Automated Market Makers for Sports Betting
In the last post, we discussed the difference between market-making in the stock market vs. traditional sportsbooks. Here is the link in case you missed it! Let us continue from the previous post and introduce automated market makers and how they work in general.
What is automated market making?
Obviously, an automated market maker (AMM) is simply “automated” + “market maker”. It is a protocol that plays the role of the market maker.
An AMM will quote prices between two assets according to a pricing algorithm. At a minimum, it takes the trade size and the existing liquidity as inputs to determine the price.
To illustrate, we will describe the constant product AMM (how Uniswap operates) which is the most basic and widely used AMM algorithm in the decentralized finance (DeFi) world. The Uniswap AMM defines the constant product curve, which is used to determine the relative price of two assets.
Suppose initially there are 50 apples and 50 bananas in the store and we are interested in trading between apples and bananas. Then, the constant product curve is defined as
Now the pricing of apples per banana (or vice versa) is determined based on preserving the constant product, 2,500 as shown in the animation below:
This animation is a great illustration. It is very intuitive! The price of a banana for an apple is determined by taking the ratio of the quantities in the store. If there are more apples than bananas in the store, then the price of bananas will be more expensive than apples and the pricing follows the constant product curve, XY=2,500.
Suppose the merchandise brought 25 apples and 25 bananas the next day. Then, the constant K formula will adjust to
Similarly, if the store manager ate 1 apple and 2 bananas for lunch, then the constant K becomes:
In this analogy, the store manager is the liquidity provider. Something that is not obvious from the illustration is the concept of impermanent loss and impermanent gain which will be explored further in a future article. The key point is that liquidity providers will lose money if the price moves away from the initial ratio. As long as the ratio comes back to the original balance, the loss is “impermanent”.
Up to here, we just illustrated the basics of the simplest AMMs and gave a taste of how the pricing mechanism works.
How does AMM work in the sports prediction market?
You can probably already imagine how the pricing mechanism works for the sports prediction market. If we replace apples and bananas with sports teams or replace them with other binary market events, then the same logic applies straightforwardly. For example, replace apples and bananas with Tottenham and Manchester United respectively — ratio determines the relative price of each outcome. In other words, the odds.
Now, there are a couple of points that we would like to highlight!
It is important to be aware that the price is determined depending on the wager size. In our previous article, we compared sports betting to the stock market and pointed out “slippage” that depends on how much liquidity is in the market, and the size of the order/wager. In other words, the larger the wager size, the more expensive the offered price will be.
This is different from the fixed odds in traditional sports betting where sportsbooks limit the size of bets because of fixed odds.In the previous section, we briefly mentioned impermanent loss when the price moves away from an initial ratio. This is actually the worst-case scenario for sports prediction because the price will always move away to an extreme by the end of the event (when the outcome is determined). This means the loss for liquidity providers becomes large and permanent. A fixed product market maker is thus not suitable for sports betting.
So far we talked about pricing between two assets, but you might be wondering how the pricing works with respect to the USD. In the end, you are purchasing some assets using USDC on our platform (USDC is a digital “stablecoin” currency that is backed by USD which makes the price one-to-one). We will come back to explain this in a later article and leave this as an exercise for you to think about.
Maybe you are also wondering how you apply this to multiple outcome markets. For example, soccer consists of three potential outcomes: win, lose, and draw. We will also leave this as an exercise for you to think about.
The Uniswap / constant product AMM is the simplest form of an automated market maker. There are many other variations of AMMs that were studied both in industry and academia. Which one is most suitable for the sports prediction market is an open question, and obviously, UBet is developing its own way of pricing markets. We will post another blog that gives an overview of different AMMs and their properties in the coming weeks.
Website: http://ubetsports.io/
Twitter: https://twitter.com/UBet__Sports
Discord: https://discord.gg/B7brvUEAFG
LinkedIn: https://www.linkedin.com/company/ubetsports/
Reference:
[1] The figure from another medium post: https://medium.com/dragonfly-research/what-explains-the-rise-of-amms-7d008af1c399
Disclaimer: The above content is intended for information purposes only and is not intended to be gambling, trading, nor investment advice. Please perform your own analysis before making any betting, trading, or investments based on your own personal circumstances or seek a duly licensed professional for gambling and trading advice. Accordingly, neither UBet nor its partners, directors, shareholders or employees are liable for any damage, expense, or other loss that you may incur out of reliance on the above information.