Two thought experiments to evaluate automated stablecoins | Hacker News Reader

Two thought experiments to evaluate automated stablecoins

vitalik.ca

210 pointsbpierre250 comments

Here's a different thought experiment.

A sponsor organization maintains a stable coin pegged 1:1 to the US dollar. The token is bought and sold on the open market to keep the peg.

Because it works so well, demand for the stable coin rises for years on end, resulting in the issuance of trillions of dollars worth of stable coin. Eventually the stable coin market cap hits 50% of all US dollars in existence. Numerous audits prove that the dollars are indeed responsibly held by the sponsor.

Now, what specifically are the assets being held by the sponsor?

This might seem like a trick question, but it's not. In what form would a stable coin of that magnitude keep its assets to ensure sufficient liquidity that the peg is never broken and can't be attacked successfully?

This depends very much on the definition of "all US dollars in existence".

If I put a dollar in my savings account and the bank loans it out to someone else, who has the dollar? It's on my balance, but it's in their pocket. How many dollars are there?

This is the question of money supply, it's very complicated but the TLDR is there is no fixed "all dollar" amount. This scenario just further muddies the question, but also explains why stable coins keep central bankers awake at night.

> If I put a dollar in my savings account and the bank loans it out to someone else, who has the dollar?

Note that for every dollar you put into a U.S. bank account that's subject to reserve requirements (I believe just checking accounts?), the bank can loan out much more than 1 dollar, based on the legally required liquidity ratio or "reserve requirement ratio."

Note also that since 2020 that reserve requirement ratio is zero.

That is a common misconception that got started as a misunderstanding of Milton Friedman's (reasonable) money multiplier equation. Banks cannot lend more money than they have - if they could, then you could start a bank and multiply your own money. The money multiplication that banks can do is take the money that party B has received for selling something and loan it back to to party A, who can use it to buy something from party B again.

A zero percent reserve requirement does not lead in all cases to infinite inflation, because banks are also afraid of a default. Bank's desire to avoid defaults adds another fraction to their reserve ratio and limits money multiplication even when the Fed would technically allow it to proceed infinitely.

Another thing you should know is that Fed chairmen have said that Friedman's equation underestimates bank's ability to create money, because in practice the Fed will allow any bank that desires liquidity to borrow it. The Fed, especially post-great-depression, will happily loan banks however much they needed to cover their reserve requirements, meaning that even before the ratio was set at zero, the number of times a single dollar could be spent was decided primarily by a balance of interest rates and fear of defaults.

Is there a misconception in the text of my previous comment? You seem to be describing a misconception that as far as I can tell isn't one I hold or one I expressed in my previous comment.

> loan out much more than 1 dollar

They can't loan out more than the deposits they have.

Why not? A bank could have 0 depsitors, but could have received money from investors.

That's not the sort of bank being talked about here in terms of "money multiplication." That's more like a VC firm or other institution.

If you could open a bank, accept 10 dollars in deposits, and loan out 100, you'd do it! You'd loan it out to your best friend and keep the cycle going until you're rich AF.

You can't loan out more than you have, but even loaning out 10 of 10 dollars has some risk - if someone comes for their money and you can't get your money back from the people you loaned it to. And it definitely increases money in circulation since it's not parked there. But it's not a magic money printing machine.

The Federal Reserve says you may loan out 100% of your deposits. They didn't say you could loan out however much you have in investors' money, or 200% of your deposits.

The misconception in your comment is that the reserve requirement is what limits a bank's ability to create money by extending loans.

I'm no expert, and I know that banking is a lot more complicated in practice, and that reserve requirements in the U.S. (even before they went to zero) were apparently not a significant factor in what actually put a ceiling on banks' ability to lend. There were apparently all sorts of tricks banks used to not be bound in practice to reserve requirements.

Yes, reserve requirements are not a significant factor, because in addition to reserve requirements banks are also subject to capital requirements. Moreover banks need reserves to operate anyway regardless of whether they're mandated by law or not. Finally they're also constrained by the demand for loans, which isn't infinite.

it's the capital requirements or capital to risk weighted assets ratio that restrict lending, not reserve requirements. reserve requirements are to protect runs by depositors and capital requirements are to protect the bank against bad loans.

at the end of every month and especially every quarter the bank submits its capital ratio and a bunch of other calculations to the regulators. if the ratio isn't good it has to increase capital reserves. if the ratio is good it can do share buybacks and dividends.

https://en.wikipedia.org/wiki/Basel_III

Can you share some resources where I can learn about this? Thanks

The reserve requirement ratio does not do what you think it does. It is the amount of money that the bank is required to keep in its account at a Federal Reserve Bank as a ratio of deposits. Note that literal cash sitting in a vault does not count one iota towards satisfying this reserve ratio.

Bank regulations have moved one from having such a specific requirement. Nowadays, you essentially need a minimum amount of equity per risk-weighted assets. So you need something like $8 in cash for every $100 in loans you give out.

> but also explains why stable coins keep central bankers awake at night.

Can you elaborate on this?

The sounds of the crashes are very loud and occur all hours.

There are far more "fake dollars" in the world than just that, even.

Ask someone with non-cash investments their net worth. They'll give you a number that includes a lot of non-cash assets. They can borrow against them. They expect to be able to cash many of them out, so they often spend as if they're "real" dollars. But they aren't "real" printed dollars.

> If I put a dollar in my savings account and the bank loans it out to someone else, who has the dollar?

I don't think that's how banks work; at least in the US and England.

Below is an extract from an excellent article/tutorial from The Bank of England[1]

In the modern economy, most money takes the form of bank deposits. But how those bank deposits are created is often misunderstood: the principal way is through commercial banks making loans. Whenever a bank makes a loan, it simultaneously creates a matching deposit in the borrower’s bank account, thereby creating new money.

The reality of how money is created today differs from the description found in some economics textbooks:

1. Rather than banks receiving deposits when households save and then lending them out, bank lending creates deposits.

2. In normal times, the central bank does not fix the amount of money in circulation, nor is central bank money ‘multiplied up’ into more loans and deposits.

[1] https://www.bankofengland.co.uk/quarterly-bulletin/2014/q1/m...

My guess is low risk debt. Most likely treasury bonds would be the majority of their holdings.

Is there enough supply of treasury bonds and other low risk debt to do this?

There are $23T treasury bonds outstanding, which is more than enough to back the $6T "M0" US money supply.

(This is kind of an exceptional circumstance, though)

It feels to me like a logical conclusion would be to have a government manage such a stablecoin. But at that point, it's just normal money but less convenient.

well normal money, but in a revolutionary form factor. metal coin usd -> paper usd -> crypto usd

Smaller jump then gold -> usd -> btc which actually means changing the global reserve currency.

If you think about it didn't the U.S. government did a similar thing by gold. USD was pegged to gold initially, then they changed it.

I think you missed a step:

metal coin usd -> paper usd -> digital usd (closed, permissioned system) -> crypto usd (transparant, permissionless system)

The jump is a lot less between digital and crypto should be small.

true, revolutionary (compared to even digital usd) as in It shouldn't require a third party (aka bank etc) for 2 individuals to transact.

No bank involved, but for every transaction you burn down an unspecified amount of the Amazon rainforest. But hey, you win some, you lose some.

You always need a 3rd party for two individuals to transact. It's banks for dollars, miners + the whole infrastructure for crypto.

In principle, miners or the infrastructure maintainers could be forced to never mine your transactions (well, transactions to/from particular wallet IDs), by the way. It would be of course harder, since your money is in a single bank, but all miners would have to agree to the ban for it to work.

>It feels to me like a logical conclusion would be to have a government manage such a stablecoin.

I think you give Governments far to much credit, they can’t even manage to peg the penny to 1cent or a nickel to 5cents.

In 2020 a penny cost 1.76cents and a nickel cost 7.42cents.

You’re confusing seniorage with value..

Seniorage is the profit between cost of producing currency and the value of the currency.

Technically there is no seniorage when there is a loss, historically it was called an inflationary tax, meaning a loss in value to the existing money supply you hold.

I think you're mistaken -- seigniorage is the difference between the cost and the value, it doesn't need to be a profit. Negative seigniorage is still seigniorage and many such papers treat it as such, e.g.;

https://sci-hubtw.hkvisa.net/10.2307/3552184

>Negative seigniorage is still seigniorage and many such papers treat it as such

Notice how you qualify your statement as many such papers treat it as such.

Yes, you can find some people, articles that refer to a loss as seniorage, but that’s a misuse and inconsistent with the actual definition. That’s also why when they do use the term improperly they typically add “negative”, specifically because the definition of Seniorage refers specifically to revenue/profit from printing money.

I think Ambani (the Indian billionaire) actually made some of his money in early fays melting down coins, and did it at such a scale government officials had to investigate what was going on

In the US it’s not illegal to melt pennys, but it’s illegal to melt pennys to profit on the raw materials.

Interesting side story: for centuries coins were intentionally made to have a ribbed side surface because otherwise people would sand/cut off the sides and sell the obtained metal.

Nowadays coins are usually not made of pure and/or very valuable metals, but the notches have stayed to help impaired people and automatic identification of coins.

Why do you think this would be best facilitated by a government as opposed to a large group of stakeholders?

Because if it's that central to the economy, representing half of all outstanding dollars, it needs to be managed centrally or bad things are gonna happen?

I mean who watch the watcher, we have democracy for the government right. Even that's an oversimplification, we have the 3 branches and check and balance so that everyone watches everyone else.

>Even that's an oversimplification, we have the 3 branches and check and balance so that everyone watches everyone else.

The Federal Reserve, which manages the current currency, was explicitly made independent of such checks and balances.

Because it’s easy to credibly promise someone their claim is worth X dollars when you (and you alone) control the creation of dollars.

I've been annoying people lately by referring to Black Wednesday as "the collapse of the stablepound against the algorithmic stablecoin the ERM". https://en.wikipedia.org/wiki/Black_Wednesday

I have a simpler thought experiment. Every time you own a tether, you buy it for a dollar. That dollar goes into a large room. Now each tether is backed by a literal dollar. What’s the problem? I can’t wrap my head around what’s so hard about this.

What happens when a bunch of literal or metaphorical rats eat the dollars?

Who pays for security for the room with the dollars?

Okay buy US government bonds with them. What else am I not seeing?

Now you have a drawdown and bond prices have declined because interest rates increased. Whoopsie, you’re fucked.

I was referring to fixed rate bonds. Alternatively charge 1% on each transaction for security.

You don't need crypto to know how this experiment ends.

It was called the Argentinian Peso, and it was pegged to the dollar. It didn't get to 50% of the existing dollars, it crashed and burned way before that

I guess they need to be an actual US based bank themselves to hold that much cash/bank deposits. And to also be trustworthy.

Your question is a trick question because the first few assumptions beg all the detail that would be required to answer the question.

But fundamentally if you were to reach that size the assets you were using to collateralize the coin would in some sense have to have some leverage against the coin itself. Ie be loans backed partially by the coin or something. But because you can’t pull yourself up by your own bootstraps that would never work, and because 50% of all the US dollars in existence is a mindbogglingly large number no player in any market could ever be that large. In particular, regulation actually prevents you from being that large in lots of important ways. For example if you try to participate in equity markets more than a certain percentage you will have your exchange access shut off and you will be investigated for trying to run an illegal market corner. Even if you set regulation aside, as you try to buy 50% say of a government bond issue to have those assets to pledge you will find the price moves further and further against you. This is the supply/demand mechanism forcing you to give more than 50% of the dollars away to other market participants meaning you can never get to the goal that you mention in your hypothesis.

Quite separately your thought experiment doesn’t apply to the situation Vitalik is talking about- he specifically says in the article he is talking about pure algorithmic stablecoins rather than ones backed by assets held by custodians of the type you mention here.

> In what form would a stable coin of that magnitude keep its assets

US Treasury bills?

Can they put it in a bank? Do banks have maximum account sizes?

In practice, yes. Because you're just pushing the backing question on to them; banks rely on withdrawals being de-correlated. Not everybody takes their money out at once. A single bank-sized account ruins that model.

Also, the larger it gets, the bigger the money laundering question gets. "We're providing seventy billion dollars of financial services to unknown parties" is not a sustainable position with a reputable bank. Tether have already got banned from several banks and are being very cagey about who they bank with.

> Eventually the stable coin market cap hits 50% of all US dollars in existence. Numerous audits prove that the dollars are indeed responsibly held by the sponsor.

> Now, what specifically are the assets being held by the sponsor?

By definition, they hold every dollar in existence. Holding any other type of good would wipe them out as soon as the exchange rate to the dollar changed.

> RAI's security depends on an asset external to the RAI system (ETH), so RAI has a much easier time safely winding down

Yay, we've "discovered" currency boards [1].

(The thoughts on negative rates are genuinely interesting, given their relation to present thinking on the subject.)

[1] https://en.wikipedia.org/wiki/Currency_board

[2] https://www.researchgate.net/publication/282613501_History_o...

There's nothing wrong with doing an old thing in a new, more automated and efficient way. In this case, taking something that's normally done by national banks, and doing it instead on a peer-to-peer network, implemented by a handful of developers.

> taking something that's normally done by national banks, and doing it instead on a peer-to-peer network, implemented by a handful of developers

It's not.

A currency board tied to the U.S. dollar holds U.S. dollars. A board--or ETF--pegged to the dollar holding rubles (or pegged to gold holding dollars or vice versa) is stable under a specific set of conditions. Outside those conditions, in the real world, it's trivially defeated.

I like the write up, I think it articulates some principles well.

I saw all the shitting on "algorithmic stablecoins" and found it a bit absurd. Not all algorithms are the same, this is plain as day. Not all game theoretical systems are identical. One very badly designed system fails, predictably, and all of a sudden every system that (conveniently) doesn't include a custodian is snake oil.

I'm finding it difficult to engage with the "Let's pretend USD is a ponzi scheme" pre-condition of some of this.

I also think that the premise that RAI is distinct from ETH is... tenuous. The problem that UST-TERRA had was that it was trivially the same, but that means what we're saying is that if RAI succeeds to any significant extent then it puts ETH in a situation where it may also death spiral. I feel like I'm agreeing with Vitalik there.

>Another extreme case worth examining is where RAI becomes the primary appliation on Ethereum. In this case, a reduction in expected future demand for RAI would crater the price of ETH. In the extreme case, a cascade of liquidations is possible, leading to a messy collapse of the system. But RAI is far more robust against this possibility than a Terra-style system.

I think the conclusion of the 2nd experimnet is "Don't buy RAI". It's just that if the return on RAI is positive, well, buy it but know it's going to explode. If it's negative well... You're earning a negative return well done.

Unlike UST/LUNA, there's no scenario in which more ETH will be minted to support the price of RAI.

Correct, but the mechanism of safe liquidation could look kind of like minting in the right context. Suppose RAI became dominant. Then the old safes of ETH locked away could become really large. In theory, they could even be larger than the ETH float. Increasing ETH prices would fuel lots of RAI withdrawals. But then if the price of the limited amount of floating ETH dropped, suddenly these safes would be forced to liquidate. This liquidation would release a lot more ETH, further depressing the price and driving increasingly larger liquidations.

Of course this is pretty unlikely, because RAI is peanuts compared to ETH.

Thought experiment #3

Create a hypothetical new stable coin. Issue one coin for every dollar put into the stable coin. For every dollar subtracted pay out the dollar and take the coin out of circulation until another USD comes back in. Make money on exchange fees ONLY. No fair using the coins or dollars in any other way.

The schemes for stable coins ALL have failure mods until someone does #3. The most likely entity to do #3 is the US treasury (or other national entity). But if someone is willing to live with income from fees only, they could do #3. Crypto currencies have efficiencies enough to make this a viable option.

How is this different from a bank account?

The bank doesn’t actually store your dollars and your bank account can’t make transactions on the blockchain.

And where would the issuer of such a hypothetical stablecoin store actual dollars, if not in bank accounts or government bonds?

I’ve been hoping we get #3 but with commodities and precious metals.

I want a “stable token” that represents barrels of oil, wheat, etc or physical gold, silver, etc.

The argument against the gold standard is that there isn’t “enough” to represent money — but I think we’d gain a lot of stability if prices were denominated in a basket of commodities.

Not really sure this would work. The problem with physical commodities is that you actually have to store them somewhere.

Having to store digital crap securely turns out to be a far bigger problem than most people realized.

Maybe still less than physical commodities, but let's not pretend crypto is "free" and easy to store and takes up no space.

The cost of storing a digital asset be it USD, a futures contract, or crypto is significantly less than paying to store physical goods. I don't see how this is up for debate.

Not to mention that there is already a small industry which does this electronically (LME warehouses) and in a way which is trusted by the largest traders in the business. Why wouldn't you just use that instead?

There are already ETFs which hold precious metals and other commodities. But most commodities aren't really "stable" stores of value themselves. Large quantities of wheat or oil can only be stored for a few years at most before they start to rot or decay. Storage is also quite expensive, especially when you run out of big oil tanks and have to charter tankers to hold the overflow.

They have XAUT and PAXG as ERC20 tokenized gold. Article (sorry if shady):

https://www.publish0x.com/journey-to-the-cryptocurrency-ocea...

>> The argument against the gold standard is that there isn’t “enough” to represent money

I hadn't heard anyone seriously say that. What amount of additional gold should be mined to fix this alleged problem?

Since there are 84M Litecoin, is LTC 4x better than BTC with 21M?

https://www.investopedia.com/articles/investing/040515/what-...

> The argument against the gold standard is that there isn’t “enough” to represent money

I literally have never heard this argument against the gold standard.

What amount of gold would be needed for it to be able to sufficiently "represent money"?

And why do you need gold to "represent money" anyway? Isn't gold itself money under the gold standard?

During the 19th century, the gold standard led to crazy inflation as new gold was found, followed by crazy deflation as the gold strikes got used up. Farmers were struggling because the money they borrowed to be able to grow crops would drop so much in value by harvest time.

This is what William Jennings Bryan's "Cross of Gold" speech was about: he was pushing for the free coinage of silver at a ratio of silver to gold of 16 to 1. This would have increased the money supply (silver as well as gold would be money) and replaced a deflationary environment to an inflationary one. Now, you can argue that this proposal was a gimmick and argue about whether it would have worked. But the point remains that the gold standard as implemented in the 19th century was a disaster for debtors and farmers had to borrow every year, unless new gold was discovered somewhere recently.

yes and a lot of these coins claim they are all backed but it's just a lip service, it's too lucrative to not do full 1-1 backing, you are right only something like US treasury can do it.

https://markets.businessinsider.com/news/currencies/tether-c...

Sure, direct 1:1 collateralized stable coins work, but look at the "algorithm" behind the underlying asset.

USD itself is an interesting stablecoin, algorithmic in nature with a board able to make decisions to change the algorithm. It is not pegged to an asset, rather it attempts to be pegged to an economic state, primarily an inflation rate, using issuance and purchase of other assets. It has failure modes as well.

Note that algorithmic and reserve based crypto stablecoins are both exposed to this, since one is backed and the other is pegged to it.

So if you want to avoid that, you need to peg your cryptocoin to something without that, i.e not a fiat currency. This is hard to do with collateral, some gold backed coins try. Pegging a stablecoin to some commodity or asset or index or "basket" algorithmically is much easier, if somewhat less stable depending on the system that is built to do it.

>> So if you want to avoid that, you need to peg your cryptocoin to something without that, i.e not a fiat currency.

Except is that really the issue? We're not trying to worried that dollar slides and hence the stablecoin is worth less. We're mostly worried that there arent dollars backing the stablecoin to begin with.

Mainly with stablecoins we are worried that there's no 1:1 correlation between the two assets. Whether they're backed or not is an implementation approach.

People are worried that the dollar slides, which was a big motivator for bitcoin in the first place. But my point is simply that the dollar is an algorithmic asset who's algorithm is governed by a governance body and has a targeted value based on economic factors, and if you don't want that, you should peg to an asset that does not have those properties.

If you want to do backing with dollars it's easy, just spin up a corporation, keep dollars on a balance sheet and you're done. Doing it with other assets if you want requires vaults and things, it's much easier to do it algorithmically, and the only reason reserves are easier with dollars than with other assets is that other assets actually exist, dollars are just a ledger in a computer, again, controlled algorithmically and governed by a board.

You’re arguing past the other poster. The dollar sliding shouldn't affect stable coins, and is unconnected to your inflation hedge in Bitcoin.

Why don’t any entities exist that do this “easy” method? Like you said, it should be straightforward. Well, because of their #3, stables want to make more money from their cash and aren’t satisfied with just tx fees.

This already exists, Circle Finance + Coinbase issue USDC, for example. It's just backed by 1:1 actual USD (at least it will be 1:1 very soon, as in 100% reserve, according to public statements, I haven't checked recently if that is complete).

They definitely do not back with 100% paper. They hold bonds and potentially corporate paper as well.

Is there a central location where the public can see the public statements frequently updated? I'm very skeptical about a public filing from three months ago when collateral was worth 2x what it is today...

https://www.centre.io/usdc-transparency

The extent of the detail that is given in the latest report is that their backing assets are "limited to cash and short-dated U.S. government obligations".

Yeah, but now people have to actually trust that you are being truthful about your backing.

Isn't this what Tether purports to do?

Tether is the next Big Short of the crypto community. I’m yet to find someone who doesn’t think tether will blow up. It’s going to be an interesting week when that happens.

I mean if you are so sure about that the trade is pretty simple. You can trade USDT perpetual futures currently at 0.9989.

You are probably going to respond with the "markets can stay irrational longer than you can stay solvent" meme but I don't see any downside to this trade other than the opportunity cost of investing your dollars somewhere else. There is no scenario where USDT goes to $10 and you lose your money.

> I don't see any downside to this trade other than the opportunity cost of investing your dollars somewhere else.

Shorting comes with interest fees for the borrowed asset. If Tether holds off a collapse for a few years, that can get substantial.

I am talking about futures not shorting, no borrowing fee if you don’t do it on margin.

Then you're talking about a setup where a) the counterparty can fail a margin call and b) the exchange it's happening on can get hacked or go bankrupt.

No fucking thank you.

1. I'm not sure what you mean by this. The closest thing to a margin call is an auto-deleveraging. Markets can get messy during periods of mass liquidations but the risk is knowable, and it's in the size of the insurance fund.

2. Counterparty risk exists in every traded market, not just crypto.

It's clear that you and most of the critics here don't even know the basics of market structure, which should cast doubt on your claims. I don't know any other field in which you can provably not know anything about a topic and yet be taken so seriously than in crypto, with both bears _and_ bulls.

> Markets can get messy during periods of mass liquidations but the risk is knowable, and it's in the size of the insurance fund.

The risk of a Tether collapse is knowable?

> Counterparty risk exists in every traded market, not just crypto.

Sure, and I’d avoid any of them whose daily operations are based heavily on a known fraudulent actor, too.

Thanks for this comment.

Sorry for my naive response.

Which product exactly are you speaking about? Could you link to the product and/or exchange where "USDT perpetual futures" are traded? I can see BTCUSDT futures (but then, you're taking another risk on BTC itsself sinking), is there a way to short-USDT-long-fiatUSD?

This one on FTX for example https://ftx.com/trade/USDT-PERP

1) Have some collateral (BTC, ETH, USDC, something you somewhat trust);

2) Use collateral to borrow 1M USDT from AAVE/Compound/whatever;

3) Trade 1M USDT for ~1M of some USD-stablecoin you trust (USDC?);

4) Wait for USDT to crash;

5) Buy 1M cheap USDT from decentralized exchanges (using the USDC you kept aside), which will cost you less than 1M USD;

6) Pay back loan (keeping the difference as profit) and get back your collateral.

You can't just short something you know is worthless. Maintaining the position has costs, but that's not even the problem. If the asset collapses to zero like you believe it should, that usually comes with a halt on trading it. That means you can't actually buy the assets you need to close your position and profit. You can actually be left on the hook for interest on a loan that can never be exited.

Making money as a result of knowing something is fundamentally worthless is actually quite difficult.

That's not how perpetual futures work. Perps settle in cash, not the underlying. You pay interest in the form of funding, which is paid every x hours. The primary risk of using perps is if the insurance fund gets wiped out due to slippage.

I too encourage bearish folks to short Tether. Otherwise, there's simply nothing else to back their claim other than their uninvested belief.

And when tether is frozen on the reference for the future settlement?

Ah – a perpetual swap (perp) doesn't actually settle like a futures contract does. It is purely a synthetic instrument that tracks the price of the underlying instrument (Tether, in this case). The contract technically does not settle (hence the "perpetual") and is priced constantly. There is no moment at which an exchange determines a settlement price because no Tether is held for the sake of delivery.

You may be wondering how the price of the perp tracks the price of the asset if there is no underlying asset to moor the price to? It's done through what's called a funding rate. Basically, there are two prices relevant to a perp contract: the price of the perpetual contract on the exchange and the price of a reference index that is meant to accurately represent the price of the asset on a fiat exchange.

The price of the perp should follow the index price, but there are often deviations. That's where the funding rate kicks in. To moor the price of the perp to the price of the underlying asset, traders pay/receive a funding rate (think of it like an interest rate) every x number of hours (x is determined by the exchange).

If the contract price is higher than the index price, that indicates there is excess demand for the contract on the exchange and, thus, traders who are long the contract have to pay a funding rate to those that are short the contract. That is because the system is trying to incentivize traders to sell (short) the perp in order to move the price down, in line with the reference index. The further the perp price deviates from the index, the higher the funding rate. If the price of the perp is below the index price, the funding rate flips negative. That means that those who are short the contract have to pay a funding rate to those that are long, thereby creating incentive for people to buy the perp and raise the price.

My question is what is that reference index? Tether is so large that the reference can’t be any of the big established crypto exchanges because they might be incentivized to “halt trading” to let the market “stabilize” rather than actually let it breakdown.

FTX lists their reference pairs [here](https://help.ftx.com/hc/en-us/articles/360027668812-Index-Ca...).

  > FTX USD
  > FTXUS USD
  > Binance USDC
  > Binance TUSD
  > Kraken USD
  > Bittrex USD

If there was a run on tether and exchanges shut down trading, I cannot imagine that making the situation any better. Traders would just move off exchange to some place where there's actual liquidity. Moreover, traders would never use that exchange again. Halting would do nothing like it does in traditional markets. In crypto, price discovery cannot be stopped.

In regards to those pairs being used as reference prices, there's no chance they freeze prices just to stabilize the derivatives. Open interest on tether is only $328m on $73b Tether outstanding

> I mean if you are so sure about that the trade is pretty simple. You can trade USDT perpetual futures currently at 0.9989.

On an unregulated exchange where the operator: can see your stops, knows your margin/liquidation price, and can manipulate the price/order book to liquidate you and take your money? I think I’ll pass.

If you're worried about stop/liquidation hunting, you can establish a 1x position that involves zero leverage and zero chance of getting ADL'd out of a position.

> There is no scenario where USDT goes to $10 and you lose your money.

It is possible (even if unlikely) because cryptocurrencies backed by absolutely nothing have value - and arbitrary value at that.

To make it happen Tether could announce they are unpegging and will spend 1Bn to buy back tether at any market price on exchanges. The gambling masses join in. Suddenly Tether is another mooncoin.

This couldn't be further from the truth. Market makers would be glad to sell Tether at a premium. Bitfinex only has enough money so as to retire all Tether coins and buy not a cent more. That said, if you're truly bearish on Tether, then you should believe that Finex doesn't have enough money to retire all Tether coins.

Even this scheme is subject to the risk that the "issuer" of the stablecurrrency stays solvent enough to pay the operating costs of those trades (or at least that there is a sufficient supply of replacement issuers and the legal system and government where the issuer is located appropriately recognise segregation of those assets on a bankruptcy and that it has a low cost and efficient bankruptcy regime), or if it's the US treasury the risk of expropriation of a change to the system which could be different from the risk with traditional currency. Ultimately you're taking credit and/or performance risk on someone.

What’s the practical difference between this and a debit card?

The 2nd though experiment about a stablecoin that that goes up 20% per year is interesting.

So in order to guarantee a stable price, the algorithm has to be able to confiscate some of your coins whenever it needs. This makes sense, but it makes stablecoins less appealing as a store of value.

I've been thinking about a way hypothetical idea for a stablecoin that uses price information. If the purpose of a stablecoin is to aid transactions, why not flip the contract around to be transaction first?

I list a couch for sale on the blockchain. And I say I will exchange it for 799 LegitCoin. And now the blockchain mints out 799 LegitCoin, and auctions them to the highest bidder. And now there is 799 sitting on the blockchain while the product is listed as available.

Because there will only be a 1:1 ration of coins available to coins spendable, you eliminate the role speculation will play. There is no rational reason for the value of the coin to rise above the peg, but there will almost always be arbitrage reasons to drive up the value of the coin to the peg.

I may be silly, but isn’t this proof of stake, where you can stake anything, including a couch?

How does the blockchain know you actually have a couch? How do you prevent someone from listing their toenail clippings for $1T?

I had this idea, too. The problem is exactly what you have stated. Then you have "trusted" nodes that hold the physical asset in hock until someone buys it. A bank, of sorts, to hold your valuables. This bank (and its associated trust) would issue the coinage once it receives the item in question.

Obviously the bank charges you for holding onto your goods (they have to make money somehow) and for the privilege of issuing you cash.

Then, you have to know your banker. Can you trust Muhammid from down the street to issue your coin? Maybe. 3rd party services would pop up, telling you whether a certain issuer had defaulted on delivering some hard good. Turns out Muhammid is a stand up guy, and he delivers when people come calling. So he takes your couch and sends you 799 LegitCoin.

Now Muhammid has options. He doesn't have to hold onto the couch. He has a couch that he valued at 799LC, but maybe he knows someone who would buy it for 815LC. This depends on whether the exchange is a loan or a sale.

[edit]

And what happens if you come back to Muhammid and he doesn't have your asset? Now you tell one of those 3rd party services. If there's a run on Bank Muhammid, and he defaults on everything, his coins are worthless. Now every coin he issued on the network is worthless. Just like small banks in the Western United States before we had the dollar.

Technically speaking, that means any node on the LegitCoin network can mint a node-specific coin. The value of that coin is not locked against any other coin. This necessitates exchanges and means to evaluate the value of said coins against each other.

Because of the cost of holding goods, most banks will hold high value assets; gold, land, rare goods, etc. And, because of the cost of holding those goods, only wealthier individuals will be able to trade assets for LegitCoin.

[/edit]

And so on. Crypto is simply reinventing existing monetary schemes.

Everything old is new again.

If you expand this transaction to an entire economy, I think this is in the neighborhood of how a central bank works. Sora XOR is/was an attempt in this direction, where a smart contract central bank buys and sells XOR on a curve to respond to economic demand. (The price crashed before it could collateralize the curve, so afaik it's never functioned as intended)

>The targeting mechanism is completely decentralized, and free of protocol-level dependencies on specific trusted actors.

...

>As long as you can find an oracle to prove the index,

Isn't the oracle a trusted actor?

Forgive my ignorance: What the heck is RAI?

He throws that at some point in the article out of the blue ...

I used the website google.com and found this: https://reflexer.finance/

The only way I think a stablecoin can work without trust is to be backed by another coin which is deflationary and always in demand. Such as a gas coin that has no block rewards but burns some percent of the mining fees on every transaction. As long as the network is used for transactions, people will voluntarily pay gas fees and deflate the coin (unlike Safemoon which burns it on transfer, paying to secure transactions is a service and the miners can just raise their fees if they want to make more).

Then theoretically as long as there is demand for this coin, it will go up in price and therefore can back any sidechain coin that will grow slower in price (or better yet, gradually drop in price relatively to it like the dollar).

That’s what we are planning to do with Intercoin: https://community.intercoin.org/t/intercoin-application-virt...

> The only way I think a stablecoin can work without trust is to be backed by another coin which is deflationary and always in demand.

OMG. I feel like a fucking alien. People are looking at me like have 2 heads.

There is no such thing as a "stablecoin" ... It doesn't exist! It can't possibly mathematically exist.

Because it's the USA Dollar itself.

That hardly made sense lol

just as something of a useful aside:

the difference between real costs of operations and the price you pay for these operations, and argue that in the long run, costs are more stable and more important than prices

just want to mention the proper business vocabulary (standardized over accounting, finance, marketing) if you want to be understood across your enterprise.

costs are monies you spend for necessities, they entail quantities purchased wrt prices and they reduce your profitability.

Prices are without regard to quantity, and refer to market competitveness, especially "what price should I sell my product at, as I have a choice about that, but also competitive prices in the market"

or as a shorthand, costs have to do with your spending, prices with your selling.

The article is kind of confusing. It starts off noob-friendly, defining what a stablecoin is and the relevant terms. Then suddenly it starts talking about "redemption rate" without ever defining that.

It's a big jump from "we're going to be noob-friendly" to "sorry, only experts who already understand these systems beyond this point".

Rai was initially released based on a value of 3.14 usd, it's down from that. Since usd has experienced inflation one would expect an unpegged stablecoin like rai to be worth more not less usd over time.

Vitalik's requirement for an automated stablecoin to only hold crypto assets is quite severe. If there's a general crypto downturn, people are going to want their money back, which turns into a bank run. It only takes a minority of holders to want their money back in order to create a bank run (and bank runs can start small and get larger because the debt/equity ratio goes down if you are already not at 100% of debt backed by equity and if you pay holders 100% value when they get out of the stablecoin). Basically the only reason to hold a stablecoin instead of the underlying assets is convenience, because if you own an automated stablecoin you don't own any upside but you do own downside risk (e.g. Terra)

But you do own upside---of the collateral.

You're right and my above comment is wrong, but I still think the downside risk is bigger than the upside risk. It's ETH with 100% leverage, except beyond the risk of collapse of ETH there is also the risk of the collapse of RAI. Either of those happening would result in severe losses.

But you could have that upside by owning the collateral.

If I own the collateral, I own the upside and downside of the collateral. If I own a stablecoin pegged to the collateral, I own the upside and downside of the collateral, plus the risk of the stablecoin collapsing.

The stablecoin doesn't offer me any upside compensating for the risk, so we are left arguing that the risk of collapse is negligible, or arguing that there is some other benefit of owning the stablecoin to compensate for the additional risk of owning stablecoins instead of owning the collateral.

Sometimes you can't buy the same collateral e.g. if the basket has many elements/currencies/stocks.

Very true, that case parallels non-crypto assets like index funds. You own the upside and downside of the collateral, and you are also exposed to some risk that the company managing the funds does something extremely stupid and/or malicious.

I said above:

> we are left arguing that the risk of collapse is negligible, or arguing that there is some other benefit of owning the stablecoin to compensate for the additional risk of owning stablecoins instead of owning the collateral

In the case of an index fund, the typical purchaser is motivated by both a belief that the risk is negligible based on the reputation and track record of the fund manager, plus the "other benefit" of the convenience of investing in a single mutual fund rather than trying to purchase the same basket of stocks at small scale.

I agree that an automated stablecoin might offer sufficient convenience to be attractive to some investors, provided they consider the risk of collapse to be negligible.

You can do other things with the stablecoin. You own the upside and downside of the collateral, and you also own the downside of the collateral, and the upside of whatever you bought with the collateral.

Stablecoins are not meant to be an investment; they're for leverage. I agree it is probably a terrible idea to borrow a bunch of stablecoins and then just sit on them.

But by borrowing against the collateral (in this manner), you get some optionality along with the collateral's upside: if it crashes, you get to keep the amount you borrowed[1], thus hedging the loss. Plus any interest earned on it.

[1] Depending on the stablecoin's dependencies you might want to have converted it to dollars outside their platform first.

If you're only holding crypto assets to back your stablecoin, doesn't it eventually work out to 1 BTC = 1 BTC? (Or ETH in this case, I guess).

Not if it's a basket of different crypto. You could peg it to the overall market cap of crypto. It would then by definition be stable in reference to itself, each token would always represent a fixed % of the pool. It still wouldn't be stable in relation to other non-crypto markets, but it's instability would be relative to crypto as a whole rather than any particular coin.

But this is only helpful if you don't need to have your money going back & forth between crypto & traditional investments or currency. It wouldn't be stable relative to fiat currencies. Relative to outside systems it would be more stable than riskier coins but probably less stable than something like BTC. Individual coins are still highly correlated to the crypto market as a whole, so any single coin's stability (or lack thereof) could still move the value of a stable coin like this significantly in relation to outside financial systems.

The code of an automated stable coin exists as a contract in the blockchain. That code can only directly measure and influence things on the blockchain. Thus it can only directly manage assets in the form of crypto.

The contract can indirectly learn about things that are off-blockchain, such as the $USD - $COIN exchange rate, but that requires someone to input that value into the blockchain, and use a complicated set of incentives, incentives stronger than the one to input a manipulated one. Performing off-blockchain actions, would require incentived agents, and be even more difficult or even impossible.

Where in the write up did he state that as a requirement? He clearly says he uses rai as an example for simplicity in explaining the mechanisms. Explaining the systems with DAI would've been much more difficult.

Collateralized loans can't result in bank runs, I don't know how you envision this working. In a downturn, people get their collateral liquidated, or exchange their stable tokens for their collateral.

The user @Proven has a comment that was downvoted and is now dead, but I'm not sure why. The comment seems reasonable. At least this part, I agree with:

----------------------------

I can buy real assets rather than assume that these oracles, "Smart" Contracts, code dependencies, blockchain (network), and developers will remain stable in extreme conditions.

If you need stablecoin, that's because cryptocurrencies are useless to denominate prices. We may as well use a Central Bankster-backed coin (which no one invested in crypto wants to admit).

It may be harder to remain anonymous, but at least there's much less to fail (mostly just the currency itself, which is why buying real-world assets represented by tokens is a much better approach).

Proven's comment is dead because they are banned. on hackernews banned users can still comment, but their comments start out [dead] until vouched. (comments that get downvoted and flagged enough also go [dead], as do comments deleted by HN mods)

if you have enough karma, you can vouch for banned user's comments by clicking the vouch button, it only shows on the direct page for the comment, you get by clicking the date of the comment.

I know all of this because I am banned.

Why are you banned? What's the process for being unbanned?

What dang taketh, he giveth back at his discretion.

As it happens, https://news.ycombinator.com/item?id=29558596 answers both of those questions.

you spooked me with the last line. very dramatic—the imagery is of talking to an old man in a cemetary.

Comments from banned accounts get killed by default. If you see a good comment that's dead, please don't copy it. Instead, vouch for it (https://news.ycombinator.com/newsfaq.html#cvouch). That's the entire reason for the vouching feature.

Other users vouched for that comment (https://news.ycombinator.com/item?id=31521548), so now we have two copies of it in the thread—the original, plus yours—and a split discussion, which makes merging the threads a pain. I'll move the replies to the proper place, but please don't do this again. (edit: especially since you've done it before - e.g. https://news.ycombinator.com/item?id=31178929)

Sorry, I didn't know about "vouching". I'll do that in the future.

Appreciated!

"What happens if we look at a stablecoin from the bold and radical perspective that the system's ability to avoid collapsing and losing huge amounts of user funds should not depend on a constant influx of new users?"

Buterin's idea of humor.

I mean... he is absolutely right, what's your problem with that? Do you think this proposition is radical indeed? If so, it shouldn't look like a joke to you. Or do you think it isn't bold and radical, but an obvious prerequisite for a sustainable stablecoin? Well, then it is really funny, because people are investing fucking millions into a bucket full of holes, without bothering to notice that water is gushing from its bottom like crazy.

The joke is that all crypto is like that. No coin has any actual utility (incl eth). As soon as the supply of greater fools dwindles, the price crashes.

With stablecoins the crash is more spectacular because they are binary nature. Either they are equal to $1 or $0.

Eth can and often does drop without impacting USDC, DAI & others. The overall network and process of filling blocks continues just fine.

Well, you can also say that all modern money is like that. USD is worthless. As soon as supply of fools that believe you can actually exchange it for useful stuff dwindles, the price crashes.

So, no, this isn't the joke and this isn't the point. Either you accept the assumption that crypto has some utility, or you don't (i.e. you either truly believe that, or agree to play the game temporarily, because you believe you can jump off before the assumption becomes false). If you don't — then don't bother with that, it isn't your game anyway. If you do: well, now we can discuss algorithmic stablecoins. They aren't something you are supposed to believe in, they are supposed to be some clever technology that secures the constant price (i.e., the peg) for them, relying on some supposedly safe assumptions. The thing is, it is a relatively new technology with a lot of buzz and not so much proven facts, so if you want to play the game, please review these assumptions for yourself and see if they seem to be safe indeed. This is best done with some thought experiments.

The difference is that it's not "fools" that will accept USD, but people who are taxed by the US government. That's what keeps the USD from being worth nothing; even if you do all your transactions in crypto (which you shouldn't), you need to convert it to USD at some point to pay your taxes.

This is why actual currency and cryptocoins are not equivalent; there's no one forcing someone to eventually convert into cryptocoins, so the value of any individual coin can go quickly to 0 based purely on whims - or someone rug pulls and freezes sales, making off with all the money from people who have bought in.

In order for the USD to go to 0, the US Government would have to collapse. We might argue about how likely that is, but I think we can all agree that it's less likely than a cryptocoin staying at the top of the hype cycle.

Another difference between cryptocoins and currency is the deflationary nature of cryptocoins, which encourages holding, which (if used as a currency) slows the velocity of money in the system, which leads to all sorts of macroeconomic problems like shortages - because no one wants to buy things, people stop making things.

Stablecoins and every other crypto scheme are just speed running the lessons the financial crashes should have taught us over the last 20 years. It shows there's a large swathe of people who saw the crash and thought that they want to be the ones who get rich and then bailed out, rather than preventing them from happening in the first place.

This is only partially true. Yes, USD is sort of backed by the fact there are people with guns extorting taxes from you in USD. Except, if you earn no USD, you shall pay no USD. And if "fools" stop believing anybody will sell you any food for USD tomorrow (because there is a civil war and USA is no more for all practivcal purposes) then USD is worth pretty much nothing, whether you are still required (legally) to pay taxes or not. Reality is stronger than the government laws, always.

So, nope. All this "backed by USA economy" bullshit is just fancy way of saying "backed by nothing". If people don't believe USD is worth anything, it is worth nothing. If people believe BTC is worth a ton of bananas on Marrakech market, it costs approximately as much as whatever a ton of bananas costs in any other currency. Now why people believe that — because there are people with guns extorting the taxes, or for some other reason — is another story.

I.e., all money exists only because people believe it exists and works as supposed.

Anyway, as I've said already, this isn't the point. Everything I've said above I had to say in order to reply, but I don't want to discuss that, it's irrelevant. Stablecoins aren't supposed to be something you believe in or not. They are supposed to just work reliably as long as people believe some other non-fiat currency works. At least, they are promised to do that (and we are discussing if this promise may be any true). I.e., DAI will obviously be worth nothing if ETH is completely forbidden (and that is enforced), so it stops being exchanged anywhere, so then ETH is worth nothing, and DAI is also worth nothing. This is ok, that isn't something stablecoins are promised to magically counter. They are promised, however, to magically maintain their peg, and if that promise is any true is what we are discussing (at least, that's the premise of Buterin's post).

So, to reiterate once more, the question being discussed isn't whether you believe in BTC/ETH/LUNA/DAI/UST/USD/gold. The question is whether if DAI/UST can maintain its peg to 1 USD under the assumption ETH/LUNA isn't totally worthless. This is the only thing that matters for the discussion. The rest is completely irrelevant.

(To be completely fair, it still can end up with the conclusion that you don't believe in the currency on which a stablecoin is based. For me, the mere fact that the whole LUNA's existence was basically justified by UST should've raised some questions about its sustainability. But it's still about "what must happen in order for UST to lose its peg", and not some quasi-philosophical discussions on whether money — be it BTC or USD or even gold — is real.)

> Except, if you earn no USD, you shall pay no USD.

I thought that you do have to pay taxes in USD even if your earnings were in something else.

Of course you have to (or, well, obviously it depends on the laws of your country, but let's be real — what country would agree that you don't?). But there's no earnings, if all your earnings are in magical Warhammer 5000 Gold™ that your government doesn't recognize as money.

(Of course, to be fair, income tax isn't the only tax you are supposed to pay, so I'm vastly oversimplifying everything here, but, again, I'm doing so because all of this isn't the point.)

It's hard to imagine but there are people who aren't American ;)

Btw, while the tax argument was the last thread the fiat fans were clinging to, it should be noted that this is obsolete now since there's at least one country where you can already pay tax with cryptocurrency and possibly a second one that recently introduced Bitcoin as legal tender. As well as the Próspera SEZ in Honduras. Things have been moving fast, it's highly unlikely that this trend can be stopped now, even if some countries aren't going to accept it any time soon.

I'm not a crypto fan but I don't think every crypto is like that. You don't need more users to keep Bitcoin's price stable.

At https://www.f2pool.com/coins under Pow Produced (24h) you can see for each PoW coin, the dollar-value of newly minted coins per day. All these new coins must be bought, for the price to remain stable (miners have huge electricity bills to pay).

Either by existing users, or new users. Probably some mix...

That's the humour, it's not "bold and radical" but obviously true.

First off, I'm not that impudent to assume that it must be "obviously true" for the GP, so I have to review both possibilities. Second, as I already said, if it's "obviously true" then I don't see what GP's problem is, since it actually is funny.

Who says they have a problem with it?

> In the non-crypto real world, nothing lasts forever.

Gold does.

> You can have a stablecoin pegged to a basket of assets, a consumer price index, or some arbitrarily complex formula ("a quantity of value sufficient to buy {global average CO2 concentration minus 375} hectares of land in the forests of Yakutia"). As long as you can find an oracle to prove the index, and people to participate on all sides of the market, you can make such a stablecoin work.

Why would I want any of that?

I can buy real assets rather than assume that these oracles, "Smart" Contracts, code dependencies, blockchain (network), and developers will remain stable in extreme conditions.

If you need stablecoin, that's because cryptocurrencies are useless to denominate prices. We may as well use a Central Bankster-backed coin (which no one invested in crypto wants to admit).

It may be harder to remain anonymous, but at least there's much less to fail (mostly just the currency itself, which is why buying real-world assets represented by tokens is a much better approach).

(The charts and formulas made me laugh - seriously? Who wants to take Cryptocurrencies 101, read some "white paper" and do "research" before they decide to park savings in a "stable" coin??? Give up, it's ridiculous!)

Central banks and large institutional banks are researching blockchain solutions for many of the things they do, not for the rebellious crypto reasons but because a distributed ledger with crypto guarantees can actually be a lot simpler to manage than a bunch of mainframe business logic developed decades past.

This would be less for consumers paying for things but the many methods banks use to settle accounts between themselves.

Example: https://www.bloomberg.com/news/articles/2022-05-26/jpmorgan-...

Nope, that's entirely for PR reasons. For some reason "blockchain" is still hype. Does it even matter there's a blockchain if it's centralised at a single party?

Central banks are looking in digital currencies (the so called CBDC), but blockchain is completely useless for that.

It makes the most sense when you're managing liquidity between multiple parties. I bet it would still be a logical next step if they couldn't get a PR boost off the hype.

If you have a group that has a trusted party (even if said party is one of the group, or a new group made from the group itself), that party can just run a bog-standard database and there's no need for a blockchain.

Of course, it’s an architecture decision. They do different things and different considerations may lead you to one or the other.

Please understand that these types of decisions come straight from management so that they can tick off a box somewhere with the word blockchain. You may have heard this bank invest in IoT and AI/ML solutions as well. It's just buzzwords.

There is no technical reason why this needs to be on a blockchain and it is very likely that the ""blockchain"" is run on some mainframe or 'private cloud' because that is how banks do tech.

I think the value in cryptocurrency has always been tokens which represent real goods intermixed with a digital, distributed transaction network. Smart contracts are a nice bonus.

Unfortunately, early conmen seized the helm of Bitcoin, removed the smart contract opcodes, and led us down the “digital (fools) gold” path.

people constantly see blockchain as a currency replacement. while this is a consequence it is not its raison d'etre. instead it is a decentralized, bank-less, accounting mechanism with its own denomination. it is up to you if you see value in this or not. i personally see value in being able to move my assets with involvement of a minimal (preferably but not necessarily 0) number of third parties

My understanding is that a blockchain is a distributed, publicly inspectable, append-only database is IRB some special properties. Is that not accurate?

there is no incentive to run it without the currency part. you water it down further and it just becomes git

yes or any vc, or an incoming/outgoings journal, or a bank account ... except ... trustless and decentralized

I mean git is also trustless and decentralised

git is trustless to the extent of vc purposes. it is not trustless to the level of financial accounting. for example git has no concept of double spending, while in a blockchain there is (usually?) no concept of branching and merging. but you are right in that they are both based on merkle trees

Double spending avoidance cannot happen without a financial incentive ie currency is what I meant

> Double spending avoidance cannot happen without a financial incentive ie currency

the incentive (reward) is to make decentralized and trustless booking keeping possible, not the other way around

That's just a misunderstanding of why (most) people use stablecoins in the first place. It's not ideological, it's practical.

When you sell crypto and want to exchange it for "real" money, say USD, some exchanges don't support that, they only support crypto to crypto conversions. On exchanges that do support it, it's sometimes slow and often comes with a fee. Both the exchange and your bank may impose arbitrary limits on it. It may create a taxable event.

By comparison, a crypto to crypto conversion (say BTC to USDC) is instant, often free, and without much limitation.

If actual USD had none of the above limitations, nobody would need or use a stablecoin. It's a utility, not some bet against the dollar.

The exchanges that don't support cashing out direct to your USD bank are unlicensed and unregulated and they're dealing in Tether to skirt the KYC and AML requirements that a real financial business has to have. Those exchanges deal in scam coins that are pure pump and dump. Those exchanges don't have to keep client funds in reserve.

This is not the point. Even the most legal/regulated have to ACH/wire cash.

If you’re a trader and want to raise your cash portion of a portfolio it’s much faster/easier/cheaper to hold stablecoins. The cost is that it’s riskier.

I don't know what you mean, for instance because Gemini follows all the regulations, I can hold USD cash in my account and it's even FDIC insured. Wiring back to a checking account is an option, but I can keep liquidity at the ready without dealing in stablecoins.

You're ignoring distributed exchanges like Uniswap. They only deal in crypto-crypto because they exist only on the blockchain.

A significant usage of stable coins is also as margin for crypto derivatives contracts.

But ultimately for the same reason gp stated. It’s easier for the exchange to deal in all crypto

That's not a limitation of USD, it's a limitation of sleazy crypto exchanges. If I sell stock through my stockbroker, I can have funds in a bank account within hours. You ought to be able to sell BTC and get funds in your account in Chase or Barclays via wire transfer within hours.

But no. Crypto exchanges hate to pay out real money. They don't even like paying out cryptocurrencies to external wallets. They want you to just bet within their closed system.

Real brokerages don't care whether you're buying or selling. They get commissions either way. Crypto exchanges have a strong bias towards your buying what they're selling.

The exchanges don't want (or pretend to ignore) the Know Your Customer (KYC) rules, which your brokerage does not.

So your brokerage is fine sending you the funds immediately. They also are fine with the audits, etc that are required to prove that they're not playing sillybuggers with your stocks or funds.

Just a side note, but it's funny how the Patriot Act is commonly decried by everyone as an overreach of government power & invasion of privacy.

But, when you mention these KYC laws (which were put into place in the US by the Patriot Act), no one bats an eye and just assumes KYC rules are reasonable.

Party > Policy

The Patriot Act is a broad-ranging piece of legislation, and the anti-money laundering provisions in Title III are less controversial than other titles.

What is unreasonable about brokerages and crypto exchanges being subject to KYC and AML regulations? How is that at all comparable to the government being able to eavesdrop on my private communications and violate my 4th amendment rights without a warrant?

Financial transactions are private communications.

This is becoming an increasingly rare sentiment, and oddly even on HN. Thank you.

You can simultaneously believe that financial transactions are private communications, and also that a securities broker being required to verify their customer's identity pursuant to KYC and AML regulations is reasonable. Your post is very vague (since the comment you're giving thanks to focused on a semantics "gotcha" and completely avoided the actual point) so I'm curious as to why you think KYC/AML regulations are invalid and brokerages and centralized crypto exchanges shouldn't be subject to them in the light of "private financial transactions".

And how does that make KYC unreasonable?

KYC is the reason we have an archaic banking system while the rest of the world leap frogs us.

Tough to solve when petrodollars are still the reserve currency globally.

Every country I’ve lived in across continents has KYC & AML. Not that it solves corruption, but it’s the bare minimum for functioning societies.

Do tell us which countries have no KYC regulations and are leapfrogging the rest?

In fairness it took me about 2 months to withdraw from an "instant access" account around the new year, because my bank demanded documents that don't exist in the country I'm living in. They also refused to even stop paying me interest that creates a tax liability, so I'm going to have to file a tax return to report less than £2 of interest this year.

>If I sell stock through my stockbroker, I can have funds in a bank account within hours.

Wait, what? I'm pretty sure you're exaggerating there. Stocks have a two day settlement period. (I know because that knowledge gets firehosed every time Robinhood/GME comes up.)

Last year when I sold stock in one account (Wealthfront), for the proceeds to deposited into another, it took four business days (edit: using ACH). When I complained on social media, my finance friends said that was typical. Now, it might have been faster with a wire, but it's not the hours you talk about.

>That's not a limitation of USD, it's a limitation of sleazy crypto exchanges.

It would still be an issue if you want to convert to USD purely on the blockchain because you're interacting with multiple smartcontracts. The USD would need to be a cryptocurrency that lives there.

Robinhood didn't have enough cash on hand for the business they are in.

It's not that brokers are required to wait for settlement. They can pay out as soon as the transaction is logged. They have the option of delaying until settlement, but big customers don't like that, so, usually, they don't. Online-only brokers tend to be sleazier about this.

So, in other words, if you want cash quickly, you can borrow against an asset? (In this case, the unsettled proceeds of the stock sale.)

Good news: you can do that with smartcontracts too! (e.g. Compound/AAVE)

No, the broker is borrowing against their own assets. They already did the transaction. They just haven't been paid for it yet. It's their accounts receivable problem, not the customer's. That's what it means to be a broker, rather than an exchange.

Since brokers usually have transactions flowing in both directions, it's usually a wash.

Except that’s not “what being a broker is”, because not all brokers offer that, and not to all clients. And if you can remember back to your original comment, you were calling exchanges sleazy for not having insta-withdrawal (which they can’t in the regulated markets because of settlement time), and now you recognize this is a service provided by brokers as an abstraction on top of the actual exchange, not what said (non-shady) exchange actually offers.

Furthermore, the broker is taking a risk by extending that credit. If it were riskless, there wouldn’t be the 2 day settlement period or the requirement to post collateral (whose necessity everyone accepts with an eyeroll at those who don’t get it on the Robinhood/GME threads).

Using TD Ameritrade I can close out positions and get paid by EOD.

ETrade and other brokers are similar (though the first outbound payment may take extra time since they need to run KYC checks).

How about someone on any other country that does not use USD?

Aside from some smaller countries in South America and others that fully adopted the USD for their economy, you can not get exchange other assets for USD without significant overages.

Not sure this makes sense. I can exchange USD for my local currency, GBP (including physical notes) for significantly less in commission and bid/offer spread than crypto exchanges charge for crypto to USD.

Try the same with the Brazilian Real or the Argentinian Peso.

Also, try doing that with more than 10k USD.

Also, try sending it to someone overseas.

Yes - countries with capital controls and/or corruption problems make life harder in many ways. But I'm not sure I understand why stablecoins make that any less of a problem when compared with holding USD in a US based account - unless the point is to avoid AML/KYC requirements.

Same for larger amounts and sending money internationally - I've done both quite frequently, and it's much cheaper to do than using crypto would be.

In a global economy, it is a lot easier to acquire/transact/hold stable tokens than actual USD, that is the point. If all you care about is the developed bubble, crypto makes little sense.

> unless the point is to avoid AML/KYC requirements

That's the entire raison d'etre of these stablecoins.

Also, try smuggling some cocaine into any country.

Right, because an immigrant working in the US and helping their family to buy a house in their home country is exactly the same as being a drug dealer.

If you're an immigrant working in the US, the rules don't apply to you?

Even though I could just tell you that blindly following rules is a trait of morons and authoritarians who have a control fetish, or go on a diatribe about "legal != moral"... notice how I didn't say anything about not following the rules. The point was about the cost of doing large transfers with crypto vs a traditional bank or currency exchange shop. You can report the crypto transactions just the same, you know?

(Maybe it is time to change HackerNews' name to something more reflective of the current audience. What do you think of "Conformist 'R Us"?)

Of course you didn't say explicitly, you implied it... since if you add the bureaucratic costs of international money transfers, then crypto is much more expensive than a wire transfer, it would make absolutely no sense to use crypto if you followed the rules.

> since if you add the bureaucratic costs of international money transfers it will be more expensive

That is plain, utterly, provably wrong. You have no idea of what kind of fees a bank will charge to exchange a wire of tens of thousands of dollars.

> it would make absolutely no sense to use crypto if you followed the rules.

Some of the rules are very specific at about the source and means of the funds. E.g, some taxes in Brazil are applied only for purchases done through credit cards. Others apply for financial operations between different banks. It is not illegal (and much less immoral) to know about the loopholes that allows you to avoid paying the exorbitant fees.

> provably wrong

A distributed system that employs a consensus mechanism where service providers compete by wasting as much electricity as possible is provably less expensive that a centralised system where providers compete by trying to become more efficient at providing services?

Then prove it.

Nice gaslighting. We were talking about the cost of sending a transaction vs the fees that a bank will charge to make an international transfer.

Now, if you really want to talk about the cost of blockchains: Ethereum's transition to PoS will mean a 10000x reduction in electricity consumption. If you look at the total number of validators vs the required power to run one, the most pessimistic estimation puts the cost of securing the whole network at $8M/year. That would be less that the amount that banks spend on physical security of their armored cars alone.

Ethereum doesn't use PoS. We're not discussing science fiction or hypothetical scenarios, here. We are discussing the reality of crypto-currencies vis-a-vis the banking industry.

As I mentioned, when you remove trade barriers, such as the EU did, the cost of international transactions falls to zero. This is because 1) the technology that banks use to transfer money is efficient, and 2) competition drives prices down to the average cost per transactions (which is near zero, thanks to 1). Now, in the crypto-currency industry you do have competition among crypto-currency transmitters, but you don't have an efficient technology to transfer crypto-currencies, instead you have a technology that performs horrendously in terms of cost-effectiveness, so transacting in crypto-currencies will always be more expensive than using banks. If this is the case, why do international transactions sometimes cost more if you use a bank? I already explained, it's because governments require banks to perform a series of checks therefore incurring additional costs that they pass on to customers. Transmitters of crypto-currencies don't perform these checks, and therefore they don't incur these extra costs, but of course this results in transactions that are not compliant with financial laws and regulations, and so the end-user will likely have to spend additional money laundering the funds and evading law enforcement. If you say that you can prove me wrong (this is what 'provably' means), then prove me wrong.

There is no science fiction about PoS, the transition is well under way and it is a matter of a few months.

> when you remove trade barriers, such as the EU did, the cost of international transactions falls to zero.

See, the whole point of crypto is that it gives us a globally interconnected trade network. If I just want to send money to someone in the EU, then of course I will just use SEPA. But SEPA means absolutely jack shit for someone trying to merely send money to a place where these basic freedoms are not granted.

Why should people should put up with the artificial barriers, when there is a parallel network that can work better for them? Do you think they can just be waiting until the powers-that-be at their local sphere to reach enlightenment and remove the barriers? I think it is a lot more efficient if we continue to work on a competing alternative for the people, at the very least to keep the goverments in check.

Yes, you do not live in a place with rampant corruption and/or capital controls.

As far as I know, every country in the world has capital controls.

So it is just a matter of how hard you like the boot stomping on your face, huh?

The idea that promoting a tool intended to undermine governing institutions, democratic and undemocratic alike, will result in a world with less oppression is one of the dumbest pitches the crypto community has come up with so far.

the argument is that some people have value in holding dollars. if stablecoin enables getting people those dollars cheaper than they could get them before, that is good for those people.

As I was saying, some people have value in holding cocaine.

Okay?

Please read https://news.ycombinator.com/item?id=31463534 before creating more strawmen.

I'm not sure why the comment was worded "some exchanges" but notably, this is the only way that a DEX (distributed exchange) can work. A DEX exchanges crypto directly on the blockchain between wallets/accounts via smart contract. These are literally just dapps (distributed apps) and no company needs to be involved; they are the most open exchanges of all.

So you can get settled by your stockbroker during a weekend?

Most payments providers will not settle during weekends and most will settle at T+2 or T+1. Instant settlements and weekend settlements are basically unheard of in the financial and payments worlds, trust me, I work in this world.

Stable coins allow almost instant settlement and weekend settlements, that's why merchants and individuals are interested in them.

Stablecoins aren't settlement - even with fiat backed stablecoins, you're an owner of a portion of the most junior debt of a private company.

You can pay with them, but you need to convert to fiat to settle.

You can sell VOO at midnight and transfer the money to your bank account? I don't think so or at least I'm not sufficiently privileged to do so.

> If I sell stock through my stockbroker, I can have funds in a bank account within hours

What brokerage are you using that ignores days-long settlement times? The only way you're actually getting this is a margin account in the background, and that comes with its own risks and limitations.

And even then, it takes days for ACH to clear, or fees for wires or debit deposit.

>If I sell stock through my stockbroker, I can have funds in a bank account within hours. You ought to be able to sell BTC and get funds in your account in Chase or Barclays via wire transfer within hours.

This sounds false on several levels. Stock settlement occurs 2 days after the trade executes. Then transferring the money from the brokerage to the bank account via ACH takes 3 business days. Some brokerages may call the cash settled after 1 day, but the ACH still takes multiple days to land in your bank account.

Wire transfers often come with fees outside of Credit Unions who may generously waive it but the cutoff time is 5PM EST at best and in some cases earlier than that, otherwise you need to wait until the next business day.

Maybe they're in a country with a bank transfer system that doesn't suck. In most of the OEDC there are free transfers that take <2 hours even in the worst case.

>It may create a taxable event.

What’s the difference in tax liability from trading your BTC for dollars vs. a stablecoin?

I’m aware that there are places where you can exchange crypto for crypto that won’t report it to the IRS, but that doesn’t change whether it is a taxable event.

By the way, there’s no statute of limitations for tax fraud.

In some countries, crypto-for-crypto transactions are surprisingly really not taxable events!

I guess you could try to argue it is a 1031 exchange (but that's now real estate only, basically). I suspect it's much more that less-than-fully-compliant exchanges will only show "inputs and outputs" and not any trades in-between.

https://www.irs.gov/businesses/small-businesses-self-employe...

The IRS has offered clear guidance that these sorts of exchanges are taxable events. You could argue all you want, but ultimately it is taxable.

It's been discussed several times... Many people are "trapped" into cryptos and can not get out (for example because their bank will close their account and kick them out if they do anything crypto related). So in several countries it's only when you sell for a currency that is legal tender that it's a taxable event. It's the case in France and, if I'm not mistaken, it's been clarified that it's now also the case in Germany.

It makes sense in a world where people are taking a very big risk by selling for a "stable" coin. For the last thing you'd want is people selling BTC for Luna / UST, and then owing the state, say, 100 K EUR, only to then see Luna going to zero. Some of these people would have their lives ruined if the taxable event happens before they're able to cash out to a currency that is "real".

If anything the Luna / UST fiasco as shown to many that it makes sense to only tax when something is sold for a real currency and not for monopoly money.

>Some of these people would have their lives ruined if the taxable event happens before they're able to cash out to a currency that is "real".

Sure this is a downside of tax law on the US, but it’s not unique to crypto. Suppose you sold one stock and used the proceeds to buy another stock, which then went to zero. It doesn’t matter whether you went to cash at some point, or if somebody exchanged the stocks with you directly. Either way, you owe taxes on the gains in the initial investment. Depending on timing, the gains and losses might cancel each other out, or you might end up with a tax liability in some year. You still need to report it.

This is a special case of an even more general problem: “I was rich, paid taxes, then later became poor.” A lot of the time, the tax man is unsympathetic.

Will the utility persist as governments ramp up compliance requirements on those exchanges? So far they've largely flown under the radar. But eventually they will be forced to do more to comply with AML/KYC and tax reporting rules.

The phrasing here seems to imply that exchanging one cryptocoin for a stablecoin should NOT create a taxable event. That's clearly not a long term reasonable expectation to hold as it's an obvious loophole all tax authorities will be looking to shut in the longer term.

Why should it, though? It's still just another cryptocurrency at the end of the day. The government shouldn't be looking for their vig until you exchange for legal tender, and I don't see how a third-party organization promising a specific exchange rate for the token changes that.

How is it a loophole? Tax applies to your capital (gains) denominated in USD.

Example 1: I deposit 10K USD and buy BTC for it. After a while I sell the BTC and get back 15K USD. I've now gained 5K in USD, which is taxable.

Example 2: I deposit 10K USD and buy BTC for it. Next I swap the BTC to ETH, a stablecoin, or any other crypto token. I've gained nothing in USD. I still have 10K USD worth of crypto. I didn't sell crypto, so there should be no tax.

Unless...you formally acknowledge a stablecoin to be representative of USD. Which may have all kinds of complicated implications.

It's not just about central exchanges. A stablecoin lets you use on-chain exchanges like Uniswap, loan money on chain, etc.

(I don't think there's a tax advantage though. I'm not a CPA but my understanding is that in the US, any exchange of one token for another is a taxable event.)

For cryptocurrencies with fees priced with gas, price volatility is not a problem. As the price grows, number of gas units per operation will likely go down, but will maintain its "stable currency" price.

Ether's price doesn't have to stay fixed. It can grow or fall, increasing or decreasing economic security of the network (assuming Proof of Stake). There is no good way to ensure price stability, and there's no reason to. We don't complain that stock prices change.

That's why smart contract cryptocurrencies have stablecoins: to address "stable use case". But this is just one of many possible "dapps", many don't need external peg to function (NFTs can be priced in the volatile ETH just fine).

Note that gold can have long run negative returns if you count, storage costs, transportation costs, insurance costs etc. plus it's not very stable in the short run.

How about leaving fiat to the retail and central banks and not trying to invent wheel on top of a rocket ship

You are posting on hacker news.

Hacker.

The defining thing of hackers is taking things apart, understanding them, modifying, breaking, rebuilding, improving, dreaming!

What a sad ideology of “don’t touch, leave it to the authorities”

Satoshi was a hacker. These Defi charlatans are scammers at best.

World of Crypto (WoC) has already addressed this issue [0].

[0]: https://www.worldofcrypto.io/blog/building-a-decentralised-m...

Reply on news.ycombinator.com