BitShares XT - Security against Market Manipulation FIND ATTACKS FOR TIPS

[tonyk]

!!!This is not an ‘attack’!!!

OK – please move this discussion to something else – like ‘Miner feasibility issues on trading platforms with no real deliveries’ or even better  ‘OMG, OMG margin calls make the shorts real uncomfortable’ or whatever you want to call it.

!!!This is not an ‘attack’!!!

[bytemaster]

Actually, I assumed that, by your own design, the BTS he received post-attack, while being at a lower market price, would be at a higher market quantity and still retain all of their value. The attack-profit comes from the initial raiding of the shorts, but also by selling BTS before anyone else can.

To trigger a margin call the value of BitUSD must rise rapidly against the collateral.  This requires people to buy BitUSD in a market where no one is selling BitUSD or SHORTING BitUSD.  The only reason for people not to short BitUSD into buying demand is if the real value of BTS were falling relative to USD.    Holding BitUSD is like lending the network $1 worth of value by not redeeming it immediately for shares.   The more demand for BitUSD with no sellers means the credit rating of the network is growing as people are willing to accept and hold IOUs from the network.

In the event of a network default, those that hold IOUs are effectively converted to equity positions.  A default can occur only when the real value of BTS falls very rapidly.   Manipulating the real value of BTS is like manipulating a penny stock (initially) but becomes increasingly difficult as the network grows.   

Unfortunately the network does not know the real BTS / USD price and is operating entirely on the BTS / BitUSD price.   So to crash BTS in the BTS/BitUSD price means you must buy up a lot of BitUSD in a market where there is no one willing to short BitUSD.   So the window of attack is in the arbitrage delay between BTS/USD price changes and BTS/BitUSD price changes.  Thus it requires speed to trigger a 'false margin call' and this is why having minimum market depth and maximum price movements helps give all market participants TIME to assess the reality of the value change in BTS/USD before a chain reaction can gain speed and force shorts out of their position at a loss based upon noise rather than fundamentals. 

Like we said earlier, if you could require 100x margin then you would be unable to shake someone from their position, but  you expose yourself to the monopoly attack when someone buys up all of the BitUSD to force the margin call.    Having 2x margin increases the risks of the shorts, but the good news is that 2x is just the minimum margin and any short that wishes to have a defense against certain market manipulation attacks would be proactive about maintaining higher collateral than everyone else.   In time the market will learn what to expect in terms of volatility.

[bytemaster]

Really? I can go long an ounce of gold, and short a barrel of oil. Each of those transactions could be described as "buying some amount of USD in the future", but I would NOT then describe them as "an investment in USD". You must be saying that the BTS system will be valuable, and BTS-owners will receive a sizable quantity of fees (paid in BTS, I assume). Even if this were true, it wouldn't be a reason NOT to sell those BTS off during a collapse in value.

The difference is that BitUSD is only redeemable in BTS where as gold/oil contracts are redeemable in gold/oil (in theory).   So at the end of the day BitUSD is just a claim on BTS.  You do not buy BitUSD unless you expect BTS to have value in the future with less volatility than the collateral can support. 

So think of BitUSD as an alternative investment that is highly correlated to USD but is not USD.

[Agent86]

I thought that, to go long BitUSD, you spent BTS to receive BitUSD. That's usually what people mean when they say they went long stocks (that they spent dollars on stocks). If the price of stocks in terms of usd has gone up, then the price of usd in terms of stock has gone down. So buying BitUSD is putting pressure on the price of BTS, not supporting it. Or have I misunderstood?

Yes, you have misunderstood.  There are two ways to buy bitUSD:

1st way
you can pay someone real dollars for their bitUSD
effect:
no direct effect on BTS price.
If bitUSD is sold on an exchange this will drive up bitUSD price relative to real USD

2nd way:
Buy BTS first and then use it to buy bitUSD within the BTS system.
Net effect: you had to buy BTS on an exchange and you drove up the real value of BTS.
The internal creation of bitUSD was not done on exchange; it happened within the BTS system and has no direct effect on an exchange with real dollars.

The secondary effect of people buying BitUSD the 1st way is that it makes bitUSD scarce so people try to get bitUSD the 2nd way; which again pushes up the value of BTS.

[MolonLabe]

you claim they can make money is by selling BTS off chain and buying BitUSD on chain... but these two actions are mutually exclusive as buying BitUSD is ultimately buying a variable amount of BTS.

Perhaps I did not explain...you either use BTS for one or the other. You go long whenever you can and sell on an external exchange otherwise (as long as you are long somewhere). This is less 'attack' and more 'defense', and does not need to be deliberate contrived by anyone.

So buying BitUSD is generating demand for BTS and supporting the price.   

I thought that, to go long BitUSD, you spent BTS to receive BitUSD. That's usually what people mean when they say they went long stocks (that they spent dollars on stocks). If the price of stocks in terms of usd has gone up, then the price of usd in terms of stock has gone down. So buying BitUSD is putting pressure on the price of BTS, not supporting it. Or have I misunderstood?

So if you want to execute this attack for the purpose of making money the only way to do so is to short BTS off chain which means borrowing from someone off chain.

You might do this, but you could also just sell the BTS that you currently own. If BTS/USD price is falling, many will decide to sell (surely you do not believe that the BTS/USD price will always go up?).

So if buying BitUSD is viewed as a proxy for buying some amount of BTS in the future where the amount depends upon the future market value of the dollar or 2x the BTS you could have purchased with USD today.    Thus BitUSD is merely a partial hedge against BTS value falling and is still an investment in BTS.

Really? I can go long an ounce of gold, and short a barrel of oil. Each of those transactions could be described as "buying some amount of USD in the future", but I would NOT then describe them as "an investment in USD". You must be saying that the BTS system will be valuable, and BTS-owners will receive a sizable quantity of fees (paid in BTS, I assume). Even if this were true, it wouldn't be a reason NOT to sell those BTS off during a collapse in value.

So if the system has 0 value then it does not work, but as long as it has some value it will work.  An attacker must buy before he can sell and must own a large amount.  Attacks are not likely to be profitable because at the end of the day he ends up with BitUSD converted to BTS which is worthless as a result of the attack than it was when he purchased it originally in order to execute the attack.

Actually, I assumed that, by your own design, the BTS he received post-attack, while being at a lower market price, would be at a higher market quantity and still retain all of their value. The attack-profit comes from the initial raiding of the shorts, but also by selling BTS before anyone else can.

So selling 'off-chain' is the only way to making money and in this sense it is like bitcoin, very hard to make money on the down side because few will lend BTC or BTS in the real world necessary to short it in the real world.

Why do you assume that people need to borrow BTS to sell? Even those who believe in the BTS philosophy prefer having more money to less money. They will sell if it benefits them, and one way to do that is to go long an asset (of course, my attack murders all the shorts, so the only other way is to sell on an external exchange).

So I still conclude that your attack is not 'automatic' nor 'assumed' and that it is not profitable.    It would have to be triggered by first buying into the system to prop up the price...

Not true. Can (and will) be triggered by existing True "BTS-Believers".

then buying a ton of BitUSD which will cause the price to rise

Each rational person will buy a little BitUSD, if only to protect themselves from this attack. This itself is enough to start the attack.

and cause new shorts to enter.

Anyone who shorts inside of BitsharesX is going to regret it. I have already explained why. Those who don't learn the easy way will eventually learn the hard way, or run out of money. The BitAssets allow people to flee BTS even more easily than they would otherwise.

The strategy of buying BitUSD until a short squeeze results in more and more collateral being put up for the later BitUSD purchases.   So some early positions may get squeezed but the later ones will not. 

In fact I think attempts at buying up BitUSD will only strengthen the network... so all that remains is attempting to manipulate the real world values. 

Sorry, this is too vague for me. "later ones", "strengthen", "real world values"?

[bytemaster]

Yes I was stating that volatility is always present not that its magnitude is constant.
Yes I expect a short squeeze to occur at some point in time.
Yes I assume that any market move is possible for natural causes. (Natural being defined as changing market perception rather than attack).

Given any market move is possible for natural causes, it is therefore to be expected that someday shorts will be blown out and BitUSD holders will have to eat any losses due to insufficient collateral.  BitUSD will then remain pegged to BTS until the holders of BitUSD accept their losses or the value of BTS recovers. 

So the question becomes one of unnatural causes that are not based upon economic fundamentals but based upon one individual or groups attempt to manipulate the price to profit.  Like you said people like to make money any way they can. 

So the way you claim they can make money is by selling BTS off chain and buying BitUSD on chain... but these two actions are mutually exclusive as buying BitUSD is ultimately buying a variable amount of BTS.   

So buying BitUSD is generating demand for BTS and supporting the price.   

So if you want to execute this attack for the purpose of making money the only way to do so is to short BTS off chain which means borrowing from someone off chain. 

So if buying BitUSD is viewed as a proxy for buying some amount of BTS in the future where the amount depends upon the future market value of the dollar or 2x the BTS you could have purchased with USD today.    Thus BitUSD is merely a partial hedge against BTS value falling and is still an investment in BTS.   

So if the system has 0 value then it does not work, but as long as it has some value it will work.  An attacker must buy before he can sell and must own a large amount.  Attacks are not likely to be profitable because at the end of the day he ends up with BitUSD converted to BTS which is worthless as a result of the attack than it was when he purchased it originally in order to execute the attack.   

So selling 'off-chain' is the only way to making money and in this sense it is like bitcoin, very hard to make money on the down side because few will lend BTC or BTS in the real world necessary to short it in the real world.

So I still conclude that your attack is not 'automatic' nor 'assumed' and that it is not profitable.    It would have to be triggered by first buying into the system to prop up the price... then buying a ton of BitUSD which will cause the price to rise and cause new shorts to enter.    The strategy of buying BitUSD until a short squeeze results in more and more collateral being put up for the later BitUSD purchases.   So some early positions may get squeezed but the later ones will not. 

In fact I think attempts at buying up BitUSD will only strengthen the network... so all that remains is attempting to manipulate the real world values. 

[MolonLabe]

   Your argument that you should be long BitUSD because you can get more BTS when margin is called because in your estimation the network is worthless.

Not true. I ONLY assume that people prefer having more money to having less money. I do, separately, believe that the network is worthless, but I plan to carefully build that argument up from a few pieces, including this one.

Volatility is constant and taking your bet is extremely risky because of the single unknown component: time.

I assume you mean that volatility is 'always present' (obviously it is highly variable). We agree that volatility exists, I think.

You are betting a run-away short squeeze is inevitable

Surely you agree that one margin-call will happen eventually? I do believe that I can prove that one margin leads inevitably to a run-away shorts squeeze, and (moreover) the protocol must be prepared to handle one even if you never expected one.

and when this happens the value of BitUSD becomes pegged to the value of the collateral.

At the instant of the margin call, yes. Which is why manipulated prices will stick.

The market participants recognizing that going long BitUSD is just like going long BTS and that they therefore have nothing to gain by attempting to 'join an attack'.

The other half of the equilibrium is selling BTS on external exchanges. This is where the attack continues (until all shorts are squeezed).

The fact remains that BitUSD would be serving its purpose most of the time with any deviations being short-lived.

Why wouldn't the attack happen immediately of its own volition, as a result of greed, and partially as a result of risk-aversion on the part of innocent traders? With some heavy-handed and expensive support from a dedicated team, the market may appear to be healthy (until this support is withdrawn).

So by your estimate no-one should go short which means no one will be able to go long.  So if a market forms at all then your assumption is wrong. 

I am describing equilibrium strategic behavior, if a market forms at all, it will mean that some participants are irrational (prefer having less money to more money). These individuals will lose their money, quickly if other participants are rational, slowly if few are. I am NOT assuming that everyone is rational, or that everyone is aware of my attack/the risks involved in going short.

[bytemaster]

MolonLabe,
   Your argument that you should be long BitUSD because you can get more BTS when margin is called because in your estimation the network is worthless.  I see your point and think you are missing something very critical:

Volatility is constant and taking your bet is extremely risky because of the single unknown component: time.   You are betting a run-away short squeeze is inevitable and when this happens the value of BitUSD becomes pegged to the value of the collateral.    The market participants recognizing that going long BitUSD is just like going long BTS and that they therefore have nothing to gain by attempting to 'join an attack'.   The fact remains that BitUSD would be serving its purpose most of the time with any deviations being short-lived.

So by your estimate no-one should go short which means no one will be able to go long.  So if a market forms at all then your assumption is wrong. 

All market participants will understand the risk involved and that they are subject to the risk of manipulation like everyone else.  These same market participants will still recognize the value of hedging.   

Fortunately this is an experiment, like bitcoin, that we can run and learn from.

[MolonLabe]

Does the fact that bts earns income from bitusd and bts transactions change the argument? It seems like if I had no guess about their relative directions I would go long bts over bitusd to capture income. I'd much rather x% of current tx fees as income than not.

Why not??   Because the market values bts at non 0 value.
Sent from my iPhone using Tapatalk

I don't think I'm getting my point across. Imagine that everyone is doing as I suggest in bold (we can go over "Why is everyone acting this way?" and "Is this behavior inevitable?" after we sort this first part out).

individuals should always be long BitAssets and short BTS, both within BitsharesX and on external exchanges, and always sell the BTS they receive as soon as possible

Then none of what you are saying matters, because if you deviate from this strategy (and hold BTS), very soon you will be margin called for your BTS, and lose them. Does this make sense?

[fuzzy]

It has been over a month, and no response to my critical manipulation. I myself have not thought of a way to prevent it.

Neither "limiting trading based on market depth" nor "maximum price movement per block" will prevent a sudden decrease in the market price of USD/BTS on external exchanges (which are absolutely required if one is using PoS). This price decrease will de-collateralize everything, resulting in cascading margin calls that cause BTS to fail to track every asset (as they will all be doubling in value toward the same infinity, as shorts sell to avoid Mcalls and longs sell to cash out their BTS [both at rates "faster" than the collapse in external USD/BTS thanks to the new 'maximum price movement' requirement (which is artificially slowing the internal USD/BTS)]).

Moreover, I feel that the fee structure...

Fees are calculated based upon how much of the order book you walk.

Given bids of 99, 98, 97, and 95.... to buy it all up in one go you enter an ask for 4@95.   You will receive 4*95=380 rather than 99+98+97 = 389 for a fee of '9'.    Obviously, if you attempted walk the book all the way down to 50, your fees are going to average 25%.

...encourages low volume (more trades = higher fees, as is obvious in the above example, 1 bid=0 fees), and large spreads (as you literally pay for every dollar the market-price moves past your existing bids/asks). As volume collapses, the market may lose all of its liquidity completely. Strategically, people will be hesitant to enter a soon-to-be-illiquid market.

All phenomena create an equilibrium of non-tracking and no trade.

( This thread promises tips in exchange for "finding attacks", but I have not yet received any tips. My Bitcoin address is 1DSrFGXJdsFw2MsrgwHeQxWq1djQk4jcyD )

If some wanted to tip u in pts, is this impossible or do u have an address?  This conversation seems worth it.

[bytemaster]

Why not??   Because the market values bts at non 0 value.


Sent from my iPhone using Tapatalk

[toast]

As in, I don't completely understand how the to 0 payoff difference in the two versiona of the game looks like in terms of actors. Like they think the income is already priced in in the +0 version?

Sent from my SCH-I535 using Tapatalk

[toast]

Does the fact that bts earns income from bitusd and bts transactions change the argument? It seems like if I had no guess about their relative directions I would go long bts over bitusd to capture income. I'd much rather x% of current tx fees as income than not.

Sent from my SCH-I535 using Tapatalk

[MolonLabe]

Ok I feel at least that you essentially understand what I am concerned about.

You are right when you say that it is all largely arbitrary, and traders will be able to influence the price through their shared opinion of what it should be.

However, although the price is largely arbitrary, it is not completely arbitrary, and this tiny constraint (the margin call) is what I want to emphasize.

To be clear, the new attack is:
1] Start the "Up and Away" attack (either directly, or through rumor/panic/misinformation) on a tradable BitAsset (such as BitBTC or BitUSD).
2] Sell BTS for the real-life version of that BitAsset.

Let me try a metaphor. Marbles are rolling down a halfpipe in a hill, randomly coasting to the left and right of the halfpipe nadir. If they somehow make it up and over the left side, they land in a new, lower, halfpipe (margin call), yet if they make it up the right side nothing happens (the other halfpipe is too high).

I would expect that, eventually, the marbles will land in the lower/lowest halfpipes. For this reason, individuals should always be long BitAssets and short BTS, both within BitsharesX and on external exchanges, and always sell the BTS they receive as soon as possible. The only logic available to traders points in the direction of this attack, everything else is just an arbitrary "I hope".

If they do this, they have a free option to double their money. If everyone does this, the first people who tried it will succeed the most, so people are rewarded for trying it immediately, and following the first movers. If no one did this, purely random market volatility would eventually cause a margin call somewhere, so users are still incentivized to "lean long", ie be more long than short (than they otherwise would be), foreseeing this, traders would themselves go long purely to capture this bias. Assuming reasonable time value of money and convenience yield, there is no profitable deviation from the boldface strategy (that I can see), and anyone who does not follow it would do better if he/she followed it, so I therefore assert that it is the unique Nash Equilibrium to this payoff scheme.

When you say that "traders will restore the price to the real price", aren't you invoking circular reasoning? BTS does not know what the "real price" is, it gets it from the traders. I realize that we can talk about it here but the software wouldn't understand us (the way it would understand a margin call). 2x, 3x, 10x, any amount of margin would not be enough.

It isn't just a bubble, where you hope to sell to a greater fool, the margin call guarantees that you can force other people to buy your stuff. The bubble doesn't become unsustainable and pop, it continues indefinitely. Why shouldn't BitBTC cost 40 trillion BTS? Why not 80 Trillion? Why not 9999 trillion or more?

[MolonLabe]

Hey these responses seem very interesting and to take the attack seriously but I'm on vacation so I'll get back to everyone on Thursday probably.

I think I did mean a positive up left cell, perhaps my previous formatting experience trained me to avoid the + 1 . Of course as you must know it really makes no difference. I can't remember what I was thinking but I do remember that the post-matrix post replaces that one.