Hacker News new | comments | show | ask | jobs | submit login
MakerDAO’s “DAI” stablecoin is breaking, as predicted (prestonbyrne.com)
109 points by matthewbauer 11 months ago | hide | past | web | favorite | 76 comments

The thing the author seems to be confused about is that when people talk about stable coin they aren't claiming that the market can't be irrational. They are claiming that over a sufficient period of time the coin will trend towards its target. In the case of DAI today, that target is 1 DAI == 1 USD. This means that while the market may irrationally buy/sell DAI for more or less than 1 USD, over time it will trend towards 1 USD.

For the vast majority of people looking for a stable coin, this is exactly what they want. They don't care so much that on a particular exchange DAI is worth more or less than $1, they care that in a week, or a month or a year DAI will still be worth "about a dollar" and not "100x USD" or "x/100 USD". DAI has the mechanisms necessary to make it profitable to drive the price of DAI back toward 1 USD if it loses its peg and those mechanisms have been shown to generally work in the wild so far.

I have personally made a few bucks when DAI transiently slipped its peg in the past. There are people actively authoring bots that profit off of DAI losing its peg and as those bots become more efficient and gain capital, I think we'll see DAI slip its peg less and less often.

> They are claiming that over a sufficient period of time the coin will trend towards its target

This is not what the MakerDAO team, nor any other stablecoin shill, is claiming. They are claiming the system "ensures" that the stablecoin is "a money that will always maintain its purchasing power."

"Ensure" and 'always' don't mean "trend towards over time." They mean always, as in all the time. If their intention is to sell a different product, I can only suggest they make more representations which are a little harder to falsify.

If what you want is "something that always is worth _exactly_ 1 USD then you are asking for the unatainable. As others have mentioned, it is _impossible_ in an open market to have a true 100% no variability peg. Markets work on supply and demand and the instant that one exceeds the other the peg will drift.

A good stable coin will have pressures that push that drift back toward the target, but it is impossible to stop the drift from ever occurring. The USD itself presumably targets some basket of goods, but it drifts away from that basket of goods. So even the USD isn't a "stable coin" by your definition.

I mean the notion of "exactly" is nonsensical in so many ways, especially when it comes to thinly traded exchanges. I can sell $1 to my friend for $1.10 break the US dollar. Lol.

But isn't 'hover around the same value long term' maintaing the purchasing power? I guess if you interpret the marketing as 'will literally always be worth $1.00' you'll be disappointed but it's weird to take that interpretation when it's not really attainable for anything that isn't a US dollar.

It's not even technically attainable for USD. If China places huge market orders for a basket of commodities with USD, USD would lose purchasing power wrt the basket for a short amount of time (I think a few seconds is possible).

The USD is not pegged to commodity prices.

You're right - what would you say it is pegged to?

Literally nothing.

Then what does the feds 2% interest rate target? Nothing?

The fear of Inflation. But USD is not pegged to anything.

Inflation measures USD purchasing power- how do you think that's measured?

So your claim is "USD is backed by X", it's just that I can't see the X, so tell me, what is USD backed by, without mounting an assault on the term "backed by". Backing means, some concrete promise or claim. USD has no concrete claim or promise behind it.

>stablecoin shill

You make some good points but I'm not sure labeling everyone and everything as "shilling" is helping your case at all.

But how much is DAI straying long term? USDT tethers are also often worth a few percent more or less than a dollar in periods of volatility. Is there hope that the 80% will zero in closer to a few percent in the future like USDT?

If there are thin enough markets of course "always maintains purchasing power" is impossible. This is true for USD ad well.

The difference being that USD doesn't pretend to be pegged to anything. My point about MakerDAO is that the development team is claiming the impossible is possible.

One of the many aims of US monetary policy is that USD doesn't lose too much purchasing power too suddenly (eg no unexpected hyperinflation). Of course any currency from a decent country would aim for this kind of soft peg.

I guess you're technically correct that their marketing material claims to go beyond this in a eay which is technically impossible

It may attempt to trend towards the dollar over time, but will it be trending that way during the time that you personally need it?

"The market can remain irrational longer than you can remain solvent", springs immediately to mind.

edit: Also, pegging a cryptocurrency to the dollar seems almost comedically perverse. What is it for?

> Also, pegging a cryptocurrency to the dollar seems almost comedically perverse. What is it for?

The frequently cited money transmission benefits of a cryptocurrency, without value instability of playing games with monetary policy.

If blockchain based currencies prove themselves to be a significant improvement on current monetary systems then state currencies, such as the US dollar, will adopt the new technology, as they have done with all other previous developments in money.

This would seem to indicate that a blockchain coin pegged to the dollar only keeps any marginal utility at all during the period where the technology is novel, as blockchain tech either crashing or becoming successful, would both be events that kill it.

> If blockchain based currencies prove themselves to be a significant improvement on current monetary systems then state currencies, such as the US dollar, will adopt the new technology, as they have done with all other previous developments in money.

The US dollar, as such, often hasn't adopted new developments in money transmission directly (at least, not at the consumer level); the circle of businesses facilitating USD transactions has, but a dollar-pegged cryptocurrency is essentially just a new part of that ecosystem.

OTOH, the dollar has adopted, directly, new tools in money supply management, several times in its history.

> The US dollar, as such, often hasn't adopted new developments in money transmission directly (at least, not at the consumer level)

The Federal Reserve began moving money electronically in 1915 [1]. (You read that right. Nineteen fifteen. In the early 80s, telex transmission services began.) The back offices of finance are surprisingly reactive to new technology, when the risks and rewards balance properly.

[1] https://en.m.wikipedia.org/wiki/Fedwire

As DAI gains trading volume and gets listed on more exchanges you'll likely see it stabilize. Any active trader who believes the price pressures will work has incentive to buy DAI when it is below 1 USD and sell DAI when it is over 1 USD as this is a low-risk trade (opportunity cost only). Right now, there isn't much volume so it is pretty easy for someone to come in and want a bunch of DAI or to liquidate a bunch of DAI and cause the market to shift suddenly. It will then take a few minutes for the market to correct, and with per-transaction fees the market is hesitant to correct small errors due to the lack of volume.

TL;DR: The more volume DAI has on public exchanges, the tighter it will hold its peg.

> Also, pegging a cryptocurrency to the dollar seems almost comedically perverse. What is it for?

The general thing you want is for your currency to have stable purchasing power (for goods and services) in the short term, maybe inflate it slowly to encourage investment - monetary policy basically. Targeting a USD peg is a way to do this, of course ideally you'd want your currency to have a monetary policy independent of USD

Won't someone learn to play the peg slip robots? If you can hack a PS4 you can trick some simple arbitrage robots.

Not saying it's easy, or I could do it... just that it's an inevitability, right? What happens then?

I mean presumably if you’re “tricking the arb bots” you’ve just found some additional arb of your own, which should make the market more efficient, and apparently in this case (I don’t know if this is correct, but this is the claim made by TFA), drive the price towards dollar parity because of this thing’s built in incentives. Or the arb isn’t from dollar parity mismatch but from the bots themselves, in which case it doesn’t matter to the market because it was dumb and you made it go away.

> if you’re “tricking the arb bots” you’ve just found some additional arb of your own, which should make the market more efficient

Do these assertions really hold?

1) That if you are able to make money on a market your presence makes that market more efficient

2) That tricking the arb bots into losing their money means you take it

? The Parity hack wasn’t a money-making endeavor, it just burned the cash.

In this thread cryptocurrency community will rediscover banking and money.

We have stablecoins or tethers (manny attempts to create something people want) - people want some form of money for peer to peer value transfer, that will allow arbitrage between exchanges and connection to "real world" as almost nobody is pricing their services in ethereum/bitcoin.

Let's say we have 3 competing stablecoins and Gresham's law will apply = people will be moving to "best" coin and leaving others holding bad ones. Imagine that you have 2 banks - one with good reputation and large reserves and second one - almost without them, yet both will be issuing private money redeemable for each other. So how come one can attract customers to his specific stablecoins solution? By offering interest rate - a stablecoin that pays dividend will be better than stablecoin without it.

It's complicated - ideologically stablecoins are similar to economical perpetuum mobile devices, yet the first are impossible due to laws of physics and second one may work because economy depends on human behaviour and humans may act irrationally.

A stablecoin isn't a store of value. It's an instrument for holding pretend Fiat. It's a way for holders to pretend they are the Fed, and say "Now I have USD! You can have it. Just sell it back to someone else eventually because it's pretend."

It's an index basically, not an actual commodity.

(Except of course in the sense that Everything Is A Commodity in the crypto world.)

Eh, I wouldn't quite call it an index. While I agree that it isn't _actually_ 1USD (or whatever it is pegged to), it is backed by other assets (in the case of DAI today, it is backed by ETH) and the system is setup such that there should always be more ETH backing the DAI than necessary to make everyone whole.

Actually, for such laughably small trading volume DAI is as stable as mountain:


I do expect DAI to be more stable as adoption and trading volume grows.

Author of is article is a troll. In his last post he wrote literally:

> Having taken fifteen minutes to review the MakerDAO paper

MakerDAO is probably one of the most ambitious projects on Ethereum with pretty complicated structure. Draw conclusions on 15 minutes of review is nonsense.

> Author of is article is a troll.

He may have an incendiary and flippant writing style, but it’s worth noting that he was a founder and COO of Monax (née Eris) which released an open source private-chain fork of Ethereum in 2014. He’s qualified to have an opinion on complex smart contracts and not be called a troll. If anything he’s emblematic of the frustration felt by many early adopters who understand that the hype bubble is making it hard for this rather experimental technology to meet inflated expectations (Vitalik recently being another prime example).


"Author of this article is a troll" != "Author of this article doesn't have a technical background"

I also have a strong background in the field and the MakerDAO project took me a long time to understand. The author makes several mistakes in his opening paragraph.

I'm not entirely convinced that the project will work as intended, but the author clearly doesn't even understand the basic ideas.

False. He actually has went and reworked troll posts he made in the past. Thats what this guy does. He has no intellectual integrity. His technical background (calling yourself COO doesn't really count as a technical background)

I'm a bit skeptical of derivative-based stablecoins in general, but Dai dropped briefly and then went right back to about a dollar, currently $1.02. I don't think they ever claimed that their price stabilization mechanisms would work instantaneously.

And while it dropped momentarily, it dropped to 70%-80% of the USD. In the world where most other cryptocurrencies move that much in a regular day without anyone breaking a sweat - that seems like a very good deal to me.

It is funny because this guy also fundamentally doesn't understand how Basecoin works.

"If the price of 1 BASE is above $1, the blockchain prints new $BASE to holders of “BASE Shares” (mother of god), a standalone cryptocurrency which is issued in the genesis block and held by early adopters, which then distributes new Basecoins to them as a “dividend.” Holders of “BASE Shares” are free to either hold onto their Dividend-Basecoins (thus not bringing the price down) or sell them into the market, pushing out the supply curve and bringing down 1 BASE’s price. This process continues until the price of 1 BASE drops to $1."[0]

Actually, the Basecoin will first go to the Bond holders. I think there are reasons to be skeptical of Basecoin and stable cryptocurrencies, but this guy's "articles" don't really seem to touch on those.


Well, of course it's easy to ensure that the price doesn't stay above $1. You can make more of them. It's ensuring that the price doesn't stay below $1 that's hard. You have to draw value from some reserve. If that reserve is depleted, you're screwed. All attempts to peg currencies run into that problem.

Would it not be possible to just destroy some of the coins? If they were destroyed for everyone, proportionally to how much they hold, wouldn't that be in interest of everyone? Because it would make the coin itself worth more?

As I see it, you have two options.

The first is distributed destruction. Everyone destroys some of their coins to raise the value of remaining coins. The problem is that this is a prisoner's dilemma: defection is the best strategy. Some people will not destroy their coins, thus increasing their relative position to others who do. Over time "destroy your coin" signals will simply be ignored.

Not to mention cold wallets, which simply don't participate in the ledger unless it suits them to. Are they meant to simply nuke coins upon rejoining? This introduces an incentive to stockpile coins in the hope that the policy will change in future.

The second approach is centralised. Someone has the authority to mark some fraction of coins as destroyed. Rather undermines the decentralisation thing people like.

But the destruction would of course (obviously) be implemented in the protocol.

Just like in Bitcoin, it's not up to each user to be honest and prevent double-spends, it's instead up to the network that checks and re-checks everything - the destruction of those coins in the example would be enforced by the network. Not relied on users.

This is why I brought up the cold wallet.

The network can check that transactions occurred, but it can't really make people perform them. It's a ledger system.

If someone doesn't post a coin destruction event, how do I know why? Is it because they're cheating? Because of a cold wallet?

You might say "I will make coin destruction a precondition for any other transaction". Now the incentive of a hoarder is to sit on their wallet while destruction is high and only enter the market when destruction is low. This means liquidity is tied to how far off the peg you are, which isn't a desirable property.

OK, how about summing up the destruction tally for any wallet that's inactive, to be paid in a lump sum? Now you're telling me that my wallet can be completely drained over time by a destruction tax that realises only if I open my wallet. Not very attractive.

OK, maybe you just destroy the net of coin creation and coin destruction at transaction time. This seems more likely to work, though I'm unsure whether I can evade it by changing addresses. I'm also unsure what incentives miners have to validate coin destruction. There's also the problem of how it cope with truly heavy amounts of capital flow. Watching a numerical balance surge up and down during a single day might be unattractive to a lot of holders.

"Distributed destruction" is just inflation in another form.

That would remove supply, but how would it stop dwindling demand?

If I have two moldy banana peels worth approx $0 each and I burn one of them, the other will presumably still be worth $0.

The whole idea is that demand will not be 0. If the demand will be non-existent, then none of this is even a problem. If no one wants that coin, it doesn't matter how it works or how many coins are destroyed, etc.

But this whole conversation even comes up because there are people that want to use such a coin. That means demand >0.

>"If the price of 1 BASE is above $1, the blockchain prints new $BASE to holders of “BASE Shares” (mother of god), a standalone cryptocurrency which is issued in the genesis block and held by early adopters, which then distributes new Basecoins to them as a “dividend.”

Isn't that a freakin' textbook definition of a Ponzi scheme?

> the Basecoin will first go to the Bond holders

Only if the price is below $1. You have to read the next paragraph of the blog post to get to that.

In the past, this person trolled Bitshares to make a name for themselves. They selectively cut'n'pasted snippets of the Bitshares forum as some form of evidence in other posts. There is a ton of context there that was removed. A blockchain can give an asset value, but that doesn't mean every market will guarantee those prices. He is now going after the Maker coin. The guy does well SEOing his website with this rubbish but he doesn't seem to understand much outside of that. oh well, no big deal until google/ycombinator point to him as an authority.

It seems to me that any decentralized stablecoin would break the impossible trinity. [0]

Basically, since the interest rate is fixed (at 0 in this case, I believe), there's an incentive to sell DAI in favour of some higher-yielding currency, which creates a downward pressure on the price. So short of preventing DAI from being sold somehow, in order to maintain a peg, someone needs to pump money into DAI at a loss.

You can try to build a complicated system to hide all that, but it seems the goal of a stable peg is simply impossible. Even for nation states it is only possible in the case they have sufficient resources and the will to maintain a peg.


Why do you say interest rate is fixed at 0. Isn't the whole bond mechanism supposed to introduce an adaptive interest rate?

Author has misunderstandings of what he's criticizing. From the article I suspect he thought the crash was caused by undercollateralization at first before his friend corrected him. His 15 minute attempt to understand the system didn't include enough time to check the collateralization ratio limits.

Also claims that a collateralized stablecoins "only works if the collateral value keeps rising", and that selling something is equivalent to using it as collateral

How is this mud slinging article so high on hacker news? This guy is just like every other troll who attempts to forge a name for themselves by tearing down other people's hard work.

Yeah, I hate security researchers too. They should shut up about TLS and processor vulnerabilities.

This does read more like a blogger trying to self-promote than anything else.

The author's main point is that since the token is collaterized by ether, the value of the coin depends on ETH rising in value. The token's creators attempt to solve this by over collateralizing 50% which he considers foolish because it just adds leverage to the system so when losses occur, they are exaggerated.

Sure, but they are planning on adding other assets as collateral. ETH is just the first version, the proof of concept. I think the author knows this, though, and chooses to dismiss the project anyway.

FUD. Insufficient liquidity on some exchange might cause the price to fluctuate briefly, of course. But it quickly corrected back to $1.

And unless I'm mistaken, the people at Maker never said it was hard-pegged to the dollar. It was a soft peg, so the price could fluctuate here and there, but remain roughly stable.

If a group wants to peg a coin to the dollar, why don't they just issue coins for $1, put that directly into a savings account, and allow redemption of any coin for $1?

That's called Tether, and its primary problem is that everyone is taking the issuers at their word that the money is going into a savings account, and that they are only issuing it when they receive $1.

What's the incentive for you to make your token stable at $1? You can print the tokens, so it's worthless to you. Why not take the $1, spend them on, say, a house, but not tell anyone?

This is the situation with Tethers right now.

cryptocurrency-people have trouble trusting. this same line of questioning leads to "why even use a blockchain" when you could just have a central database track transactions much more efficiently.

You have to trust the group that they won't for eg suddenly take the money in the savings account and run off (not honor redemption) some point in the future

Would US corporate structure offer some security for these funds? So it becomes a case of US embezzlement vs trying to make yourself whole via the blockchain?

Conceivably, but I'm not sure what regulations such a company would have to comply with or even if they can incorporate in the US at all. Eg would they need to be responsible for not letting North Koreans redeem? (probably.) What else?

I do remember tethers parent company basically couldn't get any US bank accounts

introduction bashes ethereum over its price jump and seems to go for tulips explanation. that's where I stopped reading

Another misinformed article from Preston Byrne. It's unbelievable how out of touch this guy is!

Tether gets criticized for being fractional, but it's more stable than fancy collateralized derivative products. Cryptocurrency never ceases to amaze me.

Tether gets criticized for having no released third party audits that they actually hold the hundreds of millions of dollars that is backing their token. They say they have frequent audits on their website, their transparency update that they called the Friedman audit, in the same paragraph states that it does not constitute as an audit[0].

In the same “audit”, Friedman stated they did not evaluate if tether owned those accounts, had access to the funds, if those funds represented anything else besides the backing of tether[1].



It doesn't even matter if they don't hold the USD to back tether. They make it clear in their ToS that users cannot redeem their USDT for USD. There is no known case of anybody being able to cash out USDT directly from tether.to. It is only possible by exchanging on third party sites or with other users.

Tether is a money-printing scam by a company who had their bank account frozen, so instead of handling USD they came up with the wonderful scheme of USDT.


would have included that, but tether changed that part of their legal section. it now includes more legalese that i'm interpreting as can redeem, but US residents cannot be customers. no idea of the implications of the latter for people trading it on exchanges.

"Absent a reasonable legal justification not to redeem Tether Tokens, and provided that you are a fully verified customer of Tether, your Tether Tokens are freely redeemable.

Persons ordinarily resident in, and nationals of, Prohibited Jurisdictions or Sanctioned Persons under Applicable Sanctions Laws; Persons and Government Officials believed or suspected to be transacting in the proceeds of corruption, bribery, or other crimes under Applicable ABAC Laws; and Persons believed or suspected to be engaged in money laundering or terrorist financing under Applicable AML/CTF Laws are not permitted to be customers of Tether; are not permitted to cause Tethers to be issued or redeemed; and, are not permitted to hold or transact in Tether Tokens.

Furthermore, residents of certain U.S. states are not permitted to be customers of Tether; are not permitted to cause Tethers to be issued or redeemed; and, are not permitted to hold Tether Tokens.

Beginning on January 1, 2018, Tether Tokens will no longer be issued to U.S. Persons."

Decentralized trustless systems are always going to be less efficient than centralized ones.

Wait, tether is fractional now? Wasn't one of their big promises that Tether is always backed up 1:1 by the dollar?

I believe you are correct that it's not fractional, their website claims "100% Backed - Every tether is always backed 1-to-1, by traditional currency held in our reserves. So 1 USD₮ is always equivalent to 1 USD." [1]

The whitepaper abstract also states "issued tokens are fully backed and reserved at all times." [2]

[1] https://tether.to/

[2] https://tether.to/wp-content/uploads/2016/06/TetherWhitePape...

Tether claims to be full reserve. The contention people have is that there is little to know proof that they are _actually_ full reserve. Also, there is risk in Tether that some government will seize all of their assets (there is lots of precedence for this).

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact