59 comments

[ 3.1 ms ] story [ 79.7 ms ] thread
This post doesn’t so much refute the idea that the mango markets manipulation is nihilistic as confirm it.

I personally am fine with a strict view of these cryptos where the code is law. But it is a very paired down vision compared to existing systems and thus the “nihilism” label.

> one of the advantages of cryptocurrencies as the foundation of a separate financial system is that systems are, by necessity, subject to adversarial hardening

This is what the law is for.

> Perhaps it is true that Avraham committed illegal market manipulation (or some other violation of the law). However, does it really make sense to call for prosecution and imprisonment?

Yes. They broke the law. Laws should be fairly enforced.

> Furthermore, if there is any core ethos of crypto, it is one of independence from the authority of the state.

No, that's the core ethos of a lot of people in cryptocurrencies. It turns out that there are a lot of "investors" in cryptocurrencies who do not feel that way :)

The whole point of this experiment is not to create an environment where law can be skirted, but to create a new type of environment where the law cannot be violated at all, by anyone. And of course it will attract lawbreakers, we need them to test our security. "The law" as it exists now, written in the wind and enforced when convenient or profitable, needs to go the way of the dinosaur. The law as written in self enforced, freely entered agreements that is public to all and impossible yo viate sounds like a better system to me.
> Illegality of market manipulation. Perhaps it is true that Avraham committed illegal market manipulation (or some other violation of the law).... Let us remember that Mango Markets is an unlicensed derivatives exchange run by a team of US citizens who also held an ICO for their protocol token, which likely qualifies as an illegal security ー do we really want to be tallying up legal infractions in this situation?... Almost everything in decentralized finance is illegal on multiple levels and we are all operating on very shaky ground here.

What a absolutely stupid argument.

Mango Markets ICO didn't allow US citizens to buy tokens, so it wouldn't be regulated as a security - so it is legal. But even so it's pretty easy to argue that the security itself should be legal if certain education requirements about risk are met.

That's a very different class of offense to bald market manipulation for profit.

> Furthermore, if there is any core ethos of crypto, it is one of independence from the authority of the state.

So this piece argues. Personally, I'm all for regulation of financial markets - as someone on HN mentioned "crypto seems to be speedrunning the history of financial regulation".

I don't think it's legal for secondary markets (e.g. Kraken) to allow Americans to trade ICO tokens either. Basically no ICOs have ever been done legally.
Depends on the token type. There's a distinction between governance tokens (~= securities) and utility tokens (~= coin|currency|value store|something).

Utility tokens don't fall under the Howey test so are ok to trade: https://www.blockchain-council.org/blockchain/security-token...

Realistically, there are virtually no utility tokens.
100% agree.

"He might have done market manipulation, but it was on an exchange that might have also done some(unrelated) illegal thing earlier, so ....". That's just not how laws and regulations work at all.

I don't have a particular axe to grind here. In the markets I worked in, what he did would have 100% been market manipulation but as I have argued elsewhere he may successfully be able to claim normal market practise here idk. It would be risky because I don't think it's been tested in the courts before. From my perspective crypto markets should be regulated in the exact same way as conventional markets because regulations are needed for the exact same reasons. The idea that you can just say "code is law" and pretend that laws don't exist seems unlikely to be true.

Also while I respect his decision to make losses from depositors whole, I'm not sure that would change much if a regulator did decide to try an enforcement action here.

> while I respect his decision to make losses from depositors whole,

Noting that he only made this offer after he had been named on Twitter.

This is some paper thin armor you're hiding behind.

Everyone knows how the game is played. Turn on your VPN and you're in. There is no real attempt to stop Americans from buying.

Again, this is the wild west. Rules for the and not for me deserves merciless mockery and derision.

KYC is a real and increasing barrier to this.
> In my view, one of the advantages of cryptocurrencies as the foundation of a separate financial system is that systems are, by necessity, subject to adversarial hardening. In this interpretation, Avraham’s actions act to strengthen the system in the long run.

The language of "adversarial hardening" is torturous, and amounts to a tacit admission of failure: that cryptocurrency somehow needs to go through a process of making every single mistake that traditional finance has made and corrected for.

The necessity of this repeated rake-to-face process is not elaborated or further justified.

Why must it be justified? This is a financial system, not a battle of rhetoric. If you don’t want to participate, don’t. Evidently, enough people want to participate.
> This is a financial system, not a battle of rhetoric.

No, this is a blog post about a well-understood failure mode in traditional finance. If cryptocurrency advocates want to lay claim to the "financial system" label, they should be prepared to explain how their schemes are efficient mechanisms for value transaction, not sandboxes for making sure the rules of financial crime haven't changed.

No one needs to explain anything. I transact in it and it works, therefore it is an efficient mechanism for value transaction. If you don't want to participate then fine, but as the person you replied to said, evidently enough people do.
Sure. Let me rephrase: being taken seriously requires you to explain &c. This is a public forum; nobody particularly likes it when someone says "this is so" and doesn't offer an intelligible explanation.

With enough incentives (including a large pool of greater fools), you can convince otherwise reasonable parties to use your scheme. But that doesn't make it a financial scheme; that makes it a financialized scam.

I take it seriously. You don't.

That's the thing about this stuff: it's voluntary. That's the point the above commenters are trying to make. We don't need a consensus of you all to do what we want, and if you don't want to be involved, don't.

I take it seriously enough; I have a civic interest in harm reduction in this area.

“Voluntary” cuts it when you’re talking amongst yourselves, but again: this is a public forum, not one specially for your interests. I get to voice common sense objections; that’s the point of this place.

Sure you get a voice. We can have a discussion.

But the point being made above is that we need someone's respect, someone's consensus, someone's approval to do what we are doing. And to that I say if you don't like what we are doing don't get involved.

> We can have a discussion.

That's all this has ever been!

> And to that I say if you don't like what we are doing don't get involved.

This, of course, only holds as far as cryptocurrencies remain a non-risk to most "main street" investors. But they're creeping towards that group (as the next biggest free pool of capital), and I have a civic interest in those people not losing their shirts.

No man is an island, &c.

You transact in it, it works, and then sometime later everyone's funds are wiped out by hacking, arbitrage and others.
> and amounts to a tacit admission of failure: that cryptocurrency somehow needs to go through a process of making every single mistake that traditional finance has made and corrected for

These problems are different and the solutions are different. The way traditional finance solved this was to incorporate the big scary people with guns to enforce their rules.

It would be an admission of failure if they called for government regulation, but that’s not what’s happening here. They are going back to the drawing board to keep hardening the technology in the face of openly adversarial actors.

> These problems are different and the solutions are different. The way traditional finance solved this was to incorporate the big scary people with guns to enforce their rules.

Traditional finance did not incorporate the Scary Men with Guns. They were already there, are still here, and will remain regardless of how many layers of decentralization you add.

Traditional finance's crowning achievement has been to successfully integrate itself into civil law: we don't send people to the Poor House anymore, and we don't execute bankers (or ordinary citizens) when they fail to finance the King's wars. Among the "kinds of crime" you can conceivably execute on as an individual, traditional financial crimes are among a handful that are regulated by agencies that don't have actual criminal enforcement abilities.

There is no "going back to the drawing board," only muddying the waters and hoping that someone won't notice the Scary Men with Guns (who have no intention of leaving) in the reflection.

> we don't send people to the Poor House anymore

We let them get into debt they can't repay and then send them to jail instead, but functionally it amounts to the same thing, except for being harder to get out of.

> We let them get into debt they can't repay and then send them to jail instead, but functionally it amounts to the same thing, except for being harder to get out of.

These do not amount to the same thing: one is debt, and the other is an extension of the carceral system.

As best I can tell, cryptocurrencies do not eliminate individual debt. Rather, they provide myriad new ways for ordinary laypeople to leverage themselves into lifetimes of poverty. So it's not clear what this (true!) observation has to do with the conversation.

I'm no fan of cryptocurrencies, but as far as I know they do avoid debt; while there are e.g. crypto margin loans, if you get margin called and blown out then no-one expects to be able to collect the remainder, much less use the machinery of the state to enforce that.
That's a practical matter: either cryptocurrencies will (in their quest to become actual financial systems) make debt a serious problem for the indebted, or they will continue to be financial "mystery meat."

In other words: they avoid debt because they're not doing real finance. But the entire point of this grand experiment is supposed to be a real financial system, so "it's currently broken" is not a good argument for "it will get better."

Cryptocurrencies "avoid debt" only as much as regular currencies "avoid debt". If I borrow USD from someone, I sign a legal document promising to pay it back in a certain time frame and they give me the full balance of the loan immediately - it has nothing to do with the currency itself. Cryptocurrencies don't somehow stop you from doing the same.
Sure, the point is that there's a system in place that will enforce that debt with (the threat of, and ultimately the reality of) violence. And the more subtle point is that it's impossible for a system without real-time settlement to avoid having debt baked in at a very foundational level, whereas real-time settlement and smart contracts offer hope that an alternative foundation might be possible.
This isn't how lending in the cryptocurrency world works.

Lending protocols rely on overcollateralization of loans instead of contract law. If you want to borrow $100k, you would give over (say) $200k worth of some token that you didn't want to sell. In that example, if the value of that token dropped by 50%, the lender would sell all of your collateral to recoup their $100k, not take you to court or send you to a collection agency.

If on the other hand you eventually return the $100k, you would get your collateral back.

What you are describing seems more akin to option trading in tradfi than loans. The only reason someone would be willing to give 200k worth of collateral for a 100k loan is because they are betting that the 200k worth of crypto will increase in value more than the interest on the dollars they are borrowing.

The use case for almost all individual loans in tradfi is getting a loan for more than you have in collateral.

> The only reason someone would be willing to give 200k worth of collateral for a 100k loan is because they are betting that the 200k worth of crypto will increase in value more than the interest on the dollars they are borrowing.

Well, most of the people who have crypto to borrow against would fit the description of "betting crypto will go up."

Tax implications might also play into the consideration of selling vs taking an overcollateralized loan as well.

> In that example, if the value of that token dropped by 50%, the lender would sell all of your collateral to recoup their $100k, not take you to court or send you to a collection agency.

What do they do if it drops by 90%? You'd still be holding their $100k, on only $20k of collateral.

I don't see how this avoids contract law; overcollateralization has never done so in any economic system in the past, and has historically been a ripe domain for financial fraud (since the incentive is to over-value the collateral).

> What do they do if it drops by 90%? You'd still be holding their $100k, on only $20k of collateral.

I'm not super knowledgeable here, but I expect such protocols would tend to only take the top few most liquid tokens like Bitcoin or Ethereum as collateral to somewhat mitigate this issue. Very low liquidity tokens held as collateral could be subject to manipulation to trigger loan liquidations as well.

This doesn’t meaningfully eliminate the problem; it only moves the risk proportions around. Someone is still losing incredible amounts of money in the form of worthless debt should a crash happen here, and it’s not clear what their recourses are.
Ah, I think I see what you mean now.

The ability to liquidate collateral is the the stand-in for both creditworthiness and the mechanism for recouping their principal. They would have no recourse to recover their funds if a crash were so drastic that they could not meaningfully liquidate collateral.

Pretty sure that is practically and technically wrong because you're only considering one side of the loan. When lending fiat the specific currency does matter - because often the lending institution will often create new units on the spot. We can tell by inspecting a chart of monetary aggregates that someone is creating new units out of nothing. And when financial institutions over-commit sometimes they get bailed out so they have an incentive to take on more risk than is prudent.

Both those operations are somewhere between less likely and impossible with crypto lending. If an institution lends a bitcoin, they had to acquire the bitcoin from someone else first. And if I borrow the bitcoin then forget my keys that bitcoin it isn't going to come back to them without someone else giving up a bitcoin - no matter what the regulators want to happen.

They can only avoid this by not accounting in bitcoin and working in fiat. Or by using pure credit, like what institutions do now where "your money" is actually something like an individual being a bank creditor. But it is much harder both technically and as a matter of incentives to lend crypto.

That's one take, I for one am a bit more optimistic. I think we can disempower the scary men with guns quite a bit if we take their feet out from under them.
In general, threatening the Scary Men with Guns’ money tends to make them scarier. One recalls images of Iraq divided between oil companies.
> Going back to the drawing board to keep hardening the technology in the face of openly adversarial actors

You just have to hope that those actors are altruistic enough to return some/all of the money.

Otherwise you may not get the opportunity to harden the technology.

And more to the point, there’s no suggestion of this process ever stopping. There will always be bugs—new ones discovered and old ones reintroduced.
On it's own, that would be fine! Nobody expects criminals to stop innovating on crime; it's fundamentally an arms race. The problem is not the unending process itself, but the flawed (and conceited) belief that cryptocurrencies don't have to solve the same problems.
Yes but we do expect to be unaffected by financial crime.

I have never been the victim of a financial crime, like an institution trusted to hold my assets just not doing it. Most people I know haven’t either.

Granted it’s not like it never happens. But it’s exceedingly rare and people’s standards are incredibly high. We demand and expect that if you put money in a bank you can always get it back and we mostly stick to that plan.

Well said. In a perverse way, financial schemes built on top of cryptocurrencies represent the crowning achievement of traditional finance: we've successfully made it so safe, so incredibly boring for ordinary citizens that technologists believe they can derive the same system from first principles.
When I’m dealing with a bank, I don’t have to care about the arms race. If the bank gets hacked, I haven’t lost my savings.
I don't quite buy this. "There will always be bugs" is a truism, along the lines of "all code bases will have to be maintained forever." But that's just not true, code can be done. I think it's really a problem of tooling. We have formal verification, you can come up with an idea, write a specification for it, and write code that does exactly what is intended. It does require more rigor than we currently have, someone attempting to build something must thoroughly understand what they're trying to accomplish and understand how to properly use the tooling, but it can be done.
Avraham clearly is not skilled in public relations. Perhaps there is an element of nihilism, not in his wallets actions but in his willingness to surrender to the mob.
I don't know how much work he put into this trade, but I would also prefer ~$50M free and clear over ~$110M with the threat of lawsuits attached.
There will never be enough "adversarial hardening". It will never be possible for people to verify that their Lego blocks of finance are actually robust. The only thing that will create trust is time elapsed without an exploit, but that means that the only thing trustworthy will be the oldest, most ossified pieces. But as soon as you start composing those pieces with anything custom, you're back in danger. Welcome to the dark forest.
<< It will never be possible for people to verify that their Lego blocks of finance are actually robust.

Eh. Tbh, I am not sure what robust means in this context. From what I have seen so far, a ridiculously high amount of finance moves through simple text files ( once you get down to that level ). From technical perspective, it is not hard to mess with ACH, but there are levels of compliance enforced by live humans that keep it together.

I do think I agree with your conclusion though. The most ossified piece wins.

Blacklisting/censorship of tokens/addresses is an existential crisis for defi that noone wants to talk about.

Most of defi is based on collateral, and if locked collateral can be debased by censorship, the whole thing falls apart.

A good reminder as to why crappy old dumb rock ossified btc with no features and 13 years of track record and no founder is something pretty special

I get the potential existential crisis issue - but bitcoin is subject to that, because can't governments blacklist a wallet? Then that person can't effectively take their btc away, just like in regular banking, based on some on govt identifying a criminal action, people lose their bank money when it is identified (drug dealers or whatever).
For crypto, I think it has to be "code as law". Otherwise, what is the point. People will be better off depositing their funds in a brokerage and trading that way. Trading and settlement have to be merged to get the benefits from the crypto approact to money.

I think the next step for smart contracts is to port risk engines (for leveraged trading) and treasury risk management into smart contracts. But I think there is little overlap between people who write risk engines (almost never open source) and people who write smart contracts with slick UIs, so I don't see this happening.

I think SBF (see https://twitter.com/SBF_FTX/status/1580170203664904195?s=20&...) of FTX, who knows how leveraged trading works, explained what risk controls FTX implements.

A simple risk control would be to have a single transaction rate limited in funds it can suck out of an account.

I agree 100%, those of us that embrace the crypto markets wanted a wild west. you don't get to have a wild west only on the upswing and go crying for stability when the bullets start flying in the street. Either you want it or you don't, if you don't go buy google stock or something.