Why DEXs Failed to Surpass Centralized Exchanges
This blog post is adapted from a recent talk I gave at CryptoDefiance in Singapore.
DEX volume is still ~10,000X smaller than centralized exchange volume.
Why?
One of the biggest barriers (on top of the UX challenges involved with non-custodial key management) has been low liquidity on DEXs.
While most DEXs make it relatively easy for users to connect their MetaMask, and select tokens to buy and sell, once users get to that final step of converting — they often encounter high slippage or spreads, or they can’t convert at all due to no one selling or buying the asset.
So how do we attract more liquidity to DEXs?
Liquidity pools bootstrapped by the DeFi ecosystem may one answer.
What exactly are liquidity pools?
You can think of each liquidity pool as a trading pair on an exchange.
For example ETH:BNT or DAI:BNT. On a normal exchange, each of these pairs would have an order book that aggregates buy or sell orders for these tokens.
Liquidity pools remove the need for order books and replace them with on-chain reserves. Transactions against liquidity pools are drawn from these reserves.
For instance, anyone can set up an ETH:BNT pool by depositing an equal amount of ETH and BNT tokens in a smart contract. This pool would allow the user to send ETH and receive BNT in return, or vice versa.
A small fee is taken out of each trade and redeposited into the pool and this liquidity provider fee serves as an incentive for users to stake their tokens the pool. In turn, liquidity providers earn a pro-rata share of the fees generated from conversions.
Notably, pools connect via a common network token token to form liquidity protocols. Given a ETH:BNT & DAI:BNT pool, a user is be able to convert between ETH and DAI in a single transaction, via BNT.
How Liquidity Protocols automate and decentralize DEX operations
Think about the core operations of a DEX:
- Listing: deciding what tokens to list
- Order-matching: organizing order-books and facilitating trades
- Market-making: capturing trading fees as DEX revenue
Liquidity protocols effectively automate and decentralize these operations. Anyone can list a token by creating a pool and the protocol cannot block any participant from listing. Order-matching and and market-making are performed automatically by the protocol’s smart contracts.
This last point is key.
Market-making has historically been highly technical. Up until now, it has been reserved for professional market-makers who run advanced trading strategies. And it’s one of the reasons we’ve seen so few liquidity providers on DEXs — because market-making on order book-based exchanges is out of reach for most users.
Liquidity pools turn market-making into a couple clicks. So everyday users can add liquidity to a decentralized exchange and in return receive a share of the exchange’s fees.
As you can see above, anyone can effectively obtain shares in Bancor by providing liquidity to a pool. Bancor can’t block anyone from creating a pool or adding liquidity. Participants receive pool shares in return, which generate fees. Participants can even vote on the design of each pool — such as its trading fees.
So where are liquidity protocols today?
Bancor went live in late 2017, and since then we’ve seen more and more projects using liquidity pools. The amount of locked liquidity and conversion volume processed by liquidity pools has grown steadily, especially in the last 6 months.
But we’re a long way off from seeing the kind of activity centralized exchanges see.
Key Challenges for Liquidity Pools
There are some key challenges we must address before liquidity pools are widely used in asset exchange.
- Impermanent Loss: is caused by price fluctuations of a pool’s underlying assets. This loss cuts into the trading fees generated by the pool- which has dragged down the profitability for some liquidity providers.
- Slippage: refers to the change in price during the execution of a trade. With liquidity pools, slippage is based on the transaction size relative to the amount of liquidity in the pool. Thus, larger orders can incur large slippage costs, making prices unappealing for traders and arbitrageurs.
- Visibility & Control: There is a lack of analytics and ROI tools that allow liquditiy providers to track pools and optimize their profitability.
Thankfully there are solutions on the way.
Stablecoins will play a key role in reducing impermanent loss and improving the profitability of liquidity pools.
Liquidity providers can use stablecoins to create or contribute to pools on the Bancor network without having to speculate on two volatile assets
Moreover, stable:stable pools — such as DAI:USDB, Tether:USDB — reduce impermanent loss to zero and liquidity providers just receive the fees.
To address the slippage issue, Bancor is testing a new mechanism called virtual reserve.
Essentially the pool contract points to a virtual reserve of tokens, and the slippage you see as a trader resembles that of a liquidity pool with far more physical liquidity depth — so you don’t actually need a ton of liquidity in a pool to affordably process large transaction sizes.
Blockchain data in general is becoming more sophisticated, and this is allowing more developers to build APIs and front-ends that interact directly with the blockchain, so liquidity providers can more easily access track the ROI and optimize their pool holdings.
If you are an analytics provider or building portfolio management tools, I highly suggest you start building solutions for liquidity pools.
60,000 More Liquidity Providers, Coming Soon
So how do we help liquidity pools thrive?
For Bancor, this means really focusing on the protocol level, and advancing the contracts as fast as possible to attract more liquidity providers, more developers building on this primitive and developing user-friendly front-ends, and more community governance by pool token holders and BNT holders. (See Bancor’s Roadmap Here.)
It’s worth calling out that this coming December Bancor is airdropping over $2 million worth of $ETHBNT to all BNT holders — which will create more than 60,000 liquidity providers on the network.
Without taking any action, BNT holders will see ETHBNT pool tokens land in their wallet and they will automatically start collecting fees from every ETH conversion on the bancor protocol.
A “Google moment” for DeFi?
In sum, there is an amazing potential for automated decentralized liquidity to form a vastly larger share of total crypto asset exchange.
We’re seeing liquidity pools being used in more and more projects and becoming more interoperable across networks and blockchains.
Here’s how to get involved:
- Traders: Becoming a liquidity provider (or just hold BNT in a non-custodial wallet this December)
- Dapps & Wallets: Plug into pool-based liquidity (You can earn affiliate fees!)
- Devs: Build on liquidity pools (Apply for a Bancor dev grant & check out Bancor’s docs)
- Token teams: Incentivize your community to create/add to your token’s liquidity pools (How about a monthly airdrop on your LPs?)
- Geeks: Research liquidity pools (this kind of research is badly needed!)
