Over the last two weeks, I’ve spent quite a bit of time working through what went wrong, what we could have done, and what we can do to mitigate this from happening again.
Crypto is the definition of a high-risk asset that we believe provides potential for high returns. As investors, we weigh probabilities to make what we believe to be the best risk-adjusted decisions we can using the information available to us at the time. While our investment in Terra (LUNA) did not work out, I've identified a few places where our research went wrong and ways we can attempt to avoid this type of outcome going forward.
If you’re looking for a detailed play-by-play of what happened or a quick recap, this admittedly isn’t it; please excuse my lack of brevity. This situation warrants some detailed thoughts, and you deserve clarity on what happened with our investment and why. Let’s dive in.
Part of our due diligence on crypto-assets involves evaluating network activity/growth, quality of its ecosystem, developer engagement, and investor support – all of which Terra passed with flying colors.
Unlike other Layer 1 blockchains that need to drive on-chain activity to grow, Terra was different because it was in the business of stablecoins. UST was available on almost all major blockchains, including Ethereum, Solana, Polygon, BSC, and more.
We were initially intrigued by Terra’s real-world use cases, particularly in the use case of their blockchain-native payment rail. The crypto world first learned of Terra through CHAI, a South Korean mobile payments app that processed over $1 billion in transactions per year for 2.5 million users. Chai alone made Terra the #1 blockchain in terms of real-world utility.
By design, LUNA initially acted as a defensive, counter-cyclical asset. In bear cycles, demand for UST grows as investors sell their risky crypto assets for stablecoins. As demand for UST increases, more LUNA is burned, which reduces supply and drives the price higher.
On January 22nd, we sized our LUNA position down from a ~21% weight to a 4% weight in the crypto portfolio. Why?
We were not particularly fond of Anchor Protocol’s close partnership with Abracadabra Money, the latter of which follows a leveraged stablecoin yield farming strategy coined the “Degenbox.” Leverage and crypto rarely go well together.
Anchor Protocol’s yield reserve was also depleting rapidly as a new wave of deposits outpaced borrowings.
We did not exit our LUNA position entirely for several reasons:
1.) The Terra team was aware of the issue with the reserve, and there are many potential solutions to reverse the course.
2.) Even if the yield reserve were to deplete, the natural APY that Anchor earns from staking and borrowing would still be very respectable at ~10-15%, providing some of the highest returns amongst stablecoins.
3.) Outside of Anchor, the Terra blockchain was thriving and evolving into a vibrant, multi-faceted ecosystem.
Then, we modestly increased our LUNA position on April 1st for several reasons:
1.) The degenbox strategy had wound down.
2.) Terra injected 450M UST into Anchor’s reserve on Feb 17th.
3.) The team also closed a $1 billion private token sale to bolster its UST reserve. The sale was led by Jump Crypto & Three Arrows Capital, with participation from DeFiance Capital, Republic Capital, GSR, Tribe Capital, and many others.
4.) While using capital to buy other cryptos such as BTC and AVAX are not as ideal as stable assets backed by hard collateral, it was still likely a better move than if they had done nothing (i.e., have 100% of reserve in LUNA and UST).
5.) LUNA remained one of the best-performing assets during bearish market conditions, in line with our original thesis.
6.) Terra and its ecosystem outperformed other ecosystems in several growth metrics.
UST grew too big and too fast, and the prioritization of unsustainable growth over real-world use ultimately proved its downfall.
As the Anchor protocol attracted more deposits, it also created significant systematic risk for the Terra protocol. As risk assets and crypto markets sold off during April amidst a bearish global macro backdrop, the ratio of LUNA to UST market cap quickly declined. This depleted capital from more productive projects on Terra during a time when the market was de-risking.
Productive UST deposits put to real-world use (e.g. to provide financing for businesses/projects) would have been more “sticky” when something goes wrong – but unproductive UST deposits on the Anchor protocol, by contrast, would run for the exits at the first sign of trouble.
We underappreciated the technical risks and stress scenarios that would be imposed on the systems during periods of high volatility. The Terra blockchain encountered severe network congestion and rising gas fees during high-volume periods.
To sufficiently absorb LUNA volatility on-chain and prevent UST from depegging, the protocol needed to keep raising the size of the virtual pool of the market module. However, the Terra team was slow to act as they only decided to increase the pool’s size on May 10th, three days after the first de-peg. At this point, things were so slow and so congested that exchanges began pausing withdrawals.
These factors created a perfect storm of panic in the market. This level of panic would ultimately de-peg UST beyond recovery.
At Titan, we have a process in place to reduce confirmation bias as much as possible, but unfortunately, our processes fell through this time. Our bias was informed by past crashes in which UST recovered – the May 2021 flash crash and the degenbox debacle in February 2022 are two examples.
So where do we go from here? Moving forward, we will implement an even more rigorous process to 1) identify and stress-test the tail risk of an asset (i.e. how can this go very, very badly), and 2) define the appropriate quantifiable cut-off thresholds on these metrics beyond which the tail risk evolves into a concrete risk.
During harsh global macroeconomic conditions or periods of extreme volatility, it pays to shoot first and ask questions later, at least in the case of “prove-it” scenarios such as the Terra crash. It’s better to cut losses early and re-enter later (if needed) when the team/project/plan/thesis has “proven” to work in our favor.
No crypto is too big to fail. Terra was a top-five market cap Layer 1 ecosystem with hundreds of applications and a vibrant community built on top of it. Adding more legos on top of a weak structure does not make the base stronger. It only amplifies the damage.
Reputable VCs like Polychain, Jump, Galaxy Digital, Three Arrows Capital, and Delphi all endorsed and backed Terra. The head of Galaxy got a LUNA tattoo. They wouldn’t let it fail, right? Wrong. We all suffer the same biases.
If you made it this far, kudos. Post-mortems are helpful for both clients and investors – our team learns from our mistakes, and you, our clients, receive the kind of transparency we have promised from day one.
We appreciate your trust in our ever-evolving process, and please let us know if you have any other questions.
Best, Gritt and the Titan team
Learn with titan
Become the smartest investor you've ever been through straightforward, easy-to-read investment articles.