The Deathblow to Proof of Stake

[ChuckOne]

if pos so good, why is litecoin still second. take that

Ridiculous argument. Try again.

He did his very best.

[LeChatNoir]

if pos so good, why is litecoin still second. take that

Litecoin is older and it's benefiting some kind of network effect as it is the "biggest altcoin" out there.
Give NXT a few more months 

[Chillin_with_beer]

if pos so good, why is litecoin still second. take that

Litecoin is older and it's benefiting some kind of network effect as it is the "biggest altcoin" out there.
Give PPC a few more months 

fixed

[allwelder]

if pos so good, why is litecoin still second. take that

You should know that price is not reflect the all situations,it is just a part for short time.
With time go,POS will give us exciting news.

[cwb27]

if pos so good, why is litecoin still second. take that

You should know that price is not reflect the all situations,it is just a part for short time.
With time go,POS will give us exciting news.

ohh that makes sense.

why is nxt better than ppc?

[allwelder]

if pos so good, why is litecoin still second. take that

You should know that price is not reflect the all situations,it is just a part for short time.
With time go,POS will give us exciting news.

ohh that makes sense.

why is nxt better than ppc?

I just list some difference ,but not want to offend PPCers.
1.NXT is first 100% POS with a brand new POS named TF(transparent forging)  .
2.NXT have no central checkpoint and no coin age design.
3.NXT have many features,such as Asset exchange,arbitrary message,Alias and some coming features,digital goods store,alias exchange,monetary system,and so on...

[allwelder]

if pos so good, why is litecoin still second. take that

Litecoin is older and it's benefiting some kind of network effect as it is the "biggest altcoin" out there.
Give NXT a few more months 

agree,litecoin was born in 2011,NXT 2013~2014,just time needed.

[lynn_402]

if pos so good, why is litecoin still second. take that

You should know that price is not reflect the all situations,it is just a part for short time.
With time go,POS will give us exciting news.

ohh that makes sense.

why is nxt better than ppc?

I just list some difference ,but not want to offend PPCers.
1.NXT is first 100% POS with a brand new POS named TF(transparent forging)  .
2.NXT have no central checkpoint and no coin age design.
3.NXT have many features,such as Asset exchange,arbitrary message,Alias and some coming features,digital goods store,alias exchange,monetary system,and so on...

There are arguments in the other direction, too.

1- PPC has a better distribution because of PoW.
2- PPC is less centralized, because one team doesn't have control on the whole architecture, and it was not instamined.
3- The central checkpoint are actually a good thing for now, since PoS has not been definitely proven to be secure. Central checkpoints allows some additional security for that period before the devs are 100% confident that the coin will be secure without the checkpoins.
4- There's Peershares being developped, which is an asset exchange built in through the Peerunity wallet client.

[ChronosCrypto]

A single point of failure
Mintpal was compromised. The attacker gained 30% of the total supply of Vericoin in the attack.

Well, this is the obvious problem -- Mintpal had 30% of the total supply. It could have attacked the network itself, without any hacker stealing any funds. Can you imagine one entity having 30% of the total USD in existence? Vericoin was too concentrated in the hands of too few. Over time, as a coin grows, it spreads to more people and this risk goes down.

[hicaribou]

if pos so good, why is litecoin still second. take that

Peercoin will be in the second place within several months and then compete with Bitcoin for the first place.

[ArticMine]

Can one not attack a proof of stake coin by purchasing the actual coin and at the same time selling short on the "paper / derivative" market thereby effectively maintaining a neutral or even a short position on the coin during the attack?

[devphp]

Can one not attack a proof of stake coin by purchasing the actual coin and at the same time selling short on the "paper / derivative" market thereby effectively maintaining a neutral or even a short position on the coin during the attack?

Short answer: No.

Long answer: it's a quick way for that exchange to be out of business if the exchange doesn't limit the short position of that customer (it's not like naked short selling is present on many exchanges now, because exchanges understand the limited supply nature of crypto currencies and that they can quickly go bankrupt if they allow this naked short selling, or at least with something over 1:2 leverage).

If the attacker withdraws, this will be noticeable immediately, because blockchain is transparent, people will notice the exchange's reserves are depleting and will do mass withdrawals too.

There is no economic incentive for an exchange to allow customers to have large short positions, because if the price stays the same, they lose on: a) fees (the larger the fiat price of a crypto, the larger the fee they can collect, because the fee is usually set in crypto); b) bad reputation for the exchange, people will notice that volume of trading is huge, but the price stays the same, meaning something funky is going on on that exchange, and will go to other exchanges. This funky business may have worked with Mt.Gox, because the willybot kept the price to the upside, so people were ok with that, but won't work, if volume is going up significantly, but price stays the same, it's not natural and means someone is doing something criminal.

Besides, in NXT, for example, there are plans to go to decentralized gateways and multigateways to deposit/withdraw, multigateway is already developed and functioning on live net, centralized exchanges don't fit the decentralized concept of NXT, they are a single point of failure, an entity you have to trust, hence they will not be relevant in the future.

I am sure, there are other nuances why this attack is not possible in practise, at least for an established crypto like NXT, but those above should be enough for now.

[ArticMine]

Can one not attack a proof of stake coin by purchasing the actual coin and at the same time selling short on the "paper / derivative" market thereby effectively maintaining a neutral or even a short position on the coin during the attack?

Short answer: No.

Long answer: it's a quick way for that exchange to be out of business if the exchange doesn't limit the short position of that customer (it's not like naked short selling is present on many exchanges now, because exchanges understand the limited supply nature of crypto currencies and that they can quickly go bankrupt if they allow this naked short selling, or at least with something over 1:2 leverage).

If the attacker withdraws, this will be noticeable immediately, because blockchain is transparent, people will notice the exchange's reserves are depleting and will do mass withdrawals too.

There is no economic incentive for an exchange to allow customers to have large short positions, because if the price stays the same, they lose on: a) fees (the larger the fiat price of a crypto, the larger the fee they can collect, because the fee is usually set in crypto); b) bad reputation for the exchange, people will notice that volume of trading is huge, but the price stays the same, meaning something funky is going on on that exchange, and will go to other exchanges. This funky business may have worked with Mt.Gox, because the willybot kept the price to the upside, so people were ok with that, but won't work, if volume is going up significantly, but price stays the same, it's not natural and means someone is doing something criminal.

Besides, in NXT, for example, there are plans to go to decentralized gateways and multigateways to deposit/withdraw, multigateway is already developed and functioning on live net, centralized exchanges don't fit the decentralized concept of NXT, they are a single point of failure, an entity you have to trust, hence they will not be relevant in the future.

I am sure, there are other nuances why this attack is not possible in practise, at least for an established crypto like NXT, but those above should be enough for now.

The attacker does not have to be an exchange, nor is there a need for a centralized market for the attack to work.  

It is possible today enter into BTC and LTC contracts for difference vs USD using an FX broker or brokers with something else such as USD or EUR as collateral. One can use this to create a short position on the coin. One then at the same time purchases the same amount of the actual coin, thereby not moving the market because of arbitrage between the FX broker or brokers and the actual coin market or markets. This kind of thing happens all the time with commodities, stocks and fiat currencies where the paper market exceeds the actual physical market by many orders of magnitude. Furthermore there are many valid reasons why people would do this kind of hedging. The only reason this does not work with current proof of stake coins is that they are still too young to create a viable derivatives market.

The problem with proof of stake is that the stake represented on the blockchain may not represent at all the exposure of the holder of the stake to the coin, if the holder of the stake has hedged her exposure to the coin with derivatives. The more established the coin, the more likelihood of a derivatives market around the coin and consequently the greater the risk of this kind of attack.

[devphp]

It is possible today enter into BTC and LTC contracts for difference vs USD using an FX broker with something else such as USD or EUR as collateral. One can use this to create a short position on the coin. One then at the same time purchases the same amount of the actual coin, thereby not moving the market because of arbitrage between the FX broker or brokers and the market. This kind of thing happens all the time with commodities, stocks and fiat currencies where the paper market exceeds the actual physical market by many orders of magnitude. Furthermore there are many valid reason why people would do this kind of hedging. The only reason this does not work with current proof of stake coins is that they are still young to create a derivatives market.

The problem with proof of stake is that stake represented on the blockchain may not represent at all the exposure of the holder of the stake to the coin if the holder of the stake has hedged her exposure to the coin with derivatives. The more established the coin, the more likelihood of a derivatives market around the coin and consequently the greater the risk of this attack.

Yes, I know it's possible with BTC and LTC with some brokers, but BTC and LTC are not Proof-of-Stake, hence the exchange itself doesn't have to make sure the customer is not doing funky operations with those coins. In case of PoS, it will be in the exchange's own economic interests to make sure that is not happening, because it will jeopardize their reputation (see above for huge volume and price staying the same, this is a big red flag not to trade on that exchange and be sure it will be publicized very quickly).

Paper derivatives market in cryptos is not going to be anything large to worry about due to the fact that it's very easy to take delivery of crypto currencies, whereas it would be a major hassle if you tried that with commodities in real life (95% of commodity traders do not take delivery and brokers know that, hence they allow them to do highly leveraged trading - some allow 1:10, 1:25 leverage for commodities which is crazy). This incurs the risk of the exchange's reserves being depleted, immediately noticeable by everyone on the blockchain. So, yes, commodities, fiat currencies, stocks - paper derivatives on top of them are very easy to implement. As for crypto currencies - it's not going to work on any large scale.

If the price stays the same, it will be easy for others to buy and withdraw, soon there won't be enough for everyone. The price has to rise to satisfy demand, depleting people's fiat that they can buy with. If that doesn't happen, the available supply will soon be bought up and withdrawn from the exchange.

Please see my previous post about decentralized gateways also, that NXT is implementing making centralized exchanges irrelevant.

[ArticMine]

It is possible today to get up to 20:1 margin on Bitcoin CFDs and up to 10:1 margin on Litecoin CFDs today using an FX broker not an exchange. http://www.avatrade.com/trading-info/trading-conditions There is no need to use just one FX broker for "paper" or just one exchange for "actual" for this attack to work. In fact the more decentralized the market for both the "paper" and the "actual" the better since this will allow the attacker to hide her hand.  

[devphp]

It is possible today to get up to 20:1 margin on Bitcoin CFDs and up to 10:1 margin on Litecoin CFDs today using an FX broker not an exchange. http://www.avatrade.com/trading-info/trading-conditions There is no need to use one FX broker for "paper" or one exchange for "actual" for this attack to work. In fact the more decentralized the market for both the "paper" and the "actual" the better since that will allow the attacker to hide her hand. 

Correct me if I am wrong, but CFDs is not the actual commodity being traded. There is no delivery of commodity if CFD is bought, everything is cash settled. CFD price is following price of commodity exchanges or crypto exchanges in the case of cryptos, not leading it.

So if an attacker buys actual crypto commodity in large amounts on one exchange with one hand and at the same time sells CFD with another hand at another broker/exchange, the price on the first exchange would start growing, everyone will rush to buy (people like to buy when the price grows), the price will be growing even more, arbitrageurs will quickly jump on board and make sure the price is about the same on all exchanges and with all brokers (now the attacker's short position is well under water).

If the attacker keeps persisting and demand from the attacker and other people keeps growing, the attacker will soon have to cover their short position in CFDs with a huge loss because the broker will send a margin call or would have to add cash to maintain it. In the end, because people take delivery of the crypto, the attacker would have to close short positions at a loss, no matter how much fiat the attacker has, because the crypto would be depleted at exchanges and price goes to infinity (remember: 95-99% of traders don't take delivery of wheat, copper or steel, but many take delivery of cryptos because it's very easy to do).

[ArticMine]

It is possible today to get up to 20:1 margin on Bitcoin CFDs and up to 10:1 margin on Litecoin CFDs today using an FX broker not an exchange. http://www.avatrade.com/trading-info/trading-conditions There is no need to use one FX broker for "paper" or one exchange for "actual" for this attack to work. In fact the more decentralized the market for both the "paper" and the "actual" the better since that will allow the attacker to hide her hand.  

Correct me if I am wrong, but CFDs is not the actual commodity being traded. There is no delivery of commodity if CFD is bought, everything is cash settled. CFD price is following price of commodity exchanges or crypto exchanges in the case of cryptos, not leading it.

So if an attacker buys actual crypto commodity in large amounts on one exchange with one hand and at the same time sells CFD with another hand at another broker/exchange, the price on the first exchange would start growing, everyone will rush to buy (people like to buy when the price grows), the price will be growing even more, arbitrageurs will quickly jump on board and make sure the price is about the same on all exchanges and with all brokers (now the attacker's short position is well under water).

If the attacker keeps persisting and demand from the attacker and other people keeps growing, the attacker will soon have to cover their short position in CFDs with a huge loss because the broker will send a margin call or would have to add cash to maintain it. In the end, because people take delivery of the crypto, the attacker would have to close short positions at a loss, no matter how much fiat the attacker has, because the crypto would be depleted at exchanges and price goes to infinity (remember: 95-99% of traders don't take delivery of wheat, copper or steel, but many take delivery of cryptos because it's very easy to do).

Yes CFDs are cash settled this is correct; however there is a counter party to the CFDs that has to hedge her position. This has to ultimately happen by selling the actual, either by the market makers at the brokers or by other market participants engaging in arbitrage selling the actual (at a small premium) and buying the paper at (a small discount). So it is very possible for the paper market to lead the actual market. If both the actual long and paper short are acquired gradually and across multiple brokers and exchanges the impact on the market is minimal. One must keep in mind that until the moment of the attack the attacker can keep her net position neutral, then increase her position to net short just before the attack to profit for the plunge in price caused by the chaos resulting from the attack.

Your point about taking delivery with crypto-currencies is very valid; however in this case it only serves to ensure that those writing the CFDs maintain tightly hedged positions themselves to minimize their risk, thereby keeping the paper and actual markets in sync. This will magnify the effect of the attack.

Edit: This attack works best when the market is otherwise moving sideways.

[devphp]

Yes CFDs are cash settled this is correct; however there is a counter party to the CFDs that has to hedge her position. This has to ultimately happen by selling the actual, either by the market makers at the brokers or by other market participants engaging in arbitrage selling the actual (at a small premium) and buying the paper at (a small discount). So it is very possible for the paper market to lead the actual market. If both the actual long and paper short are acquired gradually and across multiple brokers and exchanges the impact on the market is minimal. One must keep in mind that until the moment of the attack the attacker can keep her net position neutral, then increase her position to net short just before the attack to profit for the plunge in price caused by the chaos resulting from the attack.

Your point about taking delivery with crypto-currencies is very valid; however in this case it only serves to ensure that those writing the CFDs maintain tightly hedged positions themselves to minimize their risk, thereby keeping the paper and actual markets in sync. This will magnify the effect of the attack.

How do you say the impact on the market is minimal when the attacker is looking to acquire... how much, 20-50% of total PoS coin in circulation? These actions will generate huge volume across all exchanges, the attacker's own volume or counter parties' who hedge their bets. The markets of cryptos are very thin (even Bitcoin, not to mention less known coins), the price will jump and generate a lot of demand from speculators, many of whom withdraw. Soon the price on the actual crypto market will detach from paper market derivatives, the actual crypto will be traded at a premium to paper derivatives, this again noticeable to everyone. Paper derivatives will be losing their credibility, same as happening with some real commodities. The more they lose credibility, the more the premium the actual crypto/commodity trades at.

For an established PoS coin like NXT actions of a rogue trader are unlikely to cause chaos. There were a few million of NXTs dumped on the market by a hacker who stole from kLee, the price spiked down, but recovered very quickly to the level before the dump. The lower the price, the more value traders come out to buy an established valuable crypto and the more they can acquire for the same amount of fiat, so the attacker would be shooting themselves in the foot very quickly with short trading. And because the traded supply on exchanges is very thin relative to total supply of NXTs, the attacker or the attacker's counter parties would have to own too much of the coin to be able to constantly dump to break the market to the downside, because all those value traders would be out there collecting cheap coins dumped on to them. I fail to see where the counter parties of the attacker would have got that many coins from to sell at the actual market.

What you describe here is a theoretical risk that is two orders of magnitude more difficult to realize (in terms of cash you need) than the 51% attack on a PoW coin.

[ThomasCrowne]

The unprecedented solution
A rollback is terrible. Every single cryptocurrency relies on the public blockchain ledger. It is the holy grail of the entire currency. Once something is written to it and not orphaned, it's set in stone. When the team decided to initiate the rollback, they decided to use the nuclear option. They broke the entire foundation of crypto and set a new norm where it will be ok to undo transactions if the are large enough. Instead of the developers only being developers, they've now taken the option to also be the federal reserve and the police.

I'd like to understand this better, because it has been bothering me for a while. Can devs really just unilaterally decide to roll back a blockchain in PoS? Aren't there other players involved who have to go along with the idea? Lately I've seen many examples of coin devs announcing forks and changes to their coin attributes, and I've had a hard time figuring out just how much power a dev has by themself, and how much has to be a concensus decision with other parts involved - and whom are they? Miners, exchanges, holders of wallets, etc?

I can't believe devs have some secret key by which they can just single-handedly alter a coin contrary to everyone else's wishes. That would destroy all faith in such coins IMHO. So please help me understand this.

This is the first time a rollback has ever been performed due to outside circumstances like an exchange failing to implement proper security. In essence, it's a bailout for the exchange. A rollback is the nuclear option, (never been done before) it might cause fallout for every single cryptocurrency in existence.

time to put mintpal in the too big to fail category :-p

[r0ach]

Lately I've seen many examples of coin devs announcing forks and changes to their coin attributes, and I've had a hard time figuring out just how much power a dev has by themself, and how much has to be a concensus decision with other parts involved - and whom are they? Miners, exchanges, holders of wallets, etc?

Yea, it will be interesting when at some point every exchange converts to a new fork for some random coin and miners (pools) refuse, or vice versa.  Right now, everyone just kinda goes along to get along, but at some point there will be a clash of civilizations.  It's mostly exchanges that dictate everything right now since pool owners don't want to look like idiots having their clientele mining thin air that can't be cashed out.  

This would imply the power to fork comes from the developer and exchanges, and miners come 3rd, which is basically the opposite of the Satoshi view.  The Satoshi view seems to rely on the idea that there will be vastly more exchanges than mining pools, so whatever a particular exchange does has no influence on anything.  Most altcoins are entirely centralized on one exchange.  Bitcoin could be very centralized the same way in that regard having one dominant exchange in the US, and one in Europe or Hong Kong even years from now.