Building DeFi protocols on Polkadot: An Interview with Bette Chen, Co-Founder at Acala Network

I sat down with Bette Chen, co-founder of Acala Network to talk about DeFi and building decentralized protocols on the Polkadot network.

Ishan Pandey: Hi Bette, welcome to our series “Behind the Startup.” Please tell us about yourself and the story behind Acala Network?

Bette Chen: I have a background in Engineering specialized in Software. I was a developer earlier in my career, then turned - to arguably the dark side - and became a Product Manager. Like many others, a few years ago, I was drawn to the Bitcoin rabbit hole for its technology and ideology. Quickly, I discovered Ethereum and spent most of my spare time learning and building. Close to 3 years ago, I started working in the blockchain full-time as a product manager, where I met my co-founders Ruitao and Bryan.

Acala was co-founded by members of two Polkadot ecosystem teams - Laminar and Polkawallet. Laminar is building a synthetic asset and margin trading platform intending to bridge on-and-off chain liquidity and traders. Polkawallet is one of the first mobile wallets in the Polkadot ecosystem. The founders (Ruitao, Bette, Bryan, Fuyao) met in Hangzhou China in 2019 at the 1st Substrate Hackathon, where Bryan was one of the judges alongside Dr. Gavin Wood.

There was this ‘love at first sight’ of the two teams - our values are highly aligned, both are hardcore blockchainers believing in Web 3.0 and decentralization; both are dedicated builders who truly understand Substrate/Polkadot’s vision through its code.

The teams got to know each other by coding together for a month or so, before engaging in a formal relationship and co-founding Acala in October 2019.

Ishan Pandey: What is parachain economic modeling, and what is the governance architecture?

Bette Chen: A parachain is a distributed database unique to the globally coherent framework and validate by the Polkadot Relay Chain validators. A parachain will most usually take a blockchain’s shape, although there is no particular have to be real blockchains for it. The idea of parallelized chains running parallel to the Relay Chain derives its name from it. They can parallelize transaction processing due to their parallel design and achieve the Polkadot system’s scalability.

A network maintainer, known as a collator, oversees Parachains. The collator node’s task is to preserve the complete parachute node, retain all required parachute knowledge, and generate new block candidates to transfer to the relay chain’s validators for verification and inclusion in the shared Polkadot state. An implementation detail of the parachain is the incentivization of a collator node. They are not allowed to be staked on the Relay Chain or own DOT tokens until the parachain implementation stipulates.

Polkadot uses a specialized governance system that enables it to gracefully develop at the absolute behest of its mobilized stakeholders across time. The specified purpose is to enhance that the plurality of the stake will still command the infrastructure.

To accomplish this, several new structures are brought together, including an amorphous state-transition feature stored on-chain and specified in a platform-neutral intermediate language (i.e. WebAssembly) and multiple on-chain voting mechanisms, like adaptive super-majority requirements and batch acceptance voting referendums. Any protocol modifications must be decided upon through stake-weighted referendums.

There will be multiple Councils managing various aspects of the network: the General Council and ACA Referendum are the overarching chambers that govern the Acala Network. Governance on Acala is both decision making and decision execution apparatus. Network upgrade, smart contract upgrade, protocol parameter change, a state change can all be managed via governance without forking, e.g. the DAO hack loss can easily be reverted through Governance vote.

Then there are specialist Councils, for example, the Honzon Financial Council will manage stablecoin related parameter changes, the Technical Council can submit emergency bug fix or veto unanimously dangerous proposals. 

Acala also takes a phased approach to employ various governance mechanisms that will allow it to progressively decentralize and ultimately be commanded by the majority network stakeholders. The General Council will initially be appointed, then elected by token holders, and eventually, Democracy kicks in with referendum enabled.

The Polkadot Relay Chain would not explicitly implement smart contracts. Smart contracts can be enabled by parachains on Polkadot, in any case. Projects like Edgeware have already been revealed and further may embrace this functionality due to the Substrate built-in contract pallet.

Consequently, the EVM pallet enables the Ethereum Virtual Machine to be deployed by a parachain, thus enabling almost specific Ethereum contract ports.

Storage rent restricts the impact a transaction may have on the processing of a blockchain.

A contract implemented to the chain creates a code hash from which it is possible to build new chain examples, but there is no rent paid to the code hash itself. The lease only extends to cases in this arrangement who have their contract accounts. A one-time byte-fee is currently added to the transaction to deploy a code hash, but no ongoing expense.

The balance of an instance of a contract is paid proportionally to the volume of storage utilized for its account. If the balance of a contract falls beyond a given amount, the balance of the contract is converted into a “tombstone” (a hash of the current state of the contract) and its storage is cleaned up. By supplying the data that was cleared away before it became a tombstone, as well as any extra funds necessary to hold the deal intact, a tombstone contract may be resurrected. For missed rent times, this charge would occur retroactively.

Ishan Pandey: Can you explain what a Parachain Lease Offering is, its advantages, and real-world implications?

Bette Chen: The free and accessible design of blockchain networks opens up attack vectors in conventional bidding formats that are non-existent. In specific, when deployed over the internet or on a blockchain, traditional transparent auctions may be susceptible to auction sniping.

In anticipation of charging lesser than they value the commodity, auction sniping occurs whenever the end of an auction is recognized and bidders are reluctant to offer their accurate price early.

It must fill one of the usable parachain slots for a parachain to be applied to Polkadot. A slot on Polkadot is a finite resource and only a small amount would be available. There might only be a few slots that are unlocked every few months as ramp up. The target is to have 100 parachain slots accessible on Polkadot eventually (these will be split between parachains and the parathread pool). If a parachain needs to have block inclusion assured on each Relay Chain block, a parachain slot must be obtained.

Polkadot’s parachain slots will be sold according to an unpermitted candle auction which has been changed marginally to be safe on a blockchain.

A candle auction is a type of open auction where bidders send offers that are progressively higher and the winner is deemed to be the top bidder at the close of the sale. Initially, in the 16th century, candle auctions were used to sell ships and get their name from the “inch of a candle” that described the auction’s opening time. The auction would abruptly stop until the flame was extinguished and the torch went out, and the standing offer would win at that moment.

They need a random number to determine the moment of termination when candle auctions are used online. The slot auctions of Parachain would vary slightly from a standard candle auction in that the random number is not used to evaluate the length of its opening process. Instead, it has a known open stage that can be retroactively ruled to have stopped at any time in the past (at the usual close). Thus, offers may appear to be approved throughout the open stage. However, future bids are more likely to fail when the retroactively decided near moment could be considered to have followed the time an offer was sent.

As mentioned to access Polkadot’s network security, projects will need to obtain a parachain slot. The number of slots available is limited and therefore a market mechanism is required to compete for a slot, which is via a candle auction, by locking DOTs for the duration of the parachain lease.

Parachain auctions can be approached in many ways, Acala plans to bid for the first parachain slot on Polkadot (also on Kusama, but for simplicity, will only talk about Polkadot hereafter) and be the pioneer in this new crowdsourced network security paradigm - the Parachain Lease Offering (PLO).

Ishan Pandey: How is Polkadot different from Ethereum? What are its advantages?

Bette Chen: Polkadot is a layer 0 network - it is designed both to easily integrate with other blockchains like Bitcoin and Ethereum (via bridges like Interlay and RenVM) and to host several heterogeneous (independent) but interconnected chains (called parachains, like Acala). Polkadot empowers an ecosystem of domain-specific chains optimized for their use cases. Substrate-based chains like Acala enjoy full-tech-stack, e.g., EVM and smart contract capability, as well as continuous innovation and forkless upgrades. Acala occupies the domain of decentralized finance (DeFi).

Apart from being cheaper, faster, and interoperable, specialization also means you can do things far better and create categorically different innovations than generalized solutions.

Ishan Pandey: What is Decentralized Sovereign Wealth Fund? How would it work?

Bette Chen: Acala is a DAO3.0 that, by its on-chain decentralized Sovereign Wealth Fund (dSWF) and governance, obtains financial and trade (rational decision - making) sovereignty. The resources in the dSWF are Asset Under Control, but you can not fork the AUM while you still can fork the code, the liquidity, and even the environment.

A decentralized sovereign wealth fund will offer a mechanism for its people to deliver value, but with the openness and rigid obedience that blockchain enables to structured logic. It could also go faster. A traditional wealth fund only includes raising in valuation, but it may also obtain access to features in other blockchain networks depending on its investments in a system where several tokens have usefulness.

An on-chain account reflects the portfolio itself. This account does not have a private key associated with it. Instead, the general council and network administration have the right to handle the finances in the account.

Funds travel three ways into the budget. The first two come from the Honzon protocol of Acala, which functions close to Builder. Honzon offers a forum for multi-collateral loans that embraces multiple collateral types in return for a USD deposit, the stablecoin of Acala that is pegged to the US dollar. Borrowers ought to pay a stability charge on their loan, much as at Maker.

Unlike in Producer, Acala invests the charges into its sovereign wealth fund instead of burning the fee.

Second, if a loan’s equity falls below a certain threshold, the Acala network takes a penalty charge. This premium removes under collateralized debt that instead happens where any equity has to be liquidated by the network. However, instead of getting fried, much of the stabilization tax falls through the wealth pool.

Third, a liquid staking procedure called Homa is also run by Acala network. On its own, Homa merits a whole post, but it should be nutshelled here like the lending site. As other Evidence of Stake networks, Polkadot locks tokens to protect validators and save the network. Such tokens allow any return if all goes right, but if the picked validators misbehave, the tokens are liable for slashing.

This scheme is very much associated with network stability, although very much in contrast with the thesis of Acala that funds should be worthy of fulfilling new purposes. By building a staking pool where users can deposit DOT in exchange for any freely transferable L-DOTT, Homa aligns this stress (although not necessarily at a one-to-one ratio). For decisions such as a validator allocation policy, the pool is controlled by the Homa Council (again, whole other article), and L-DOT holders will return their L-DOT to the network and recover the underlying DOT plus its share of the pool’s staking income and minus any charge that goes into sovereign wealth fund.

Acala can be seen as a sovereign blockchain, as a sovereign nation or entity. The dSWF helps it to achieve economic sovereignty, by re-investing network surplus in foreign assets such as DOT and KSM, to gain access to network security, long term wealth generation for ACA holders (e.g. DOTs can be staked and generate yield), strategically holding stakes of valuable sovereign networks, just like Norway’s Sovereign Wealth Fund.

Ishan Pandey: Recently, $100 million of investor’s funds were liquidated from the DeFi protocol Compound following the oracle exploit. What are your views on it?

Bette Chen: The incident’s root cause was Compound (to my great surprise) using only Coinbase Oracle as a single price feed for DAI pairs. Hence when the Coinbase exchange had a significant price fluctuation on DAI pairs, many loans were being liquidated on Compound. A few lessons learned from our side as builders in DeFi

Oracles remain the single most critical component in a DeFi protocol that requires external price feeds. And almost all oracles have been exploited to some level at one point or another, and no one has a reasonably ‘perfect’ solution yet.

Folks in the industry have much respect to compound their professionalism, rigidness in design, etc. This incident reminds us DeFi protocols are complicated, any assumptions, any stones left unturned, can easily become an oversight that leads to exploits and significant financial losses.

Ishan Pandey: According to you, what regulations will we see around stablecoins in the future?

Bette Chen: That depends on the nature of the stablecoin mechanism. But of course, in any case, regulation on stablecoin is not a question of if, but a question of when and how.

For a decentralized stablecoin, the regulations are likely to be applied for specific jurisdictions where the technology is being applied and distributed through centralized partners, such as credit card providers, merchants, CeFi providers. These partners are most likely required to be regulated, rather than the core protocol. This is probably the current trend of regulation on the edges. But this might change as CeFi merges with DeFi.

Another aspect to consider is ‘decentralized’ stablecoin can also merge with centralized stablecoin. In Maker’s case, a significant portion of DAI is now collateralized using centralized stablecoin USDC. Many of the traditional banking system rules are applied to USDC. As a ‘customer’ of USDC, the implication is that some of these rules may also be applied to Maker.

Ultimately, we (builders in DeFi) also bear the responsibility of bringing others (e.g. the regulators) along our journey, and we can’t do all of it alone. Many existing regulations don’t make any sense if applied directly to these DeFi/CeFi fusion products. Education might be the key to expanding awareness, fostering more in-depth understanding, understanding what benefits this industry brings to everyone, and helping regulators better adapt/create their policies to/for this categorically different industry.