In “On Impossible Things Before Breakfast,” NYDIG uses the Terra/ LUNA collapse as a case study. The paper aims to prove that the algorithmic stablecoin concept is flawed in nature. It also takes aim at the current state of the DeFi stack and shows how fragile it is. The subtitle says it all, “a post-mortem on Terra, a pre-mortem on DeFi, and a glimpse of the madness to come.”

Related Reading | What Terra’s Collapse Brought For Stablecoins In Japan, New Law Passed

For those living under a rock, on May 7th Terra’s algorithmic stablecoin, UST, lost its peg to the dollar. Several key withdrawals to the Anchor protocol might’ve been the cause. Or, maybe it was an attack. The fact of the matter is  that “the system was broken.” The disturbance in the force caused a bank run from the protocol and, in turn, that caused a death spiral that lead UST and its twin sister LUNA to go to zero.

The NYDIG paper points out two design vulnerabilities that the Terra ecosystem had. Number one, “other aspects of the LUNA/UST set-up, in foresight, were even worse than the inadequate 19.5% Anchor “yield.” For example, investors needed to first buy LUNA to subsequently mint UST, and only then could they deposit the UST in Anchor.” Number two, “algorithmically permitting the printing of LUNA in “unlimited amounts” was the fatal design flaw, guaranteeing, in advance, that a UST bank run – and corresponding LUNA hyperinflation – was a possibility and, via Gresham’s Law, an inevitability.”

NYDIG‘s Definition Of Yield

The controversial Anchor protocol “advertised a 19.5% “yield.” According to NYDIG, not Anchor nor DeFi in general are using the word correctly. “The only sustainable source of yield is sustainable economic return, which in turn depends on the positive-sum game of employing capital to meet consumer needs in the real economy. There is no other source. Calling something “yield” does not make it yield.”

How did the Anchor protocol pay all of its clients, then? Simple, 

“Anchor’s “yield” was not sourced from sustainably profitable economic activity. Rather, Terra’s parent company periodically transferred portions of its $30 billion treasury to Anchor. This meant that unless Terra could raise enormous sums of new funding indefinitely, it would eventually run out of money”

Apparently, the whole Terra ecosystem was frail.

UST price chart on Gemini | Source: UST/USD on TradingView.com

Centralization And Total Value Locked

Remember, the NYDIG paper is an indictment on DeFi in general. The first points of contention are the concept of TVL or Total Value Locked and the idea that DeFi is decentralized. Nothing could be further from the truth, according to the authors. And they’ll use the Terra ecosystem as an example to prove that point

“DeFi is not decentralized. The Terra ecosystem was not decentralized. Terra initially sourced funding from LUNA token issuance apportioned to Terraform Labs at inception. Also funded by Terraform Labs, the Luna Foundation Guard (LFG) was a Singapore “non-profit” set up to help maintain the functioning of the UST system.” 

The centralized organizations that surround a supposedly decentralized one will take control if needed. This means they’ll eventually take control 100% sure. “As is so often the case in DeFi, peacetime decentralized governance swiftly gives way to wartime centralized governance when a crisis arises.” Ain’t that concept familiar. 

Speaking about familiar concepts, “Perhaps the most common metric employed to appraise and value DeFi tokens, “Total Value Locked” (TVL), represents neither “total”, nor “value”, nor “locked.” 0 for 3.” That might sound harsh, but, “It’s not value because they often rehypothecate the collateral.” That’s right, “DeFi projects often represent, and rely on, a series of rehypothecations. The “collateral” in one application can be used in others, ad infinitum.”

Can It Work, Though? NYDIG Says No

At least not yet. Not algorithmic stablecoins nor DeFi are possible in the crypto market’s current state. “No matter how well intentioned, all algorithmic stablecoins will fail and the vast majority – possibly all – of DeFi’s current versions will fail, where “fail” here means not gaining sufficient critical mass to matter, being hacked, blowing up, or being altered by regulation to the point of non-viability.”

Related Reading | Mike Novogratz Speaks: Terra’s UST Was “A Big Idea That Failed”

What does NYDIG propose instead? To build the whole DeFi stack over bitcoin’s Lightning Network. You’ll have to read the “On Impossible Things Before Breakfast” paper for details, though.