A comprehensive interpretation of what an automated market maker is?

This article was published on March 23, 2020, on Chain News, authored by Richard Burton and translated by Perry Wang. The original title is "Understanding the Rising Popular Category of Ethereum DeFi: Automated Market Makers."

Let’s start with the most basic concepts.

How do markets operate?

The relationship between supply and demand has been a topic of study for economists for a long time. As people buy and sell goods from each other, this relationship evolves over time.

In the stock market, market makers are responsible for facilitating trades between buyers and sellers. They make money through the spread. The spread refers to the difference between the price offered by one party for an asset and the price at which another party buys it.

How Automated Market Makers Work

Automated Market Makers (AMMs) have been present in traditional financial markets for some time. They allow financial professionals to establish price relationships that must be followed. If a certain asset reaches a specific price, a certain amount of that asset will be bought.

In traditional finance, AMMs are operated by humans. They are deployed, controlled, started, and stopped by people. They operate in centralized exchanges, which are also controlled by humans.

On Ethereum, this human control can be minimized. Humans still create some market-making bots, but thereafter, we can release them to operate autonomously like some economic organisms. If designed correctly, they can achieve automation and irreversibility like the Ethereum protocol.

Not just automation, but with autonomy.

Uniswap is one of the earliest market-making bots on Ethereum (of course, Bancor was the real first). In 2018, I had the privilege of seeing the Uniswap bot being built in my office.

To understand how AMMs like Uniswap operate, we need to grasp their core concepts:

? Market Makers
The human entities that trade to form a market.

? Automated Market Makers (AMMs)
Bots composed of smart contract code that perform market making.

? Fixed Product AMMs
Bots that use a specific bonding curve to determine the price at which a trade is executed.

What curves are we talking about?

? Supply Curve
A graph depicting the relationship between the cost of a good and its supply. This curve is drawn after observing data.

? Bonding Curve
The relationship between the price of a token and its supply, defined by a mathematical curve.

The asset supply is aggregated in a liquidity pool:

?‍♂️ Liquidity Pool
A smart contract that stores the tokens ready for trading.

⚖️ Arbitrage
Here, traders can profit from the price differences or imbalances in the liquidity supply.

? Exchange
Here, using curves and liquidity pools, one token is exchanged for another.

All of this leads to the concept of the bonding surface:

? Bonding Surface
A multidimensional price and supply relationship between three or more tokens. This tool makes synchronous trading across multiple assets more convenient.

? Constants and Invariants
These are mathematical functions that can be used to define various curves in a graph.

Now, that’s all the core elements of Ethereum AMMs.

What Projects Operate on AMMs?

An increasing number of new projects are emerging in the automated market-making space. You can look into the inspirational history behind Uniswap to examine the lineage of these projects.

In 2018, the Uniswap project was born, aiming to compete with Kyber, 0x, and Bancor, which had already raised over $200 million at that time. With some bonuses and loans, Uniswap's main developer Hayden Adams launched the system, immediately outperforming its competitors. Other AMM bots were too complex and difficult to use. Uniswap was user-friendly.

Uniswap and Bancor are both AMMs in the market share of decentralized exchanges.

Let’s take a look at some of the major AMM projects currently on Ethereum:

Bancor 2017

• The first major smart contract AMM from the wild
• Pioneered several key feature designs
• Used a bonding curve to calculate prices
• Tracked liquidity provider contributions via a bonding curve
• Connected multiple liquidity pools through a hub-and-spoke model
• Unfortunately, adopted BNT tokens as its hub currency (instead of Ether)

Uniswap 2018

• Main goal: A decentralized protocol for token liquidity
• A simpler bonding curve than Bancor
• Used Ether as its hub currency
• Very gas-efficient
• Not highly capital efficient
• Overall, exceptionally well-built (see Hayden's tweet from November 2018 when he launched the product)

Recommended reading: From Experimental Projects to Uniswap's Hundredfold Liquidity Growth

Curve -- 2020

• Main goal: Low-slippage trading between stablecoins
• Extremely high capital efficiency => very low slippage
• A multidimensional bonding surface similar to Balancer
• Performs very well with stablecoins, but not as well with non-stablecoins
• Its bonding curve has certain risks
• The actual bonding curve formula is opaque (see my tweet commenting on this)

Recommended reading: Understanding the Popular DeFi Project Curve: Why Is It Important? Why Was It "Attacked"?

Comparison of Curve (also known as Stableswap) bonding curve with Uniswap bonding curve.

Ethereum will soon welcome several new AMM projects:

Balancer -- Expected in 2020

• Main goal: A decentralized Vanguard (Note: Vanguard is the largest no-fee fund family globally and the second-largest fund management company in the world)
• Expanded Uniswap's bonding curve into a multidimensional surface
• Enabled token weights in liquidity pools (e.g., ETH=50%, Dai=25%, MKR=25%, etc.)
• The capital efficiency of the portfolio model is more than twice that of the hub-and-spoke model
• However, the portfolio model may lead to liquidity fragmentation
• The introduction of weights increases gas costs
• Learn more

Balancer can create many pools composed of various tokens.

Shell Protocol -- Expected in 2020

• Low slippage in trading
• High profits for liquidity providers
• Weighted, multidimensional pools
• Supports any style of stablecoins (like Dai, cDai, and Chai)
• Granular customization of AMM behavior
• Mitigates the risk of a stablecoin default
• Transfers arbitrage profits to liquidity providers

Balancer can create liquidity pools composed of various tokens.

Why Do They Perform Better?

I think Jacob Horne (former product manager at Coinbase and founder of the online marketplace Zora, which uses bonding curve pricing) said it well:

It turns out that Uniswap is the most efficient way to launch a liquid exchange for any asset. @UniswapExchange is the "preferred liquidity provider," but that doesn't mean you won't see other order book exchanges supporting asset trading. The progress we see is the following:

What Uniswap brings to the market is "uncensorable token issuance." If you want to list a stock on the New York Stock Exchange, you need to obtain permission from multiple financial institutions. The cost can often reach millions of dollars.

If you want to list a token on Ethereum, you just need to create a single transaction. The cost may be less than a dollar.

Why Do They Operate on Ethereum?

Traditional markets use order books because they can operate at high speeds. People can place orders and cancel them in less than a second.

Ethereum, on the other hand, processes a transaction block every 20 seconds. AMMs operate well because in each round on Ethereum, you can input one token and receive another in return.

Its shortcomings in price perfection are compensated by predictability.

Traditional economists see slippage in this. Ethereum economists see simplicity. It is an excellent closed loop that can handle large amounts of capital.

The market tells us that these systems are very valuable.

The steadily increasing trading volume on Uniswap shows that AMMs have high value.

Some Conclusions

We are witnessing the rise of autonomous automated market makers on Ethereum. These unstoppable bots are competing for liquidity with each other while also competing for liquidity from traditional order book exchanges.

The differences between these AMMs are subtle yet significant: their base trading pairs, the design of their bonding curves, and the reward structures for liquidity providers all vary. Some even enhance the ability to create liquidity pools containing multiple assets, a sophisticated design that can give AMMs ETF-like functionalities.

As liquidity increases, their price slippage will narrow, and efficiency will improve—better meeting the demands of the DeFi economy. The best AMM experiments are likely to survive and create lasting network effects.

We enjoy the fun of using liquidity bots, and we will watch them achieve great results.