Sorry, I don’t understand which market is which, or how this affects the “attack” above. Could you please rephrase that?
It’s not just the act of placing an order on the exchange, is it? Doesn’t the large buy order placed on BTC-e have to be actually purchased before NuBot would go ahead and replicate the order on NBT books? In that case there’s nothing for the attacker to remove (while simultaneously selling into the custodian’s order).
The way I understand it, it’s a mirroring of the open BTC-e orders, not a mirroring of already-filled orders.
Ah, Jordan already considered this attack.
Does this mean that custodial funds must now be held on the mirrored exchange(s)? For example, custodial funds would be held on BTC-e, increasing exchange default risk? Up to this point, this has not been a requirement.
There are several technical obstacles that will make the mirror quite unstable and subject to attacks.
I submitted an idea for an alternative solution to Jordan for consideration.
In short :
Alternative : Instead of depending 1:1 to other order books, I would pre-prepare N models of orders/liquidity distribution , for example flat, balanced, deep, shallow, steep, bear, bullish. Then instruct nubot which model to use and change it when necessary according to market conditions.
This will make it much more solid and reliable (no need to read/cancel/place orders multiple times a minute!) and will solve the problem of offering huge walls around 1$, will give us control over the models, will eliminate third party failure and attacks scenarios, will allow for wall shifts to happen smoothly, and will also make our current multi-custodians approach feasible.
Or maybe not. Just an idea
That is very ambitious. You would need real data to guide algorithm development, to make forward and backward tests. So far I haven’t seen real data of custodian losses to judge. Complex technical solutions tend to worry me. You would need to make simulations and test harnesses. Some of the brightest minds I know are hired full time to analyze financial data and deveolope strategies and models. Obviously we have a dedicated, intelligent and knowledgeable team. But we also have a lot to do so we will probably be facing more specialized and sophisticated attackers who can spend a lot of time finding weakness on certain technical aspect.
So I am in favor of first identifying indicators to show the problem to solve, then setting up monitors quantitative indicator, then developing robust and simple solutions.
I support this motion.
Glad to hear it. This is something the network needs. I hope you will submit a proposal for a custodial grant.
It is. But isn’t it less ambitious than making real time interpolation and mirroring of multiple order books?
I thought you were proposing to make multiple models. I was suggesting make a robust model in three steps. Both approached would need to process real time data feed.
I expect a model generic enough that requires to be switched very few times a day, versus multiple time a second
Will the algorithm be reviewed here? I think an analysis and trade off study would be helpful.
You bet!
@JordanLee This is btw the proposal I have made. [Passed] LPC grant request for supporting NBT/USD on CCEDK
At the present moment this has about 38% support. I believe I need to take the time to respond to a number reservations that have been expressed both here and privately. I will also say more about how it should be implemented, which I think will help people understand mirroring order books won’t be as technically difficult as some think right now. In fact it will greatly simplify multiple LPCs supporting the same pair. So, expect quite a bit more info about this in the coming week.
The plan expressed in this motion is the most important way we can increase NuBit adoption at this stage, in my opinion. I believe NuBits can replace both USD and BTC as intermediate currencies with the plan I have outlined above. What I mean by intermediate currencies is when I someone wants to trade their Blackcoin for Peercoin they cannot do so directly, so they use either USD or BTC as an intermediary: they trade Blackcoin for BTC and then use the BTC to buy Peercoin. In the future (when this motion is passed and implemented), people will want to use NuBits as the intermediate currency because it will offer superior liquidity. That NuBits are stable is another advantage over BTC, as people performing this kind of trade don’t have an interest in speculating on the value of BTC.
I am looking forward for the upcoming technical clarifications.
But before that, personally, I would also like to receive a clarification of what I wrote in post#46 . It basically boils down to :
Do we really want NuBits to be an intermediary coin? Why?
And can be expanded to several subquestions : Is being an intermediary coin part of our strategy to become widely adopted as a stable currency? Is there any sign that being used as an intermediary will drive adoption of NuBits in other contexts?
What are the benefits for shareholders?
Do we have clear signs from the market?
Should we update our whitepaper? Our marketing targets, slogans, and campaigns?
“The world first stable intermediary currency”
Should we change our description of “What can you do with nubits?” ?
I believe that this change - forgetting technical challenges - deserves more attention, and a proper discussion followed by a motion.
I will chip in what I think.
I have been concerned that Nubits is running on fractional reserve and there doesn’t seem to be an obvious or quick solution to it. I do realize that fractional reserve is an effcient way to use capital. The crux is where to draw the line between being so conservative that Nu economy is choked off and being so adventurous that we ended up having a catastrophic loss of the peg.
I don’t know where to draw the line yet but I think I see how to increase the safety margine, to broaden the line into a buffer zone. The trick is to create a demand for nubits such that its users don’t really care whether there is localized or short term doubt of the peg. Such use already exists. Bitcoin fluctuates greatly but people still consider it safe as a medium of exchange because it is often “good enough”. This utility of bitcoin creates a constant demand of it. People don’t rush to the exchanges to cash out bitcoin just because there is a couple of percent drop of BTC/USD. This fraction of bitcoin that is effectively permanently in users hand is what I consider safety margin.
Nubits could have this safety margin, just miles broader. The period of bitcoin’s being considered “good enough”, as mentioned above, is inversely proportional to how stable bitcoin is. If nubits is widely used as a intermediary currency in commerce, then I expect many users will hold onto it regardless where short term price might drift. I will go as far claiming people will use it even the peg will slowly drift away from at strict 1:1. This part of “nubits with velocity” is the coushion of fractional reserve. (I will post an brief review of a paper that discussed money with velocity)
Therefore I see pushing nubits to be an intermediary coin on cryptocurrecny exchanges to be of great strategic importance for Nu’s security.
I agree if it is being under control. The beauty with cryptos is that we are compelled to make everything transparent in order to keep the integrity and the trust into the trustless decentralized system.
So I think having a real time tool that lets shareholders and users know about the quantity of nubits in circulation, the quantity of nubits parked, the reserves and the buy liquidity on every NBT/* pair is very important so that we know to which extent Nu is “leveraged”.
I agree on that. I mean it is just a matter of trust. And the trust can absolutely exist under a fractional regime as long as the degree of the fractional aspect is “reasonable”, I believe.
Look at FIAT money. It is probably highly leveraged but still most people use it all the time.
We need to define and measure what is “reasonable” though.