#Definition
A bonding curve is a mathematical function that defines the relationship between an asset's price and its supply. In prediction markets, bonding curves power automated market makers (AMMs), allowing traders to buy or sell shares at any time without needing a counterparty on the other side.
When you trade against a bonding curve, the smart contract mints new shares when you buy (increasing supply and price) or burns shares when you sell (decreasing supply and price). The curve ensures there's always a price at which you can trade, providing continuous liquidity regardless of market activity.
#Why It Matters in Prediction Markets
Bonding curves solve fundamental liquidity challenges:
Always-available trading
Traditional order book markets require matching buyers and sellers. In niche prediction markets with few participants, this can mean no trades for hours or days. Bonding curves enable immediate trades 24/7.
Predictable pricing
The mathematical formula determines exactly how much your trade will cost. Before executing, you can calculate the price impact of any position size.
Automated market making
No human market maker needs to manage inventory or set spreads. The algorithm handles everything, reducing operational costs and enabling markets that wouldn't be viable with traditional infrastructure.
Permissionless market creation
Anyone can create a market with a bonding curve by providing initial liquidity. This enables rapid market creation for emerging topics.
#How It Works
#The Basic Mechanism
Bonding curves define price as a function of supply:
Price = f(Supply)
#Visualizing the Curve
Imagine a graph where the X-axis is the Supply of Shares and the Y-axis is the Price.
- Early Buyers: Buy when supply is low → Price is low.
- Late Buyers: Buy when supply is high → Price is high.
- Selling: Moves you back down the curve (reducing supply and price).
This mechanism ensures that as demand increases (more people buy), the price automatically rises.
Common curve types include:
Linear curve
Price = m × Supply + b
Price increases at a constant rate as supply grows.
Exponential curve
Price = a × e^(k × Supply)
Price increases faster as supply grows; early buyers get better prices.
Constant product (x*y=k)
x × y = k
where x = token A reserves
Used by Uniswap and many prediction market AMMs.
#Constant Product Formula in Prediction Markets
Most prediction market AMMs use variations of the constant product formula:
Yes_reserves × No_reserves = k (constant)
Example:
- Pool starts with 1,000 Yes tokens and 1,000 No tokens
- k = 1,000 × 1,000 = 1,000,000
Trader buys 100 Yes tokens:
- New Yes reserves must satisfy: Yes_new × 1,000 = 1,000,000
- Calculate after removing 100: (1,000 - 100) × No_new = 1,000,000
- No_new = 1,000,000 / 900 = 1,111.11
- Trader pays: 1,111.11 - 1,000 = 111.11 No tokens (or equivalent USDC)
The purchase cost more than 100 because buying reduces Yes supply, increasing its price.
#Numerical Example: Price Impact
Starting state:
- Yes reserves: 10,000
- No reserves: 10,000
- k = 100,000,000
- Price of Yes: 10,000 / (10,000 + 10,000) = 0.50 (50%)
Small trade (100 Yes):
- Yes after: 9,900
- No after: 100,000,000 / 9,900 = 10,101
- Cost: 101 No tokens
- Effective price: 101 / 100 = $1.01 per Yes (about 1% slippage)
Large trade (2,000 Yes):
- Yes after: 8,000
- No after: 100,000,000 / 8,000 = 12,500
- Cost: 2,500 No tokens
- Effective price: 2,500 / 2,000 = $1.25 per Yes (25% slippage)
The bonding curve automatically charges more for larger trades, protecting the pool from being drained.
#Dynamic Bonding Curves
Advanced AMMs use Dynamic Bonding Curves that adjust k or other parameters based on market conditions:
- Liquidity Concentration: Concentrating liquidity around the current price (like Uniswap v3) to reduce slippage for active trading ranges.
- Time-Decay: Adjusting the curve as the event deadline approaches to reflect the "time value" of money or certainty.
- Sigmoid Curves: Using an S-shape curve instead of an exponential one, which provides flatter prices in the middle (50% probability) and steeper prices at the edges (0% or 100%), mimicking how human traders behave.
#Examples
#Example 1: Binary Election Market
A prediction market for an election uses a bonding curve:
- Initial liquidity: $50,000 split equally between Yes and No tokens
- Early trader buys 0.51
- Price moves from 0.52
- Later buyers pay progressively more as Yes reserves deplete
The curve reflects increasing conviction in Yes while ensuring liquidity remains for No buyers.
#Example 2: New Market Creation
A user creates a market on whether a tech company will launch a product:
- Deposits $1,000 initial liquidity
- Curve initialized with equal Yes/No reserves
- Price starts at $0.50 (50% implied probability)
- Trading begins immediately; no need to wait for order book depth
#Example 3: Arbitrage Correction
The bonding curve shows Yes at 0.55:
- Arbitrageur sells Yes (or buys No)
- Each sale increases Yes reserves, lowering price
- Trading continues until price reaches ~$0.55
- Arbitrageur profits; curve now reflects updated information
#Example 4: Liquidity Provider Returns
An LP deposits $10,000 into a bonding curve pool:
- Receives LP tokens representing pool share
- Earns trading fees (typically 0.1-2% per trade)
- At resolution, receives share of remaining pool value
- Risk: Impermanent loss if one outcome dominates
#Risks, Pitfalls, and Misunderstandings
Impermanent loss
When outcome probabilities shift significantly, LPs suffer impermanent loss. In prediction markets, this is especially severe because one side eventually goes to 0.
Slippage on large trades
Bonding curves inherently create slippage. The larger your trade relative to pool size, the worse your execution price. Always calculate expected slippage before trading.
Front-running and MEV
On public blockchains, pending transactions are visible. Bots can front-run large trades by buying before you (raising the price) then selling after your trade executes.
Low liquidity amplifies volatility
Small pools have steep curves; small trades move prices dramatically. This can create erratic price behavior that doesn't reflect genuine probability updates.
Curve manipulation
With enough capital, traders can move prices on bonding curves. Unlike order books where manipulation faces opposing limit orders, curves move deterministically with trades.
Resolution loss for LPs
Unlike trading fees on exchanges, prediction market LP positions lose money at resolution because the pool necessarily holds the losing side.
#Practical Tips for Traders
-
Check pool depth before trading: Larger pools mean less slippage. Look for pools with reserves significantly larger than your intended trade size
-
Calculate slippage before executing: Use the AMM formula to determine your actual cost. Don't be surprised by price impact on large orders
-
Split large orders: Breaking a large trade into smaller pieces over time can reduce total slippage, especially in thin markets
-
Watch for sandwich attacks: On public chains, large trades attract front-runners. Consider using private transaction pools or smaller trade sizes
-
Compare AMM prices to order books: Some markets have both. The order book might offer better execution for certain sizes
-
Understand LP risks before providing liquidity: Prediction market LPs face unique risks including guaranteed losses at resolution. LP returns require high trading volume to offset these losses
#Related Terms
#FAQ
#How is a bonding curve different from an order book?
Order books match buyers with sellers: your trade executes against another person's order. Bonding curves let you trade against a mathematical formula backed by a liquidity pool. Order books can offer better prices but may have gaps; bonding curves always provide a price but with slippage proportional to trade size.
#Why do prediction markets use bonding curves?
Traditional order books require active market makers and sufficient trading volume. Prediction markets often cover niche topics with limited natural liquidity. Bonding curves provide guaranteed liquidity from day one, enabling markets that couldn't exist otherwise.
#Can I lose money providing liquidity to a bonding curve?
Yes. Liquidity providers face impermanent loss when prices move significantly. In prediction markets, this is guaranteed because markets resolve to 0 or 1. LPs profit only if trading fees exceed impermanent loss, which requires high volume relative to pool size.
#What determines the shape of the curve?
The mathematical formula chosen by the protocol. Constant product (x*y=k) is most common. The curve's steepness depends on initial liquidity; larger pools have flatter curves and less slippage. Some protocols use custom formulas like LMSR (Logarithmic Market Scoring Rule) designed specifically for prediction markets.
#How do I know if a bonding curve offers fair prices?
Compare the curve's implied probability to your own estimate and to prices on other platforms. Factor in the slippage you'll pay; the displayed price is only available for infinitesimally small trades. For large positions, calculate your actual average execution price using the curve formula.