where things are starting to go wrong. Again, interest expands the supply, and expanding the supply/sell side to get demand back up? Hmm…

Let’s take a very extreme example to check how this is going to work out just to check the general concept. Take 100% interest on just a single day. If I am not in any way interested in NuBits, that is indeed interesting! Suppose the price of one NuBit is at 90 cents. I’ll buy 1000 at 90 cents for $900 and expect to sell 2000 tomorrow at $1, so I get $1100 for free. Realistic? How about no. First of all, the demand drops again when interest rates drop back to 0% when the price is at $1. Then all that newly created money through interest has to go somewhere, which will be dumped on the market. Because I now have 2000 NuBits, I can put a ridiculously low price of $0.45 per coin in order to break even again. If I sell anywhere above this price, I make a profit. So, the price was at 90 cent, and because of the interest it’s now about to crash. There has been a temporary increase on the demand side, but the most serious effect is on the supply/sell side and the only way to fix it is by adding yet even more supply (via interest).

So this was an extreme example, but this starts happening even when rates are low as it’s the same mechanisms. In the about section https://nubits.com/about/price-stability prices are said to go to zero if the coin has failed, what I am saying is that due to this mechanism the coin is designed to fail. There will be imbalances, at the very least due the fact that there is only a soft floor at and a hard cap at $1. Once you start fixing this by expanding the supply things will just go downhill. This has little to do with speculators, as my examples show it’s just the result rational behavior.

The question: Is this indeed a critical flaw in the design of NuBits, or am I overlooking something here? Please comment below

[desrever]

Hi, great post and great explanation.

I believe you are overlooking the liquidity pool, aka the buy wall https://nubits.com/about/white-paper#buy-side-liquidity . Also, dual-side custodians , or other type of custodians who put proceedings from the sale of nubits back on buy like KTm’s custodian proposal that is in operation now

[Digiconomist]

The problem I see with the liquidity pool is that they still get paid in NuBits, adding to the supply/sell side that the liquidity buy wall is supposed to hold. Also, you can create tons of NuBits, sell them at $1 and have them bought back if the price falls below - the market will even do that for you if short selling is possible. But then you’re still adding a lot of pressure on the sell side. Every single coin that is created on the supply side needs to go somewhere, and what I see is that every options seems to create demand by adding to the supply/sell side including these additions.

I believe it is possible to create a buffer that will hold for a while, but my main concern is that it just makes it something like a time bomb. As supply/sell side pressure grows, you to strengthen your buy walls to deal with this. To enforce the buy side, you must either way add to the supply/sell side - until it eventually blows up… Hmm… I didn’t see anything that actually allows the supply side to let off some steam.

[biophil]

The TL;DR of your post is this:

“In NuBits, inflation is the solution to decreasing value of NuBits.”

This is my biggest reason for skepticism. I was coming here this morning to ask a similar question and initiate a similar discussion; I’m happy that I’m not the only person who’s noticed this problem.

My one question: Are NuBits ever destroyed? If the answer is “no,” then how do you expect “parking” to support the price long-term? Parking reduces the supply in the short-term; but as soon as the bits un-park, there are more of them. Like the OP said, these “more of them” have to go somewhere and since by hypothesis we’re already in an oversupply condition, the network may have to increase interest rates again to incentivize re-parking. And on and on it goes until the buy-side liquidity fund is exhausted.

I’d love for someone to convince me that these are wild-and-crazy doomsday scenarios, or that I’ve missed something important.

[desrever]

king reduces the supply in the short-term; but as soon as the bits un-park, there are more of them. Like the OP said, these “more of them” have to go somewhere and since by hypothesis we’re already in an oversupply condition, the network may have to increase interest rates again to incentivize re-parking. And on and on it goes until the buy-side liquidity fund is exhausted.

TX fees are burned, parking can last arbitrarily long

[biophil]

biophil said:
king reduces the supply in the short-term; but as soon as the bits un-park, there are more of them. Like the OP said, these “more of them” have to go somewhere and since by hypothesis we’re already in an oversupply condition, the network may have to increase interest rates again to incentivize re-parking. And on and on it goes until the buy-side liquidity fund is exhausted.

TX fees are burned, parking can last arbitrarily long

I’ve been skimming this discussion and I believe Jordan Lee acknowledged that burning TX fees would never be enough to meaningfully reduce supply in the kinds of conditions we’re talking about.

So, we’re left with arbitrarily long parking, but no incentive to do so. If I’m one of two people with large amounts of NuBits and some of them need to be parked for the next million years, the best scenario is when both of us park. However, if you park and I don’t, then I’m much better off personally than if we both had parked; so I have a strong incentive to defect, throw you under the bus, and make you think I’m going to park and let you take all the loss of parking. Thus, we have a Prisoners’ Dilemma, which is exactly the kind of thing that crypto is supposed to eliminate programatically. The whole point of crypto is to eliminate the need for me to trust individuals; indefinite parking is not trustless in NuBits.

[Digiconomist]

[quote=“biophil, post:4, topic:275, full:true”]
The TL;DR of your post is this:

“In NuBits, inflation is the solution to decreasing value of NuBits.”[/quote]
Great summary!

TX fees are burned, parking can last arbitrarily long

For TX feees to have any effect you need an extremely high monetary velocity, while interest tends to have a decreasing effect on velocity. Besides, how high can these go? Low transaction costs are considered one of the main benefits of cryptocurrencies, so I don’t think these can be raised high enough. I’m not sure how the length of the parking will change anything. Given a typical yield curve, a longer period just means more interest and a bigger underlying supply.

[CoinGame]

I don’t think the concerns you’re expressing are unfounded, nor are they falling on deaf ears. There’s really no perfect system right now. We’re pegging to a monetary unit that could itself destabilize. Bitcoin’s volatility has scared away people from everyday use because they’re not interested in the risk regardless of the utility.

But lots of the concern with this design seems to be “it won’t last forever”, which is really the concern that could be said to FUD (not that that’s what you’re trying to do, but we’ve seen this with Bitcoin, Peercoin, etc, etc) any sort of innovation.

Phone lines for internet access lasted as a mainstream utility to access the internet for a bit over a decade and a half in major markets. I was there when it started, and I thankfully watched it die out. The utility of it was extended through several innovations. It was still an important step to achieving widespread demand for internet access. Not many people look back at the technology as a failure because it reached a plateau which the fundamentals of the technology could not expand beyond.

I personally expect NuNet can achieve it’s goals for far greater than a decade. I personally believe that we will be able to innovate solutions into the protocol, or through ancillary mechanisms to extend the utility of the network far beyond what a lot of people are projecting.

If it does not last forever, that’s OK. Nothing does, and it doesn’t stop a majority of people from waking up in the morning and trying to make things better.

Everyone on the team is an accomplished problem solver. The distribution of NuShares ensures that a wide variety of people outside the developers are encouraged to build a system that will work for a long time. I have already begun the discussion with people internally, and I’m sure people who want to support the network will begin that discussion themselves.

We are on day one of this experiment. Bitcoin is on day 2450, and they still have lot’s of problems to sort out. So it’s truly appreciated that you have concerns with the long term viability of the project. We would absolutely love to have you join us in the journey to build a solution that meets its goals well beyond 2450 days, since that’s the only current standard to go by.

I just find it rather myopic to see “it wont last forever” posted everywhere as some sort of ground breaking argument.

[Indigoman]

I’ve been thinking about this extensively since release. The infinite supply when coins get parked makes it seem like parking wasn’t a solution to the problem but rather just a way of postponing having to deal with it. Don’t get me wrong, I’m a great enthusiast and optimist when it comes to peercoin and now Nubits. I want this to work and I have the confidence in the community to come up with solutions in the future. It’s as the devs said, we can use this innovation until it’s exhausted by which a newer innovation or improvements to it will have already fixed it or replaced it. Bitcoin just came out a couple of years ago and people are already taking the whole system for granted. It’s an innovation and innovation leads to more innovation.

Having said that, I’ve came up with two ideas of how to counter the expected infinite supply issue and since I’m currently in class should be focusing with the prof :P, I’ll just be brief and ask for your analysis and input on the ideas:

Option 1:
Use a forced transaction fee mechanism that adjusts based on the inflation rate ( maybe a derivative of the interest voted on) so the inflation by coins getting unparked balances out with destroyed coins in the transaction fees. Think of it just like miner fees in Bitcoin. Just like transaction fees go to miners in Bitcoin to incentives them to mine, the transaction fees in NBT would be, in essence, paid to the parkers. (As interest rates get higher to reward parkers, transaction get fees get higher to balance the inflation in a manner where it seems like the public is paying the parkers to park by spending higher transaction fees. which also means that the value of the unparked coins with interest will be higher at the time of unparking since the parker will have more coins while public supply of coin will get throttled by the increased transaction fees. hope I made sense lol if someone understands this please feel free to rephrase it.

Option 2:
Take the reward(interest) for parking outside of nubit to limit inflation. Pay rewards for parking in the form of Nushares. The logic is based on the idea that people would park coins because they see potential of long term investment which means they believe in the coin. The best way to reward them in that case is to give them equity in coin in form of nushares. In this manner, NBT inflation is only controlled by demand for it rather than by parking interest. Since the parker gets nushers, he will have even higher incentives to park again to protect his investment. by protecting the peg, he protects his steady income of dividends as well the value of shares he hold. Basically once a person parks, he’ll have more incentive to park the next time and the more he parks the more invested he becomes in Nubit. With all that vested interest, we shouldn’t have a shortage parkers on low demand cycles.

Would love to hear your thoughts on these two ideas. I’m not an economist. I’m a software engineering student but very interested in the topic and the analysis of the nubit system. Please include concrete analysis and test cases if you see a fault in the ideas presented. (Spoken like a true programming nerd lol)

ps: I typed this on my phone so please excuse bad spelling, grammar, missing words.

[Indigoman]

Another idea I had was to change peg from exactly $1 to $1+/- 5% (buy wall at $.95 sell wall at $1.05)

The idea behind this is provide a breathing room for traditional traders to profit from trading in NBT. Organic buy and sell walls will be created naturally by the market above and been below the walls NuBot will use which will ease the pressure off of our custodians. Is my logic flawed here?

[mquandalle]

Option 2:Take the reward(interest) for parking outside of nubit to limit inflation. Pay rewards for parking in the form of Nushares. The logic is based on the idea that people would park coins because they see potential of long term investment which means they believe in coin. The best way to reward then in that case is to give them equity in coin in form of nushares. In this manner, NBT inflation is only controlled by demand for it rather than by parking interest. Since the parker gets nushers, he will have even higher incentives to park again to protect his investment. by protecting the peg, protects his steady income of dividends as well the value of shares hold. Basically once a person parks, he’ll have more incentive to park the next time and the he parks the more invested becomes in Nubit. With all that vested interest, we shouldn’t have a shortage parkers on low demand cycles.

This second option makes a lot of sense for me. You have described the parker rationale, but it also correct the nushares valuation:

• In case there is an increase of demand for nubits, nushares holders create more nubits units (from thin air) and thus they get richer;
• In case there is a decrease of demand for nubits, nushares holders create more nushares, therefore their relative participation in the nu network capital decrease, and thus they get poorer – which makes sense.

But I think we miss a correct incentive structure for this model to work in the long term. Basicaly we rely on the honestly of nushares holders to vote on a dillution of their participation, which may be perceived against their own short term personnal interest. I’m not sure how this can be addresed, but maybe nubits holders could participate (vote) somehow.

[masterOfDisaster]

Another idea I had was to change peg from exactly $1 to $1+/- 5% (buy wall at $.95 sell wall at $1.05)

I find the idea great. It introduces a spread between buy and sell of the “central bank” represented by the custodians.
But that spread kicks only in once you trade the NBT for something else at an exchange. As long as they are kept in circulation it is no drawback (and they are meant to be kept in circulation!).
The +/- 5 % might be a bit too high, but I’d like to let a proper calculation of a feasible spread for the economists.
I like this proposal, because it can be done without changing anything in the protocol.
And by utilizing the spread for “earning” some money (at the central bank), you have money to pay the interest rates without creating inflation - assuming the spread is chosen wisely.

[RobertLloyd]

If dividends for NuShareholders are funded by earnings other than the creation of NuBits (e.g. transaction fees, spreads, loaning out NuBits, selling of NuBoxes, etc), then each NuBit will be represented by one dollar held by Nu - correct?

In such an arrangement, does this mean there will always be enough dollars held by Nu to create a buying wall to maintain the $1 value?

If so, in this scenario, is parking and the creation of unbacked NuBits, to pay interest, necessary?

[RobertLloyd]

I am assuming that, if we don’t consume incoming dollars created by the sale of NuBits for investment and dividends, we have a perfectly balanced system - increasing supply at will to create a selling wall to keep the price no higher than $1 - and using those dollars to maintain a buying wall to prevent the price dipping below $1. Which maintains a stable NuBits price, which is the object, right?

But if we deplete the dollar side, even by 1%, we compromise the integrity of that equation.

The result is not that NuBits will become a ponzi scheme as some have suggested, since NuBits will still be providing a productive and legitimate service - a stable unit of exchange.

Rather, the problem, as I see it, is public perception. Everything hinges on getting NuBits accepted as safe and sure - but if our equation is not water-tight, it risks being flawed in the minds of people, particularly if opponents keep drawing attention to it. People won’t use something they feel is not secure.

The downside in maintaining an even balance between buying and selling walls is that NuShareholders will not get dividends for their labour and financial investment - but the upside is that NuBits will have greater chance of gaining adoption, in which case the share value will rise (I presume)

Meanwhile, the parking mechanism can be used, not for the defensive reason of maintaining NuBits, but as part of a conventional banking role - borrowing money to loan out to businesses at a higher rate, etc. In time, successful profit lines can be established, thus building a dividend-base in the longer term.

My understanding is very imperfect, but this is how I see things - at the moment

[Indigoman]

I have put together a proposal of a solution addressing this issue. Would love to hear your thoughts on the matter over on that thread.

[Proposal] Solution to the problems of Asymmetrical control in NuBits

The problem was discussed on : 1- Question about selling pressure aka the buy wall 2- Asymmetrical control in NuBits and the problem it creates Proposed solution first mentioned: 1- Asymmetrical control in NuBits and the problem it creates Problem : Nubits can be created out of thin air when demand exceeds supply to maintain the peg to $1. Therefore the total supply of Nubits in the market increases. When the supply of Nubits increases beyond the demand, the parking mechanism simulates a…

[Digiconomist]

I’m never unwilling to think along, but what I would like to know then is what problem NuBits really solves. “The volatility in cryptocurrencies” by pegging to the USD by integrating central bank functionalities? Because I want to point out that Bitcoin was created as a solution to inflation and a flawed central bank. By recreating the USD in virtual currency form, you do not just get those problems back. On top of that, you also get any flaw in the new currency - another central bank mechanism in NuBits’ case. Even if we fix that mechanism, what do we achieve? Because it still seems like a step back even if it works.

[CoinGame]

I’m sure the people that bought Bitcoins at $800 a pop are fine that they’ve lost 50% of their investment within a matter of months. Way better than losing 2-3% per year /s. Bitcoin mining is becoming more centralized every day. It’s a ticking time bomb as well (to use the same language you called NuBits on your website before imploring people to stay away). In fact, you’re replying to me as if Bitcoin has some sort of one up on fiat, but reading this quote from your website seems to show the uncertainty of Bitcoins future as well. Not only the uncertainty of the future but how poor it will be as a currency because of the design.

As the value of Bitcoin must continue to increase, it prevents the digital currency from becoming a stable store of value. A store of value is “the function of an asset that can be saved, retrieved and exchanged at a later time, and be predictably useful when retrieved.” In the case of the dollar, this is ensured by existence of a monetary policy. The money supply can be influenced, so prices can remain stable or even increase to promote spending. With a fixed supply, there is no monetary policy to have a dampening effect on price levels. In other words, price volatility will remain relatively high, leaving one Bitcoin far from “predictably useful when retrieved.” On the other hand, it was also noted that Bitcoin’s deflationary nature discourages spending by design. This seems like a poor perspective for a network that is intended to be funded by transaction costs, once the release of new coins comes to a halt.

How has Bitcoin resolved this since 2009? They haven’t. They’re spiraling towards numerous outcomes with many of them not being good. Those are the problems we’re solving - to answer your question.

I’m glad that there’s criticism of the design. We’re having a lot of discussion on how to proceed forward thanks to people articulating their thoughts on solutions instead of “OMGEE CENTRALZ BANKZ! INFLATOON! SCAM! PONZI! SCAM!”.

The fact is we are not a central bank, Federal Reserve, IMF, World Bank, or Illuminati. If we had more guns and tanks that may be a different story. We don’t. We’re a decentralized crypto asset which has a solid foundation for offering the first easy to use crypto that works well as an actual currency/store of value. We did not beg for funds through a kickstarter to make this happen. We’re not forcing anyone to buy NuShares or use NuBits. We understand that there’s challenges which need to be overcome. Our design has far more flexibility to overcome the challenges ahead of us than Bitcoin or any altcoin.

We’ve had a lot of really smart people join the board to offer their assistance in meeting those challenges. Hopefully that trend continues. As someone who sees the developer discussions I can assure you that everyone is listening and notes are being taken.

[Chronos]

I’ve designed a new mechanism to address the issue of unboundedly-growing NBT supply. Although peg control remains asymmetrical, it provides a way to reduce NBT supply, improving the system’s long-term viability. I’m interested in your thoughts.

New mechanism to reduce NBT supply: NBT Burning

The idea is now better formulated in my mind. Leave the parking mechanism as it is today Add a new function, Burn NBT Shareholders vote on the amount of NuShares generated per burned NBT Default reward is 0 NuShares, so this “hard fork” will initially change nothing In times of crisis, if parking incentives can’t maintain the peg, shareholders can start voting up this reward Reward for burning 1 NBT is the median burn vote of past 2000 blocks Any number of NBT can be burned. Example: if share…

[Ben]

I’m never unwilling to think along, but what I would like to know then is what problem NuBits really solves. “The volatility in cryptocurrencies” by pegging to the USD by integrating central bank functionalities? Because I want to point out that Bitcoin was created as a solution to inflation and a flawed central bank. By recreating the USD in virtual currency form, you do not just get those problems back. On top of that, you also get any flaw in the new currency - another central bank mechanism in NuBits’ case. Even if we fix that mechanism, what do we achieve? Because it still seems like a step back even if it works.

NuBits is just the first implementation on the Nu framework. We understand that using an inflationary currency to peg price to isn’t a long-term solution that will satisfy eveyone. It is our intent to develop other forms of digital currencies that are far more adaptive to inflation and deflation of external units like the USD or EUR.

For day-to-day usage, the inflationary aspect of USD is negligible. A coffee at your local cafe is going to cost approximately the same amount oday as it will tomorrow. This price stability is sufficient to help us get Nu off the ground now, and allow us time to come up with solutions that address the larger concerns.

Only time will be the judge, but where we are today, versus where we anticipate we’ll to be in a couple of years is going to be significantly different. Just as the ecosystems supporting digital currencies are vastly more developed today when compared to what they looked like when Bitcoin was released.

By no means should you interpret this as me discounting your concerns, quite the opposite. I believe it is possible to create a consensus-based system that can keep itself solvent and adapt to the changing needs of the communities that use it.

[masterOfDisaster]

I want to point out that Bitcoin was created as a solution to inflation and a flawed central bank.

First I want to state that Bitcoin is driven by inflation.
The Bitcoin mining industry has been generating 25 new BTC each roughly 9 minutes (due to the increasing hashing power in the last quarters).
This might continue until the next halving of the coinbase reward at block 410,000. Even if the arms race at Bitcoin mining gets less aggressive and the hash rate remains unchanged, there will be 25 new BTC each 10 minutes for the next roughly 1.5 years.
With a total supply of (2014-09-29, 07:08 p.m. UTC) 13327250 and more than 500,000 BTC generated by mining in the next year, that leads to an annual inflation of almost 4%.
Without that inflation the whole mining industry did collapse. The tx fees would not be able to sustain the mining, miners would be turned to other still profitable to mine coins and the Bitcoin network would remain in huge danger of >50% attacks; bigger danger than is already given by financial (those who have money invest in mining farms) and geographical centralization (places with good insfrastructure and low energy costs attract mining farms) of Bitcoin amongst mining pools.

Secondly I want to ask how you understand Bitcoin as replacement for a flawed central bank? Bitcoin has nothing to do with a central bank. Bitcoin invented a technological solution to store and transfer something of value without the need to trust third parties. With Bitcoin you only need to trust the partner of the transaction and Bitcoin itself. That was a great invention, but has nothing to do with the repacement of a central bank.

At first sight I found it strange that the current Nu ecosystem looks like resembling a kind of central bank function. I asked myself “what the heck shall that be good for”. But then I came to the conclusion that the reason for having started this project make sense: the adoption of crypto currencies as payment method is low because of the high volatility and NuBits is trying to provide a solution for that.
Instead of having a central bank emitting a currency, we have a decentralized autonomous bank emitting a currency.
The advantage lies in the immunity against national and political abuse of that system (given that the NuShares are distributed wide enough).
When central banks fail ultimately, people in countries have to pay the bill (by having worthless currencies and having to pay for the debts of the central bank ). When “Nu bank” fails, the NuBits holders suffer from a loss, but the NuShares holders pay the rest of the bill.

Without a stable currency we have no real currency but valuable assets. I’m glad that Bitcoin created a solution for having valuable assets.

And I’m glad that Nu has started the attempt to create currency.

You ask what we achieve? A currency that might be backed by intrinsic value and not by force, threat and military. I think that alone is worth trying it.
There’s a lot to do. This is only the start. We have day 6 after the launch of Nu.
There need to be ways of generating revenue in addition to the destroyed tx fees. Once NuBits have reached some adoption one might think of loaning NuBits or investing NuBits in venture capital markets - just like the business cental banks do, but in this case based on votes from the “owners” of the “Nu bank”; votes from those who are accountable for the success of Nu.