In defence of DeFi

It’s been a mad year.

Although, I suppose every year in history has been a mad year, depending on where you live and what your interests are.

But this year, in particular, it’s been a mad year in crypto.

Right now trust in crypto is at an all-time low. Which is fair, given everything that’s happened over the last 12 months.

However, as we’ll see today, it wasn’t really the “crypto” part that went wrong.

Most of the bankruptcies and scandals this year came about because of plain old fraud and deception.

In this issue, we’re going to look at what’s caused the wave of crypto company bankruptcies, and examine how using decentralised finance (DeFi) could have avoided all of them.

Well, most of them.

The Terra Luna fiasco was caused by badly designed DeFi.

But we’ll get to that in due time.

So let’s get started…

Why DeFi?

The idea of DeFi simple.

You take the world of traditional finance and you replace the centralised authorities with decentralised alternatives built on blockchains.

(Or on distributed ledger technology (DLT) if you want to be accurate, as many crypto projects aren’t actually blockchains, but they are all DLTs.)

What’s the advantage of this?

There are many. But the main ones are fairness, transparency, security, efficiency, cost-reduction and “sticking it to the man.”

(You can read more about this in my everything you need to know about crypto essay.)

At least that’s the idea.

But after the Terra Luna fiasco, DeFi has lost much of its glitter.

However, given the ongoing FTX revelations, and centralised crypto lender collapses – like Celsius, Voyager, BlockFi, Genesis et al – DeFi’s case has been substantially strengthened.

Now instead of just stating that as fact, let’s explore why it’s true.

One of the main benefits of DeFi is that it uses over-collateralised loans and relies on smart contracts for instant liquidation.

This (in theory) makes everything less risky, including rehypothecation.

Now, if you don’t recognise the terms, “smart contracts”, “over-collateralised” and especially “rehypothecation”, don’t worry.

(And if you do, maybe you should get out more. Haha. Actually, maybe I should get out more.)

Let’s start with the simplest to explain: smart contracts.

These are simply computer programs that run automatically when certain conditions are met.

So, in the case of a loan, a smart contract can be programmed to automatically liquidate when the loan’s over-collateralisation ratio falls below a certain level.

Simple.

Ah, but what’s an over-collateralisation ratio?

Collateralisation means using collateral to secure a loan. For example, a mortgage is a collateralised loan. Your house is your collateral.

If you can’t pay back the loan the bank takes your house.

An over-collateralised loan means you put up more collateral than you borrow.

So, if you want to borrow £100, you might put up £120 worth of Ethereum as collateral.

(This would give you an over-collateralisation ratio of 120%... £120 is 120% of £100.)

Why would you ever want to take out an over-collateralised loan?

Well, you might believe that your £120 of Ethereum will be worth double that in a year.

So you take your £100 (which you just got by taking out your loan) and buy £100 of Ethereum with it.

Let’s assume you were right (wouldn’t that be nice?).

One year later, Ethereum’s price has doubled.

So you can now sell £100 worth of the £200 of Ethereum you have and use it to repay your original loan.

(Remember, your loan was £100. Your collateral to secure the loan was £120. So you only need to pay back the £100 loan amount to get back your collateral.)

You get back your £120 of Ethereum (now worth £240) and you also still have £100 of the £100 of Ethereum you bought with your loan (because that £100 of Ethereum had also doubled, to £200).

So you now have £240 + £100 = £340 of Ethereum.

Whereas, if you’d not taken out the loan and just held onto your original £120 of Ethereum, you’d now have £240 of Ethereum.

(Remember, Ethereum has doubled in price.)

So you’ve just made £100 more than you would have by simply buying and holding:  £340 - £240 = £100.

Although, in practice you’ll have paid interest on your loan.

So let’s say the interest on your loan was 10% APR. In that case, you’d have needed to use £110 of your £200 of Ethereum instead of £100 to get back your original collateral.

So, let’s say you made £90.

(This is basically how margin trading works. But that’s a whole other topic.)

In this scenario everyone is happy. You just made £90 from nothing and the person you lent the Ethereum to made £10 for doing nothing.

This tactic goes downhill fast when prices start to crash

Ah, but what if Ethereum fell in price over that year, as it just has done.

Let’s say Ethereum starts to fall.

Let’s say your agreed collateralisation ratio is 108% of your loan.

So if Ethereum’s price falls 10%, your collateral – which is currently worth 120% of your £100 loan – will only be worth 108% of your loan and you will be liquidated.

In a DeFi platform, this liquidation happens automatically. It’s carried out by a smart contract built into the DeFi platform you’re using as we already discussed.

In traditional finance, you can ring up your broker and plead with them not to liquidate you. Or you can say, you’ll add more collateral on Tuesday.

In DeFi, you can’t call up anyone, and you can’t plead with a smart contract.

So if you’re using a DeFi lending platform, as you see Ethereum’s price dropping, you have three options.

1. You could keep adding more Ethereum as collateral so that your loan stayed above its agreed collateralisation ratio.
2. You could pay off your loan, plus any interest owed before it hits its liquidation level.
3. You could do nothing and get liquidated when Ethereum falls 10%.

If you choose – or are forced to take, because you’re out of money – option three, the person you took the loan from gets to keep enough of your collateral to pay off your loan and you get whatever is left back.

But there will also be a penalty you’ll have to pay out of that collateral for getting liquidated.

Let’s keep things super simple and say that liquidation fee was 8% of your loan.

So in this scenario you get nothing back. Your £120 of Ethereum has dropped in price to £108, forcing a liquidation.

£100 of that is used to pay off your loan and £8 is paid to the lender as a liquidation fee.

In this case you are very unhappy.

The person you took the loan from is still relatively happy as they still made £8 profit.

But you lost £120 of Ethereum to gain £100 of Ethereum. So you lost £20.

But it’s actually worse than that.

Because Ethereum is now worth 10% less.

So you lost £120 of Ethereum and gained £90 of Ethereum.

So you really lost £30.

If you’d just held onto your £120 of Ethereum and not taken out that blasted loan, you’d now have £108 of Ethereum.  So you’d only have lost £12.

But it’s actually even worse than that.

Because of how our brains work, it really feels like you lost far more than £30. It feels like you lost the whole £120 of Ethereum you put in.

This is why margin trading makes people unhappy.

Oh, actually, it gets even worse.

Again because of the way our stupid brains work, you feel the pain of a loss much more strongly than you feel the joy of a gain. Roughly twice as strongly, which leads to all kinds of bad behaviour.

Okay, what about rehypothecation?

Rehypothecation is when someone takes that £120 of Ethereum you put up as collateral and uses it for collateral for something else.

This is standard practice in a lot of finance, and it allows the lender to make much more money from your loan than just the interest you pay on it. Which in turn allows them to offer you better rates.

Win-win, right?

The thing is, it can also be disastrous, especially when prices drop.

(Side note: rehypothecation of mortgages is basically what caused the 2008 financial crisis.)

In many DeFi platforms, you can actually rehypothecate your own collateral.

But – IN THEORY – it’s far less dangerous than rehypothecation in traditional finance because of the smart contracts and instant liquidations factors.

Phew!

What does all of this have to do with the slew of crypto company bankruptcies we’ve seen this year?

Glad you asked…

All of these big centralised crypto lenders were not over-collateralising their loans.

In most cases they made out like they were, but in the small print they usually admitted that they didn’t over-collateralise all of their loans.

I wrote an article in mid-2021 about the promise of a sustainable 10% yield and what that might mean for people’s savings, the stock market and life in general.

At the time you could get a staggering 12% yield on stablecoins through the DeFi platform AAVE.

However, as I pointed out:

Right now there are three big issues with DeFi for the average person.

1. It is complicated …

2. It’s incredibly risky.

In theory, if it all works as it’s supposed to, DeFi can be low risk. But, in reality, it doesn’t pan out that way.

Over the last few months there have been myriad DeFi projects that have lost people money through accidental or deliberate bugs in their code.

The deliberate kind is usually called a “rug pull”, which is a phrase you’ve probably seen popping up a lot in the cryptosphere recently. …

3. It’s expensive.

DeFi has become so popular that it’s clogging up the Ethereum network – which is the network most of DeFi uses.

And that means the fees you need to pay to use a DeFi platform like AAVE can run into hundreds of dollars.

So, unless you’ve got hundreds of thousands of dollars to play with, the interest rates aren’t really worth it.

What’s the point of converting £1,000 into a stablecoin to earn 10% interest if it’s going to cost you £100 to start earning that interest in the first place?

Maybe that would be okay if you knew that 10% rate would last for years, but it won’t. Which brings me to…

4. Rates fluctuate constantly.

The reason why I said, “right now” in my AAVE example is because by the time you read this, that rate will have changed.

Remember, those rates are being managed in real-time algorithms weighing up supply and demand for each particular pool of crypto. And they fluctuate like crazy.

One day, you might be able to get 14% on USDC. The next day the rate on USDC might have gone down to 5%, but the rate on GUSD us now up to 11%.

So you’d want to pull your USDC out of the pool, trade it for GUSD and then put it into the GUSD pool to get the best rate.

Fair enough. But remember, pulling your money out of a pool could cost you $50-$100 in gas fees. And the same would go for putting it back into a new pool.

Which is why, currently, DeFi only really works if you’ve got hundreds of thousands of dollars to play with. Otherwise, you’re going to end up losing more on fees than you make on interest.

Entrepreneurs saw what was happening in DeFi and realised they could take its idea and plug it into their own platforms that they controlled.

This way they could make the experience much easier for users, attract more customers and make a ton of money for themselves.

I called these platforms centralised decentralised finance (CeDiFi) because they were basically taking the backend of DeFi and plugging it into a platform they centrally controlled.

By doing this they could make things:

• Easy for users.
• Simple for users.
• Cheap for users.
• And they could have less fluctuation in their interest rates.

However, this ease of use came at the cost of transparency.

As I said at the time these platforms were in theory a win-win… “so long as you’re okay trusting an institution with your crypto”.

The problem with these CeDeFi platforms was the risk.

Again, from my article:

With DeFi platforms your main risk is a bug in the code and/or the developers deliberately scamming you.

From what I’ve seen, with CeDeFi bugs in the code tend to be less common, as the platforms are usually built and backed by multimillion, or even multibillion, dollar companies. But it’s still a risk.

However, the main risk here is the centralised nature of the platform. If the company running it goes out of business, it could take your money with it.

For example, in November 2020, a CeDeFi platform called Cred went bankrupt and took around $140 million of depositors’ money with it.

Or they could be hacked.

Or they might not be over-collateralising the loans and increasing the risk of defaults.

Or they might get sued by regulators and forced to freeze all the assets they hold. Remember, this is a new industry and regulations are changing all the time.

I see all of these risks as potentially major.

It turned out one of those risks was, indeed, major.

These CeDeFi companies were not over-collateralising their loans and, in the end, that’s what killed them.

I discovered that at the time, but it took a lot of digging.

Here’s what I wrote about the main risks of (now bankrupt) Celsius in that article:

Other than the general CeDeFi risks that I covered earlier, I think Celsius’ main risk is that it doesn’t over-collateralise all its loans.

If you scour its terms of use (I told you I did a ridiculous amount of research into this, didn’t I?) you can find the following:

“(v) Celsius may lend your coins to exchanges, hedge and other counterparties, which may provide full or partial collateral for any coin or fiat loan.”

See that? “Or partial collateral”.

Well, I did some digging and it turns out that in some cases they do not over-collateralise their loans.

I even emailed them to doublecheck. After a week or so (remember, slow customer service as one of the negatives) I got a canned response about the benefits of Celsius, with this paragraph at the end (emphasis mine):

“To ensure coin loans are always returned to Celsius, we require borrowers to post collateral of up to 150% (which means that the borrower gives Celsius an alternative asset as collateral for the asset they are borrowing) or we conduct thorough due diligence reviews of borrower’s financials and repayment ability.”

So it’s true. They don’t over-collateralise all their loans. This is a pretty big risk in my book.

I also want to point out that while its crypto in cold storage is insured, the crypto you deposit and earn interest on isn’t.

A lot of CeDeFi platforms play this “insurance” card as a benefit. Actually, I think they all do. But it never really means anything because your deposit isn’t insured.

As you can see, it was very difficult to get confirmation that they were not over-collateralising their loans.

And I only did this digging because I already knew a lot about how DeFi worked and why it worked – and the risks.

Celsius wanted customers to think it was over-collateralising all its loans, so it was safe from defaults.

But it wasn’t… hence why it collapsed after Three Arrows Capital – which it had given a $75 million uncollateralised loan to – defaulted on its debt.

BlockFi, which is now also bankrupt, was doing the same thing.

Again, from my article:

And just like Celsius, it [BlockFi] also doesn’t over-collateralise every loan.

From its “how does BlockFi custody assets?” page (emphasis mine):

“BlockFi requires most borrowers to post varying degrees of collateral for its lending activity dependent on the borrower’s credit profile and the size of the loan.”

So it appears that just like seemingly all other CeDeFi platforms, it will let certain “trustworthy” borrowers take out loans without over-collateralising their position.

In fact, of the three CeDeFi platforms I covered in that article, only one – Nexo – was explicitly clear that it over-collateralised every single loan.

It’s no coincidence that it’s also the only one left standing today. However, Nexo has many of its own problems. So let’s not hold it in too high esteem.

Basically all of these bankruptcies were triggered by the Terra Luna fiasco, which caused Three Arrows Capital (crypto’s biggest hedge fund at the time) to collapse and default on its many millions of dollars of loans, which bankrupt all the CeDeFi platforms that had given it those uncollateralised loans.

And because those CdDeFi platforms were companies, not DeFi protocols, when they went bankrupt, they took all their users’ assets with them.

(Turns out that insurance was useless after all.)

It’s also worth noting that despite the chaos, the big, established DeFi platforms like Maker, Uniswap, AAVE and Compound are still going strong.

And that’s despite many attempts by hackers to ruin them over the last couple of years.

Although with the same breath we have to also acknowledge this whole mess was caused by the Terra Luna debacle, which was a DeFi platform itself.

And FTX?

You can sort of forgive the CeDeFi platforms for what happened – actually, you can’t. They made it way too hard to get to the truth about how they issued their loans.

But at least they didn’t outright lie about it.

And you can sort of forgive the Terra Luna fiasco. At least that was ultimately a failure of tokenomics. It worked great until it got stress tested too hard. But everyone who looked into it knew that was a distinct possibility.

As for Three Arrows Capital – I mean, it was a hedge fund. It should be no shock that it collapsed. Hedge funds collapse all the time. It’s a risky business.

But then that takes us to FTX.

FTX wasn’t just giving uncollateralised loans to a hedge fund (Alameda Research). But that hedge fund was under its own control.

And the money for those loans was coming out of customers deposits.

Not customers who put their money into a lending program. Customers who were just using FTX as a place to do normal trades and store their funds.

Again. This would be kind of forgivable if FTX told customers it was going to do this. But it did the opposite.

It expressly stated in its terms of service:

You control the Digital Assets held in your Account.

Title to your Digital Assets shall at all times remain with you and shall not transfer to FTX Trading.

None of the Digital Assets in your Account are the property of, or shall or may be loaned to, FTX Trading; FTX Trading does not represent or treat Digital Assets in User’s Accounts as belonging to FTX Trading.

So FTX customers had no idea their fund were being stolen and given to Alameda Research.

Then when Alameda started losing money, it couldn’t pay back its loans to FTX.

Which meant FTX couldn’t give customers back their funds.

And when rumours of this started to surface, many customers tried to get their funds back… which didn’t go so well.

And the rest is history.

So in conclusion, why DeFi?

Well, because of everything that’s happened in 2022.

The FTX scandal could not have happened on a (well built) decentralised exchange or DeFi platform.

Why?

Because they are decentralised, which means they are transparent.

Anyone can look into their code and verify if they are doing what they claim to be doing.

And that’s not to say DeFi doesn’t have its own set of problems: bugs in the code, tokenomics, network congestion, rug pulls etc.

But fraud, deceit and opacity can be virtually eliminated… so long as people are actively checking the platform’s coding.

If FTX was DeFi people would have been able to look into it and see it was using customer deposits for its own purposes.

In fact, that wouldn’t even have been an issue, because the platform could have been coded in a way that made that impossible. And people could have verified that code for themselves.

In traditional finance, something like that would be impossible. But it’s standard practice in DeFi.

In traditional finance ensuring companies are doing what they say they are doing is the job of auditors, regulators and accountants… but all of those entities can also be corrupt or incompetent. Look at Enron, look at 2008… look at FTX.

With DeFi, anyone, anywhere in the world can do the auditor’s job because the code is open source and hosted on public DLTs.

The problem with DeFi is it only really works well within one chain. So DeFi on Ethereum has problems linking with DeFi on, say, Cardano.

This leads to congestion and incredibly high fees.

There are two schools of thought on how to overcome this problem:

1. Increase scalability.

If Ethereum could process 100,000 or 1 million smart contract transactions per second, then congestion and fees wouldn’t be an issue.

Although then everyone would have to use Ethereum and other projects would fade away. Some people would see this as a good thing. Some would not.

This is the approach “super fast” projects like Aptos, Radix, Solana and Algorand take.

(Those links will take you to my deep dives on those projects.)

2. Make all DLTs interoperable with each other.

This approach means every DeFi platform could have its own chain. And since that chain was only handling one (or a few) DeFi platforms that chain wouldn’t get too congested.

I guess you could say this is another way of increasing scalability. But the key point here is all the different chains would be able to seamlessly interreact with each other.

This is the approach Polkadot and Cosmos take.

(That link will take you to my deep dive on Polkadot.)

And so I guess it’s fitting that my deep dive in this month’s premium issue will be Cosmos.

As I wrote in last month’s deep dive into Aptos:

Cosmos has pretty much the opposite vision to Aptos. Like Polkadot, it sees a multi-chain world where different projects interact with each other.

And it’s already been used to launch some of the biggest projects in crypto. Terra Luna (RIP) for example was built using the Cosmos software development kit.

Here’s how Cosmos describes itself:

“Cosmos is the internet of blockchains, an ecosystem of apps and services that exchange digital assets and data using IBC (Inter-Blockchain Communication) protocol.

Cosmos envisions that interoperable blockchain technology make the world economy more resilient through decentralization, more accountable through transparency, and more efficient through programmable value. The future economy consists of over a million interconnected blockchains used to exchange digitized real-world assets and provably scarce digital assets.”

If you’re a premium subscriber, you’ll get that deep dive on Sunday the 8th of January.

And if you’re not, you can become one here. If you do, you’ll also get access to every deep dive I’ve published since I launched this site.

Okay, that’s all for today.

Thanks for reading.

Harry

Full disclosure: At time of writing, I held the following cryptos: Ethereum, IOTA, Radix, Mina Protocol, Aleph Zero.

Disclaimer: This content does not constitute financial advice, tax advice or legal advice. Your money and how you choose to spend it is your responsibility. Nothing that appears here should be construed as investment advice or recommendations to buy or sell any securities, cryptos or investments. coin confidential does not offer investment advice. We merely provide information. Crypto investing is highly risky. You should not base any investment decision solely on information we publish. We believe all information we publish to be accurate, but we cannot guarantee it. Always do your own research before making any decisions about your money. See the full disclaimer for more.