99.999% Loss and the [Mark] Cuban Missile Crisis
A review of the IRON Finance Collapse and how it got us
Can you imagine losing 99.99999% of an investment within 6 hours? The [Mark] Cuban Missile Crisis they're calling it. Yes, that Mark, from Shark Tank, the Mavericks, etc.
Mark Cuban is actually insanely technical, he's deep in decentralized finance.
I saw an episode of shark tank where he grilled a startup for abusing machine learning buzzwords to try and get ahead.
But this one got past him (and me 🙈🙈)
On June 13th he wrote a blog post on one of his liquidity pool positions. He put up $75,000+ between 2 crypto assets: TITAN and QUICK.
The idea is that people who want to trade one for the other, use his pile of both to make the switch and in exchange they give him 0.25%.
Unfortunately, TITAN had a problem. Let’s start with talking about collateralization.
Imagine you run a pawn shop: I give you my $100 watch, you lend me $100, I return the $100 (+ a fee), I get my $100 watch back. This is 1:1 collateral.
In the real world, semi-collateralization (less than 1:1) is very common.
The US banking system only keeps 10% of loans in reserves (the reserve ratio).
The way it works is:
Required Collateral = (1 - Trust Ratio) * Debt
Where the trust ratio is essentially “How much does the overall market trust my operations and long-term sustainability such that they’ll trust me with their money”.
If your pawn shop had 50% trust in me, I could give you my $50 watch (that you trust I really love and want back) in exchange for $100. Then when I pay you back $100 (+fee) I get my watch back. This is semi-collateralization.
Because US bank deposits are fully insured by the US Government, it's easy for bank trust ratios to be 0.9. Essentially 90% of the banking system is trust based. This is contentious and I’ll be writing an Off the Cuff on how banks work separately and will link it back: here.
Decentralized finance is not a very trusting place. So when MakerDAO created their Dai stablecoin, which attempts to always be worth $1 (without deposit insurance), it's trust ratio is actually *negative*. It's -0.5. Every dollar they issue, they have $1.50 in assets backing it up.
This is NOT capital efficient. To mint $1 you have to put $1.50 in a box? Well MakerDAO has got it figured out because they're a multi-billion dollar organization now!
Maybe negative trust is smarter than I realized 🤔
So back to Titan.
TITAN is the flexible collateral behind the IRON stablecoin.
IRON wanted to be like DAI, but with a trust ratio of 0.25.
75% of it's value was stored as USDC (Coinbase's 0 trust ratio dollar), while the other 25% was in their own coin, TITAN.
This works great when IRON is popular (valued >$1): TITAN grew from <$1 to $60 within weeks as people could use it to mint new IRON profitably (e.g. spending $0.75 USDC + a small amount of TITAN they paid < $0.25 for to create $1).
But when IRON loses popularity (i.e. it fails to be valued by the market at $1) more TITAN gets created to cover the difference, diluting it's price (I'm simplifying here).
The TITAN max supply was supposed to be 1 billion.
But IRON lost it's popularity so quickly that 26,999,000,000,000 more TITAN got created, destroying the value nearly instantly.
It’s hard to think about how this happened. There were always risks.
The project was a fork & pork. They copy/pasted code from a larger more well known project FRAX that is well known in the space of semi-collateralized stablecoins.
The development team didn’t understand the parameters they were changing in FRAX’s code. They created feedback loops so that TITAN could “compound” creating more of itself and more IRON when the collateral behind IRON (75% USDC, 25% TITAN) became more valuable than $1.
They heavily incentivized providing liquidity to TITAN- such that the estimated APY was several million %. Implying that $1 of TITAN could compound to be worth $100,000s within a year. People were even claiming they were making so much money they could quit their jobs over it.
The lessons for you reader are: if it’s too good to be true, it probably is; be extremely careful with trust in decentralized finance; and never gamble more than you can afford to lose. The difference between investing and gambling can be subtle and blurred in DeFi.
If you liked this off the cuff- I hope you share & subscribe!