[marcinjachymiak]

I'm not a fan or user of Celsius, but this is an incredibly inflammatory title with very little evidence to back it up. Claims like that demand greater proof. This is an incredibly lazy article

The author pretty much failed to do any research on DeFi investments (point 3 in the OP). Compound and Aave are just 2 of many places investors place their assets, and are definitely near the lower end of APYs. Badger, which Celsius has already said they used, offers much higher returns. So do other projects like e.g. Convex Finance.

Where do these high APYs come from? Well a lot of it is coming from incentives of these protocols and speculation in those native assets. But the author doesn't even know that, so I won't bother steelmanning his argument

[vesinisa]

So instead of actually debunking his arguments you published a handwavy reply? I very much think that if someone is offering insanely high ROIs but does not divulge how the value is created, you can safely assume it's a scam, and simply pointing out the disparity suffices as proof to me at least.

[tinco]

So, definitely not in defense of Celsius it smells funny to me as well. But the author doesn't seem to understand DeFi. I don't know how Celsius operates but if there's a genuine zero-knowledge proof of their operations then that is better than an actual audit. The author doesn't understand this and proceeds as if no audit actually happened. If the author instead spent some time researching the contents and the merits of the proof they'd make a better point. If the zK proof is bullshit, that's a far worse indictment than all the other circumstantial evidence in my opinion.

Anyway, I don't have much faith an operation ran by an old con artist and a 24-yr old actress would actually produce a valid zK-proof that covers their operations. That's the sort of thing you need a serious technical team for. So the article still stands in that regard.

[rglullis]

Like he said in TFA: the zk-proof only shows they have control of some funds, it doesn't say anything about the origin of those funds, much less about present and future obligations.

[ceejayoz]

> I don't know how Celsius operates but if there's a genuine zero-knowledge proof of their operations then that is better than an actual audit.

Someone might be able to demonstrate to you that they have a $20k bank balance.

Audits are where you try to do a reasonable job of making sure they aren't also hiding the fact that they've got $50k in credit card debt lying around.

[marcinjachymiak]

I don't know what happens behind the scenes at Celsius.

I do know there are many ways to get those high ROIs in DeFi.

With zero evidence suggesting they are lying about returns, the charitable guess I can make is that they are getting high DeFi returns and giving their customers slightly lower ones after taking a cut.

[adam_arthur]

Can you explain some of the ways then?

Not the user facing side, but the backend/underlying part that generates the returns necessary to sustain those yields.

[TimJRobinson]

Most of the high yields come day traders paying trading fees. For example on sushiswap on Polygon there are pools of ETH/USDC that people trade back and forth with trying to play the market. Every time they do a trade they pay a 0.3% fee. If you provide liquidity to these pools there is so much trading going on that the returns are currently 20% - 30% per year.

You can earn even more if you provide liquidity for riskier assets that have a higher chance of dropping in value.

The risks are impermanent loss and that the token you're providing liquidity for drops in value.

Yearn is a decentralized automated tool to find the platform that is giving the best returns and moving all the money they manage into it.

All of this is open source and on the blockchain.

[adam_arthur]

But the fees are taken from people speculating in crypto to make money.

Even though these are "fees" or "loans", the purpose is strictly speculation. It's not driven by intrinsic value, but people attempting to make money via speculation.

If I take a loan for a house, I can live in it. That's intrinsic value.

So the system only works to the extent that people are willing to pay for tokens without intrinsic value. Does not seem sustainable to me

[TimJRobinson]

I don't particularly care why they're trying to trade the markets, or if they're providing intrinsic value, but I'm happy to invest money and collect 20-30% dividends from them.

If everyone suddenly decides they don't want to trade any more, then I'll find somewhere else to invest my money.

[sec400]

Not Celsius specifically but you can see details about yearn's strats here: https://docs.yearn.finance/getting-started/products/yvaults/...

Specific vaults strats: https://medium.com/yearn-state-of-the-vaults/the-vaults-at-y...

Links to the actual contracts: https://yearn.watch/

[adam_arthur]

I just glanced through these and don't see any explanation.

They basically just say "we put it in a vault and harvest the rewards".

What I'm asking is where do the rewards come from. What is the underlying mechanism that makes this model sustainable.

If you invest in a REIT, tenants earn money through their business and pay rents. If you invest in a BDC, the BDC makes loans to businesses and collects interest. Relationship and risks are quite clear here.

If you're Bernie Madoff you generated high yields for investors for decades by taking money from one investor to pay another, and ultimately was not sustainable and bankrupted many people. For example.

So are DeFi yields like a BDC, or like a Bernie Madoff?

[bhouston]

I think the main answer to most of these questions is that everything works well as long as BTC and most others currencies continues to rapidly increase in value as a result of a lot of cash inflows into these cryptocurrencies. This papers over all of the fraud at least for now...

When the explanation is too complex for anyone to really grasp or verify, realize that this is probably intentionally opaque in order to hide the fact that it is either hugely risky, built on a house of cards or it may be just outright fraud.

[anonymoushn]

Some yearn vaults invest in liquidity pools that let users pay 10bps or whatever to trade one stablecoin for another. So the yield comes from that 10bps fee.

The sUSD yVault deposits sUSD into overcollateralized lending protocols and collects interest on loans to other users (who are presumably using Aave or CREAM as places to buy leverage to get super fucking long crypto). So the yield comes from other users paying interest to borrow sUSD.

[SilasX]

I think I can answer your questions.

Background/Disclaimer: I'm long crypto and have significant assets in Gemini's Earn program, a regulated variant of crypto lending that pays 8% on their stablecoin, GUSD.

I found Celsius and Yearn to be too sketchy/inscrutable to bother with[1], so I'm not going to defend anything about them, only the narrower claim that (some) DeFi liquidity pools are a non-Ponzi, sane way to earn returns in some conditions.

Liquidity pools [2] function as automatic market makers, using their assets to allow traders to trade between two cryptocurrencies. As a liquidity provider, you lock up two cryptos in return for a cut of the fees it takes from traders; your cut is proportional to how much liquidity you contributed. Those pools expose an interface to the Eth network that lets anyone put one of the cryptos in and take the other out, with a pre-determined formula for how it figures the exchange rate. Its profitability comes from the extent to which this exchange (after accounting for both pool fees and Eth network fees) gives a better deal to traders than their other alternatives (like centralized exchanges).

Like any market maker, you face the risk of "impermanent loss" from the shift in relative value of the cryptos, as market makers make standing buy/sell offers which look stupid when the market moves against them. There are also setup fees and all the usual risks associated with smartcontracts. When you consider how long you have to leave them in the pool to make a profit, and those fixed setup fees, plus the opportunity costs of lending through other providers with less volatility[4], the returns (3-100%) just about barely compensate you for the risk.

I experimented with them starting about three months ago and made a presentation, for which you're welcome to see the slides (slide 7 summarizes the downsides):

https://docs.google.com/presentation/d/1BrMMbL5vOzdnkaPVj5mO...

Contra bhouston's claim [3], these LPs don't require ETH to perpetually gain value: as long as traders continue to use the pool, ETH could stay stagnant or even fall significantly and they will still pay a return, though a long-term fall would induce a big impermanent loss (in an ETH-stablecoin pool) that would take a while for fees to compensate you for. (Edit: Although I suppose you could argue that people won't keep trading between ETH and other tokens unless that speculation, in the aggregate, is merited -- but it's not a simple matter of the pools only being profitable as some kind of derivative of ETH's value.)

[1] In the case of Celsius, largely because of this: https://prohashing.com/guides/earning-interest-on-crypto

[2] I've mainly worked with Uniswap v3; some of the details may vary slightly in other protocols.

[3] https://news.ycombinator.com/item?id=29498814

[4] For example, Gemini will lend out your GRT tokens at 6.4%, so that's your minimal return for this to be a good idea.

[adam_arthur]

Even though these are "fees" or "loans", the purpose is speculation. The whole web of connections in the flow of the money is rooted by people speculating to make money.

These aren't loans that are made to acquire some intrinsic value, like buying a house, investing into growing a business and so on. People use this liquidity or take crypto loans to try to make more money elsewhere in crypto.

Is that right? Is there a tie to the real world in there?

To me, this appears quite unsustainable and prone to failure if a risk off event comes. How would crypto have performed during the GFC? It's fallen 80% before even in times where economy has been perfectly fine. The whole ecosystem is ripe to implode due to 0 intrinsic value to owning the tokens that would otherwise cushion a fall in value.

Even in a severe recession, cashflowing businesses have value, rental properties have value. Crypto has none that I can tell.

[SilasX]

Wait, what? The entire purpose of my previous comment was to clarify the role of liquidity pools, part of something in another reply that you asked about, and you are asking about wholly unrelated things I have no expertise about. Can you at least comment on whether that (IMHO honest and thorough) attempt to explain LPs addressed anything you asked about?

>Even though these are "fees" or "loans", the purpose is speculation. The whole web of connections in the flow of the money is rooted by people speculating to make money.

LPs don't have "fees", they just have regular fees, no need for scare quotes.

>These aren't loans that are made to acquire some intrinsic value, like buying a house, investing into growing a business and so on. People use this liquidity or take crypto loans to try to make more money elsewhere in crypto.

I don't have an special expertise on what the crypto loans on Gemini/Celsius are for, beyond what their literature says. However, everything you've said there applies just as well to (secured) margin loans that brokerages make. Do you have the same objections to those?

[notahacker]

There may well be many ways to get ROIs in excess of 12.68%. There are also many ways to obtain investment capital on better terms than a fixed 12.68% APR, especially if you're really, really good at investing. And if you want as much USDC as possible it's even odder to lend some of it back out at 1%!

I guess the charitable explanation for them raising capital as expensively as possible from randoms on the internet is they're feeling charitable themselves. I seem to remember that was how all the pre-crypto HYIP "investments" explained the generosity of their compounding daily return offers

[marcinjachymiak]

There are also other ways to get these returns outside of DeFi, like by staking ETH, selling covered calls, or various other strategies on centralized services

[yellow_postit]

And these are also mostly speculation.

The crux of the OP’s argument to me is that The company and others in this space position the yields as being “in kind” to savings account yields despite it very much being a different risk ballgame. See the frequent hacks on https://rekt.news/ for example.

Do sites like Celsius explicitly claim to be as safe as insured savings banks? No of course not but they play a similar linguistic game of association to Tesla’s “autopilot” nomenclature.

As with most things do your own research and don’t risk what you can’t lose.

[gruez]

>like by staking ETH

What's the ROI on staking ETH right now? Is it anywhere near the rate that they claim?

>selling covered calls

you're taking on risk when doing that. It works well until you get assigned, in which case you take a massive loss and unable to pay back your investors.

[quickthrowman]

How do you take a “massive loss” on covered calls being exercised? They have defined risk, you give up some potential upside to collect premium. If they are exercised, the underlying is called away.

Selling naked calls has a lot more risk, see the blowup of optionsellers.com on naked natural gas options

[gruez]

>How do you take a “massive loss” on covered calls being exercised?

The loss comes from the opportunity cost of what you could have sold the underlying asset for. That might or might not be "massive", but the potential is definitely there. At the end of the day you're picking up pennies in front of a steamroller. Maybe it even has expected value greater or equal to the advertised APY, but failing to disclose that and/or pretending like it's guaranteed is deceptive. I'd be pissed if I put my money into some sort of "savings account" with 8% APY, then get wiped out next time there's a spike/dip, because it turned out that they were using it to write covered calls.

[thr0wawayf00]

Agreed. And once the irrational optimism leaves the market, we'll see just how stable this whole system is. It mirrors the irrational, misplaced confidence in the housing market back in 2007.

Tether feels like the most likely POF based purely on the attention they're getting now, but nobody knows exactly how it will bust.

[dustintrex]

> Well a lot of it is coming from incentives of these protocols

I keep seeing this for DeFi.

- So you put 1 fiat into the magic box, and 2 fiat comes out. Where did the 1 fiat profit come from?

- DeFi collateralized staking algorithms hash protocol incentives!

- Yeah whatever but no really, it's a closed system so all inputs and outputs need to sum up, where did the new fiat come from?

- YoU dOnT uNdErStAnD cRYpTo!!1!!

[berberous]

I’m not really knowledgeable about DeFi, but here’s a simplistic model (would love to be corrected by someone with deep DeFi knowledge):

1. You put $1 of ETH into a new protocol, let’s say a new Decentralized Exchange on Ethereum that is paying a high amount of interest in order to attract liquidity to their protocol. The high interest is perhaps paid in ETH or maybe in a new token for the protocol.

2. Users flock to this new DEX, and actually use it, generating trading fees for the DEX, which drains activity from CEXs like Coinbase. If in step 1, the payment was made with a new token, perhaps that new token either earns a cut of trading fees, or gets governance rights over the DEX.

3. You either earned another $1 of ETH, or of the new token.

In the above example, it’s not a closed system anymore than the US economy is. A new company was created, which created real value by creating a better exchange which attracted users over CEXs. The company has value now it can payout because it generates trading fees and the equity/governance of the company is valuable. It bootstrapped that with protocol incentives.

[bglusman]

I think you may have just also described fractional reserve banking? that is literally where much of the money supply comes from.

https://en.wikipedia.org/wiki/Fractional-reserve_banking

[Jeema101]

Except that they aren't banks, and so there's no lender of last resort to backstop them. [1] There can't be any assurance that depositors would get all of their money back in the event of a bank run if the money isn't backed by actual dollars or a lender of last resort.

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

[bglusman]

Well… I’m not expert on DeFi or anything but until the 20th century banks existed and didn’t have this right? But yes, there were runs on banks… but there are no runs on CDs right, and that’s a banking product also? Why not? Well, they’re contractually wrapped up for a time period. What if all DeFi lending is/was similar and crypto contracts locked them up for set periods… would that be an OK and legitimate system then? If not why not?

[notahacker]

Except of course, that the crypto world insists that all these cryptoinvestors borrowing crypto to buy other crypto to sell for more crypto to repay their original crypto debt is an ecosystem which isn't [even more] dependent on inflation of the crypto supply.

That and unlike stablecoins the Fed doesn't pretend it's fully backed.

[gzer0]

> Where do these high APYs come from? Well a lot of it is coming from incentives of these protocols and speculation in those native assets. But the author doesn't even know that, so I won't bother steelmanning his argument

Is this not the very definition of a ponzi scheme? The returns coming from "incentives" of these protocols and "speculation" in those native assets sounds very ponzi-like to me.

[roywiggins]

If the yields are actually coming from speculation, it's not a Ponzi. Ponzis don't generate real yields, they just shuffle money from new investors to old investors. If it's actually making risky bets and winning them it's more like a hedge fund or something.

Of course hedge funds don't have steady 10% yields, they lose money when their risky bets don't pan out.

[adam_arthur]

That's essentially exactly what a Ponzi scheme is. You have an asset with 0 intrinsic value that only goes up in value due to new investors buying in.

You will only ever make more money if somebody buys in after you.

Unlike stocks where owning a share entitles you to profit stream of the company (intrinsic value)

[roywiggins]

If they are actually lending the money out to people who aren't investors, earning interest on it and returning that to investors, then Celcius is not a Ponzi scheme, it's probably speculating on weird, opaque and shitty assets, but it's not a Ponzi.

Like, if I set up a business "investing" in worthless penny stocks and somehow manage to generate returns, I'm not running a Ponzi. Maybe I'm pumping and dumping those penny stocks. It's still a scheme, it's just not a Ponzi- I am actually earning money for my investors! I'm defrauding other people, but I'm not defrauding my investors.

[adam_arthur]

Pump and dumping penny stocks works exactly like a Ponzi scheme. It's a layer removed perhaps, but the early buyers win and the later buyers lose, by definition.

I'm not an expert in Celsius, perhaps it's structured a bit differently. But any system that relies on new investors paying out the old is structured akin to a Ponzi scheme, if there's no intrinsic value element to the speculation.

If AAPL pays a 10% dividend, huge numbers of people would buy it for the yield. The yield comes from their underlying business, not from investor money. This puts a floor on the price.

When your yield only comes from newer investors, that's not intrinsic value or sustainable

[roywiggins]

Not every financial fraud is a Ponzi scheme, and it's silly to call everything a Ponzi. It sounds like Celsius probably is one, though.

[FanaHOVA]

Yields for nascent crypto projects are basically marketing budgets and user acquisition costs passed on to the user as profits instead of going to Google Ads or being a coupon for free fries with your order. They want to incentivize people to bring liquidity to them, and pay them for it. The real question is whether or not this actually leads to retention or of it's wasted capital, but that's beyond the point of the article.

[JaimeThompson]

"Its flagship product: 10 to 12.68% annual returns on USD stable coins and this with little to no risks."

What is more likely that they figured out a way to "hack" the financial system to make such returns in a very safe fashion or that they are doing things behind the scenes which may be a little bit sketchy?

10% safe annual returns is a bit high if one thinks that the APY is reflection of how risky the investment is.

[Alex3917]

> Where do these high APYs come from

The promise of DeFi is that it's supposed to reduce risk for lenders by letting them see the assets of borrowers in real time. If the risk to lenders is lower but interest rates are higher then how can it not be a Ponzi scheme? There's literally no other possible explanation.

[reginold]

I would be interested in seeing other articles on this topic.

As someone new to DeFi, I found the article sparked new questions, even though as you said it wasn't deeply researched.

If Celcius is basically reinvesting into a bunch of other assets, it's possible, or even likely, that the high yield assets are too good to be true. That's where my concern would be. Is BadgerDAO rock solid? Compound? I would be curious to see expert analysis of these. Celcius is simply built on top.

[popol12]

This. Plus, if you deposit Eth on Aave, it's not for the 1% returns but because it then allows you to borrow other coins which you can use somewhere else in DeFi, where it can bring you much higher returns (convex/curve for instance). I'm amazed that this guy wrote such a big article on a topic he obviously doesn't fully understand.

[roywiggins]

The problem is that every Ponzi insists that their critics just don't understand their special sauce that generates their phony returns.

[max48]

The problem is also that people are using the word "ponzi" to describe every investment they think is bad and/or don't understand. offering high rates doesn't automatically means it's a ponzi. With the definition that some of you are using here, literally every single bank, currency, edge fund and all publicly traded companies would be a ponzi.

[kangnkodos]

I think it's a good rule of thumb to assume that anyone who guarantees returns greater than the historical return of the S&P 500 is a ponzi scheme.

I don't know of a single reputable bank, hedge fund or publicly traded company that does that.

A huge number of places do have high returns from time to time. But any reputable firm points out that past returns do not guarantee future results. No reputable firm guarantees returns that high in the future.

The key is the guarantee.

Anyone who GUARANTEES returns that high in the future is suspect, and therefore must be more transparent than usual in order to clear the bar. The article indicates that Celsius was less transparent than usual in describing exactly how they made their high returns.

[roywiggins]

Celsius does say "While Celsius strives to maintain stable reward rates over time, any change in circumstances may bring about changes to such rates, and in some events the rates may drop to 0%" deep in their Risk Disclosure.

https://celsius.network/static/risk-disclosure.pdf

Which is a bit odd- there's no chance you might lose money? How are they making money without risking that the return might drop below 0?