[OFFICIAL]Bitfinex.com first Bitcoin P2P lending platform for leverage trading

[deisik]

I believe that they allow 3.3x initial leverage and require 15% maintenance margin in order to maintain positions. By my calculations, this means that there would need to be a very swift change in price in excess of 10% for the possibility of individual accounts to have negative equity.

Users do not have to use the maximum amount of leverage (although lenders do not have any control over this).

I would point out that, according to https://www.bfxdata.com/positions/btcusd as of Jan 6 at 8:59, the price of BTC was 1,151.16, and there were 37,349BTC in outstanding BTCUSD long positions, and as of Jan 6 at 17:59, the price of BTC was 962.26 (down ~16.4%), and there were 26,422BTC in outstanding BTCUSD long positions (down ~10,900BTC, or ~29.25%). This did of course follow a very sharp upward movement in the BTCUSD price, and is only one example of how the market on Bitfinex specifically handled one specific quick selloff.

I don't quite understand what you mean to say

But maybe it's just me, after all. Regarding the amount of leverage, 3.3:1 leverage means that the trader can potentially have 3.3 times the amount that he actually has. In this manner, if he has, say, only 100 dollars he could buy bitcoins for the total of 330 dollars (including his own money). If the price goes higher, that would give him 3.3 times more profit, but if the price goes down dramatically he could potentially incur more losses than he has funds in his account. With 3.3:1 leverage that would happen if the price went down 33%, but since the exchange has the right to forcefully sell his bitcoins (to prevent them from losses), they will do that even before the price comes down to the point where the account will get wiped away completely

[Rogerdale]

I would point out that, according to https://www.bfxdata.com/positions/btcusd as of Jan 6 at 8:59, the price of BTC was 1,151.16, and there were 37,349BTC in outstanding BTCUSD long positions, and as of Jan 6 at 17:59, the price of BTC was 962.26 (down ~16.4%), and there were 26,422BTC in outstanding BTCUSD long positions (down ~10,900BTC, or ~29.25%). This did of course follow a very sharp upward movement in the BTCUSD price, and is only one example of how the market on Bitfinex specifically handled one specific quick selloff.

Where does bfxdata get the information from? Does it get it from all exchanges or bitfinex only?

[Sukrim]

As the name implies, bfxdata charts date from Bitfinex.

[Rogerdale]

Are they using a cold wallet now or still rely on BitGo? By the way, does anybody know how a hacker hacked both Bitfinex and BitGo and passed BitGo's 2 factor authentication?

[Ente]

They don't use BitGo any more, and are back to their original hot/cold wallet setup.
The hack was only on BitFinex. The hacker used the keys from BitFinex to change the limit on BitGo. Amazingly, up to 1M BTC a day, over the BitGO API directly, which was not a documented feature. Without raising questions or warnings from BitGo. Then the hacker went on to send many many transactions, signed with BitFinex' compromised key, to BitGo, which they co-signed without question.
That's one of he worst setups I ever heard, on BitGos side. But the original problem was the compromise of BitFinex' servers.

Ente

[deisik]

They don't use BitGo any more, and are back to their original hot/cold wallet setup.
The hack was only on BitFinex. The hacker used the keys from BitFinex to change the limit on BitGo. Amazingly, up to 1M BTC a day, over the BitGO API directly, which was not a documented feature. Without raising questions or warnings from BitGo. Then the hacker went on to send many many transactions, signed with BitFinex' compromised key, to BitGo, which they co-signed without question.
That's one of he worst setups I ever heard, on BitGos side. But the original problem was the compromise of BitFinex' servers

This raises a lot more questions really

If BitGo just blindly signed the Bitfinex transactions without asking questions, what was the purpose of them being there in the first place? I don't know how that system works or should work but common sense tells me that BitGo is useless in this setup, i.e. if we remove it completely this won't change anything at all. I assume their primary aim is to provide additional channel for confirmations (like SMS codes in 2FA). Or does it mean that they are involved in the theft too?

[Quickseller]

I believe that they allow 3.3x initial leverage and require 15% maintenance margin in order to maintain positions. By my calculations, this means that there would need to be a very swift change in price in excess of 10% for the possibility of individual accounts to have negative equity.

Users do not have to use the maximum amount of leverage (although lenders do not have any control over this).

I would point out that, according to https://www.bfxdata.com/positions/btcusd as of Jan 6 at 8:59, the price of BTC was 1,151.16, and there were 37,349BTC in outstanding BTCUSD long positions, and as of Jan 6 at 17:59, the price of BTC was 962.26 (down ~16.4%), and there were 26,422BTC in outstanding BTCUSD long positions (down ~10,900BTC, or ~29.25%). This did of course follow a very sharp upward movement in the BTCUSD price, and is only one example of how the market on Bitfinex specifically handled one specific quick selloff.

I don't quite understand what you mean to say

But maybe it's just me, after all. Regarding the amount of leverage, 3.3:1 leverage means that the trader can potentially have 3.3 times the amount that he actually has. In this manner, if he has, say, only 100 dollars he could buy bitcoins for the total of 330 dollars (including his own money). If the price goes higher, that would give him 3.3 times more profit, but if the price goes down dramatically he could potentially incur more losses than he has funds in his account. With 3.3:1 leverage that would happen if the price went down 33%, but since the exchange has the right to forcefully sell his bitcoins (to prevent them from losses), they will do that even before the price comes down to the point where the account will get wiped away completely

If someone has $1,000 in USD, they can buy 3.3BTC @$1,000 each (ignoring trading fees, and interest payments). This person will owe $3,300 to lenders. If the price of BTC declines to $850, then the 3.3BTC that he purchased is now worth $2,805, he will still have his initial $1,000 and will have total assets of $3,805; he will still owe $3,300 to lenders, so his net equity is $505, or 15.3% of what he owes to lenders. He would get a margin call at roughly this time.

If the price were to fall to $697 (18% decline from $850), then the value of that 3.3BTC would be $2,300.1, plus his initial $1,000 would mean that his total assets are $3,300.1 and the amount lent to lenders would be $3,300, however after factoring in trading fees and interest, his assets would be less than the amount owed.

So it seems that the more precise statement is that the price will need to fall by roughly 18% from a point very quickly for there to be a risk that individual accounts having negative equity.

I would point out that it is theoretically possible for someone to buy 3.3BTC in the above scenario, and for the price of BTC to remain the same for a long enough period of time, and the interest payments cause a margin call.

[Quickseller]

They don't use BitGo any more, and are back to their original hot/cold wallet setup.
The hack was only on BitFinex. The hacker used the keys from BitFinex to change the limit on BitGo. Amazingly, up to 1M BTC a day, over the BitGO API directly, which was not a documented feature. Without raising questions or warnings from BitGo. Then the hacker went on to send many many transactions, signed with BitFinex' compromised key, to BitGo, which they co-signed without question.
That's one of he worst setups I ever heard, on BitGos side. But the original problem was the compromise of BitFinex' servers

This raises a lot more questions really

If BitGo just blindly signed the Bitfinex transactions without asking questions, what was the purpose of them being there in the first place? I don't know how that system works or should work but common sense tells me that BitGo is useless in this setup, i.e. if we remove it completely this won't change anything at all. I assume their primary aim is to provide additional channel for confirmations (like SMS codes in 2FA). Or does it mean that they are involved in the theft too?

It is my understanding that the initial setup involved BitGo signing up to 2,000BTC per day out of Bitfinex's wallets/addresses. It would be possible to change this, but the channel to change this was different from the channel to send transactions to BitGo for their signature.

BitGo offered an API that it's customers could use to send transactions to BitGo for BitGo's signature. The high volume nature of Bitfienx's transactions sent to BitGo meant that the API had to always be logged in, and for obvious reasons, the Bitfinex server that connected to the BitGo API needed to always be online.

It is my understanding that not long before the hack, that BitGo changed their API so that it was possible to change security settings, including the ability to increase the maximum number of coins that BitGo will allow to be withdrawn from Bitfinex's wallets/addresses via transactions that BitGo signed.

[deisik]

I believe that they allow 3.3x initial leverage and require 15% maintenance margin in order to maintain positions. By my calculations, this means that there would need to be a very swift change in price in excess of 10% for the possibility of individual accounts to have negative equity.

Users do not have to use the maximum amount of leverage (although lenders do not have any control over this).

I would point out that, according to https://www.bfxdata.com/positions/btcusd as of Jan 6 at 8:59, the price of BTC was 1,151.16, and there were 37,349BTC in outstanding BTCUSD long positions, and as of Jan 6 at 17:59, the price of BTC was 962.26 (down ~16.4%), and there were 26,422BTC in outstanding BTCUSD long positions (down ~10,900BTC, or ~29.25%). This did of course follow a very sharp upward movement in the BTCUSD price, and is only one example of how the market on Bitfinex specifically handled one specific quick selloff.

I don't quite understand what you mean to say

But maybe it's just me, after all. Regarding the amount of leverage, 3.3:1 leverage means that the trader can potentially have 3.3 times the amount that he actually has. In this manner, if he has, say, only 100 dollars he could buy bitcoins for the total of 330 dollars (including his own money). If the price goes higher, that would give him 3.3 times more profit, but if the price goes down dramatically he could potentially incur more losses than he has funds in his account. With 3.3:1 leverage that would happen if the price went down 33%, but since the exchange has the right to forcefully sell his bitcoins (to prevent them from losses), they will do that even before the price comes down to the point where the account will get wiped away completely

If someone has $1,000 in USD, they can buy 3.3BTC @$1,000 each (ignoring trading fees, and interest payments). This person will owe $3,300 to lenders. If the price of BTC declines to $850, then the 3.3BTC that he purchased is now worth $2,805, he will still have his initial $1,000 and will have total assets of $3,805; he will still owe $3,300 to lenders, so his net equity is $505, or 15.3% of what he owes to lenders. He would get a margin call at roughly this time

It seems that I pinpointed where you make a mistake

If a trader has 1,000 dollars and the leverage ratio is 3.3:1, it means that he would owe only 2,300 dollars (not 3,300 as you say). Basically, the leverage of 1:1 just means that there is no leverage altogether and the trader is using his own funds (i.e he doesn't borrow). So he will get a margin call when the price goes down to roughly 700 dollars (which is equal to 30% as I said). In that case, his remaining funds will be just enough to cover his debt (350 dollars on every 1,150 dollars borrowed)

[ETCethereum]

i want to ask, bitfinex still trusted , deposit and withdraw normal and fast or not
thank you

[pbleak]

i want to ask, bitfinex still trusted , deposit and withdraw normal and fast or not
thank you

It's fine in general day to day terms. Just be aware that the hack was possibly insider so you can never be sure. Just don't put much on an exchange as a rule. Alternatively just use one that has never been hacked.

[Ente]

i want to ask, bitfinex still trusted , deposit and withdraw normal and fast or not
thank you

It's fine in general day to day terms. Just be aware that the hack was possibly insider so you can never be sure. Just don't put much on an exchange as a rule. Alternatively just use one that has never been hacked.

Heh.
That's like saying "only drive in cars that never had an accident".
The correct solution is to never give up your bitcoins - especially not to a place where hundredthousands of them are laying, where there are no legal securities and no insurance.
If you need to exchange coins, sure, put a few up. But as few as possible, and for as short as possible.

Many have to learn that the hard way. I learned it the hard way.

Ente

[deisik]

i want to ask, bitfinex still trusted , deposit and withdraw normal and fast or not
thank you

It's fine in general day to day terms. Just be aware that the hack was possibly insider so you can never be sure. Just don't put much on an exchange as a rule. Alternatively just use one that has never been hacked.

Heh.
That's like saying "only drive in cars that never had an accident"

In fact, there is a reason behind this

Would you drive a car which had been in an accident in the past, and you don't what was the real cause that made it happen? I guess common sense would tell you to stay away from this car until you find out the exact cause of the accident (and fix it). What if the brakes are out of order in this car? It is basically the same with Bitfinex. Basically, we don't know for certain what happened there. If it was an inside job, and they didn't "fix it", you can be dead sure that it will happen again sooner or later (that's what happened to Bter)

[Ente]

i want to ask, bitfinex still trusted , deposit and withdraw normal and fast or not
thank you

It's fine in general day to day terms. Just be aware that the hack was possibly insider so you can never be sure. Just don't put much on an exchange as a rule. Alternatively just use one that has never been hacked.

Heh.
That's like saying "only drive in cars that never had an accident"

In fact, there is a reason behind this

Would you drive a car which had been in an accident in the past, and you don't what was the real cause that made it happen? I guess common sense would tell you to stay away from this car until you find out the exact cause of the accident (and fix it). What if the brakes are out of order in this car? It is basically the same with Bitfinex. Basically, we don't know for certain what happened there. If it was an inside job, and they didn't "fix it", you can be dead sure that it will happen again sooner or later (that's what happened to Bter)

What you say isn't wrong, not at all.
What I'm saying is that all cars are insanely dangerous, high-speed high-weight deathbullets carrying fragile humans inside. And that zerohedges slogan has some truth in it.
The real question, of course, is if you trust yourself more than a centralized exchange..

Ente

[deisik]

i want to ask, bitfinex still trusted , deposit and withdraw normal and fast or not
thank you

It's fine in general day to day terms. Just be aware that the hack was possibly insider so you can never be sure. Just don't put much on an exchange as a rule. Alternatively just use one that has never been hacked.

Heh.
That's like saying "only drive in cars that never had an accident"

In fact, there is a reason behind this

Would you drive a car which had been in an accident in the past, and you don't what was the real cause that made it happen? I guess common sense would tell you to stay away from this car until you find out the exact cause of the accident (and fix it). What if the brakes are out of order in this car? It is basically the same with Bitfinex. Basically, we don't know for certain what happened there. If it was an inside job, and they didn't "fix it", you can be dead sure that it will happen again sooner or later (that's what happened to Bter)

What you say isn't wrong, not at all.
What I'm saying is that all cars are insanely dangerous, high-speed high-weight deathbullets carrying fragile humans inside. And that zerohedges slogan has some truth in it.
The real question, of course, is if you trust yourself more than a centralized exchange

What ZeroHedge slogan do you refer to?

I don't think it is a question of trust. Obviously, there is no trust in any such exchanges. But we don't have a lot of choice in this regard. If we want to earn profits or just convert our bitcoins to fiat (and vice versa) and get market prices at that, we may have to use exchanges after all (or similar centralized services). There had been plenty of talk about implementing decentralized distributed exchanges over the blockchain, but we are not there yet by any metric. Maybe, Lightning Network will change something with its close to instant transactions

[Quickseller]

There had been plenty of talk about implementing decentralized distributed exchanges over the blockchain, but we are not there yet by any metric. Maybe, Lightning Network will change something with its close to instant transactions

This will never be possible.

1 - Once you sell your btc for USD, your USD will be at risk.

2 - If you have any orders on the exchange will be at risk because the exchange is not going to let you make an order without the coins being controlled by the exchange -- otherwise customers would have the opportunity to not sign a transaction and could potentially delay signing to wait and see if they can repurchase the sold coins at a lower price very soon after they sell.

[deisik]

There had been plenty of talk about implementing decentralized distributed exchanges over the blockchain, but we are not there yet by any metric. Maybe, Lightning Network will change something with its close to instant transactions

This will never be possible.

1 - Once you sell your btc for USD, your USD will be at risk.

2 - If you have any orders on the exchange will be at risk because the exchange is not going to let you make an order without the coins being controlled by the exchange -- otherwise customers would have the opportunity to not sign a transaction and could potentially delay signing to wait and see if they can repurchase the sold coins at a lower price very soon after they sell

Previously, I also thought along these lines basically

Though, unlike you, I understood that I couldn't know everything, and I assumed that there was a way to make such an exchange irreversible. And yes, some guys have explained to me how a decentralized distributed exchange can be done in an efficient manner. Basically, if you sell or buy bitcoins, you provide collateral in the form of the same bitcoins (say, twice as much the amount you buy or sell), so if you try to welsh on the deal, you lose more. Right now, this system wouldn't be usable for obvious reasons (due to transaction fees and confirmation delays), but with Lightning Network properly implemented it should work flawlessly and reliably

[Ente]

What ZeroHedge slogan do you refer to?

Why, it's

Code:

On a long enough timeline the survival rate for everyone drops to zero.

of course :-)

I don't think it is a question of trust. Obviously, there is no trust in any such exchanges. But we don't have a lot of choice in this regard. If we want to earn profits or just convert our bitcoins to fiat (and vice versa) and get market prices at that, we may have to use exchanges after all (or similar centralized services). There had been plenty of talk about implementing decentralized distributed exchanges over the blockchain, but we are not there yet by any metric. Maybe, Lightning Network will change something with its close to instant transactions
[/quote]

Exchanges can and will show up in a decentralized, secure way, I am sure. The real gambling, err, trading, is in the margin trading. Now that's a different thing, where I don't know if this can even be possible in a decentralized, secure way. Maybe with smart contracts ^^

Ente

[deisik]

What ZeroHedge slogan do you refer to?

Why, it's

Code:

On a long enough timeline the survival rate for everyone drops to zero.

of course :-)

Kinda death is irreversible no matter what we try, life is not recyclable we are born to die. On the other hand, when we move closer and closer to a limit (i.e. zero or infinity), virtually everything becomes possible

I don't think it is a question of trust. Obviously, there is no trust in any such exchanges. But we don't have a lot of choice in this regard. If we want to earn profits or just convert our bitcoins to fiat (and vice versa) and get market prices at that, we may have to use exchanges after all (or similar centralized services). There had been plenty of talk about implementing decentralized distributed exchanges over the blockchain, but we are not there yet by any metric. Maybe, Lightning Network will change something with its close to instant transactions

Exchanges can and will show up in a decentralized, secure way, I am sure. The real gambling, err, trading, is in the margin trading. Now that's a different thing, where I don't know if this can even be possible in a decentralized, secure way. Maybe with smart contracts

Margin trading would be no different from regular trading

As to me , there is no particular difference with regular trading at all (i.e. without using borrowed funds). Just like when trading with no margin you can't sell more than you have, it is the same with margin trading, though the price change required to wipe your account will be less (it will be just above 0 by a certain margin, pardon the pun). In this way, I don't see any insurmountable difficulty in setting the limit on the price change which will make your funds forcefully overtaken. This is a minor technical issue, I can't even call it a difficulty as such

[Quickseller]

There had been plenty of talk about implementing decentralized distributed exchanges over the blockchain, but we are not there yet by any metric. Maybe, Lightning Network will change something with its close to instant transactions

This will never be possible.

1 - Once you sell your btc for USD, your USD will be at risk.

2 - If you have any orders on the exchange will be at risk because the exchange is not going to let you make an order without the coins being controlled by the exchange -- otherwise customers would have the opportunity to not sign a transaction and could potentially delay signing to wait and see if they can repurchase the sold coins at a lower price very soon after they sell

Previously, I also thought along these lines basically

Though, unlike you, I understood that I couldn't know everything, and I assumed that there was a way to make such an exchange irreversible. And yes, some guys have explained to me how a decentralized distributed exchange can be done in an efficient manner. Basically, if you sell or buy bitcoins, you provide collateral in the form of the same bitcoins (say, twice as much the amount you buy or sell), so if you try to welsh on the deal, you lose more. Right now, this system wouldn't be usable for obvious reasons (due to transaction fees and confirmation delays), but with Lightning Network properly implemented it should work flawlessly and reliably

I am not sure if you are specifically referring to this, however the system you are describing sounds a lot like bitsquare. With bitsquare, traders are mostly protected against a hack of the exchange, although the arbitrators essentially are bearing the risk of the exchange being hacked, and I don't think there is a way of getting around this. Traders also need to trust bitsquare to the extent that multiple arbitrators are not colluding with each other maliciously. Probably very importantly, users will need to know how to detect fraudulent fiat payments, and dispute the receipt of the payment -- this is a skill that many current bitcoin users do not have, and one that potential new users have in less frequently. This kind of exchange will also result in less liquidity than say bitfinex as traders will need to wait longer to receive fiat funds.

If the system you describe would work so successfully with an instant-confirmation system (like LN), then platforms like localbitcoins would implement similar systems (especially for difficult-to-reverse type payment methods).