Defi Education: Spark Protocol and how it outperformed other lending protocols in Q2

DeFi lending had a brutal Q2. The rsETH exploit shook a lot of confidence in DeFi in general. Borrowing demand was weak. Spreads were tight. Revenue fell.

Some numbers: Aave's interest income dropped 23% & Ethena's fees dropped 21%. In that same window, Spark more than doubled its share of the lending market & stayed profitable. As a defi maximalist, this headline caught my attention. Here's my takeaway's from the latest Blockworks report on their Q2.

First, what is Spark? It's a capital allocator for stablecoins. It sources liquidity from Sky (the protocol formerly known as MakerDAO) & routes that capital across a range of venues: its own market SparkLend, Morpho, institutional desks like Anchorage & Arkis, & Uniswap. It moves the money to wherever demand & spreads are best. Essentially, it's a protocol of DeFi yield farmers trying to maximize return and minimize risk.

The data - Spark went from 4.3% to 10.4% of outstanding loans across the major venues while the combined loan book for that whole group shrank about 30%. It grew its slice while the pie got smaller. Net income was ~$3.3M for the quarter, operating costs flat at $1.1M a month. Spark bought back $1.31M of SPK in Q2 straight out of operating surplus. So here we have a real, profitable business buying back their own tokens with profit.

My takeaway on risk management - In mid-April a bridge exploit elsewhere in DeFi (the rsETH depeg) stressed collateral markets across the board. Spark had built oracle killswitches that freeze loans if an asset's price moves more than 5% off peg, plus rate limits on how fast capital can move. This is growing in the industry (Morpho, Gauntlet, and Yearn and 3 protocols I'm a fan of that also have these kill-switches, but it isn't ubiquitous). During the exploit, Spark took zero loss. In the weeks following the KelpDAO exploit, Spark added ~$1B in inflows. Capital managers evaluated Spark and deployed due to the risk management system.

The third thing Spark is growing as a liquidity engine behind other people's stablecoin products. You can read the entire insight by Blockworkds, but they are basically working with PayPal & PYUSD, Uniswap, Robinhood Earn, and BitGo. The earn programs operate on an infrastructure level, simplifying UX and Spark earns a portion of the yield they generate. Vaults, but vaults with strong integrations.

A lot of Spark's revenue has shifted into one bucket. Distribution rewards (payments from Sky for driving USDS adoption) hit $4.88M in Q2 & now make up around 74% of net revenue, while lending-spread income fell 59% as rates compressed. Distributing a stablecoin at scale is a real, hard job, & demand for it is exploding. Spark does it well, but there's a concentration risk as it makes up so much of the business. At the same time, Spark holds more than double the lending market share it had, so when borrowing demand recovers, that share converts straight into profit. They're positioned for the recovery while having another business that can generate income while waiting on a recovery.

The big takeaway is DeFi is no longer a 1-trick game for the protocols. They need multiple, diverse revenue streams if they're going to survive the competitive landscape and crypto winters. Spark seems to have that. We've seen Aave & Morpho getting into more verticals also.

Some may be bearish on the future of DeFi, but I've never been more bullish.