When we set out to build a token with a guaranteed floor price, we started where everyone starts: Uniswap. Take the x*y=k formula, manage the liquidity positions well enough, and enforce a floor on top.

This is the story of how we got here. Three versions, two chains, eight audits, and a fundamental rethink of how on-chain liquidity should work. It's also the first article in a series about what floor-backed tokens actually are, how they work, and why they're about to change how tokens are issued.

What "floor-backed" actually means

A floor-backed token has a minimum redemption price that is enforced on-chain, backed by real reserves, and mathematically guaranteed to never decrease. Three conditions have to hold simultaneously:

The floor must be backed by real liquidity. The reserves must be locked in smart contracts, always available for redemption.
Any holder, at any time, must be able to sell at or above the floor price.
The floor must only go up.

Satisfying all three at once is a mechanism design no existing AMM was built to solve.

Constant product pools (x*y=k pools), pioneered by Uniswap, have no concept of circulating supply. The formula only tracks the token balance inside the pool, not the total supply in existence. The pool literally cannot tell a token's market cap because it's not part of the equation.

This creates cascading problems for anyone deploying token liquidity. The initial ratio of tokens to reserves permanently determines price sensitivity, liquidity depth, and supply distribution for all future states of the pool. A minor difference in that ratio has massive consequences as supply expands and contracts along the curve.

The math breaks entirely when the pool holds more than half the total supply. The TVL of the pool becomes more valuable than the entire FDV of the token, violating all financial rationality. This is how $SLERF ended up with 75,000 SOL of permanently inaccessible capital.

The pool was working exactly as designed. The problem is bigger than any single pool. Tokens built on x*y=k are broken by default. Value leaks out like water through a leaky bucket: to arbitrageurs, to mercenary LPs, to dead capital sitting in unreachable price ranges. The industry has spent years building increasingly clever ways to slow the bleed. Ve(3,3). Concentrated liquidity. On-chain market making for token issuers. None of them fix the crack. They just catch some of the water.

85% of tokens launched in 2025 ended negative. Only 9% that dropped in week one ever recovered. The damage is baked into the curve before the token even hits the chart.

Protocol Owned Liquidity

OlympusDAO pioneered the idea that protocols should own their own liquidity. We agreed.

But OHM's floor was implicit, not programmatic. There was no smart contract enforcing a minimum redemption price. The "backing per OHM" metric looked like a floor on a dashboard, but holders couldn't redeem at that price on-chain. When sentiment turned, the price crashed well below backing.

That's where Baseline started.

The Baseline evolution

We took that lesson and started building. What followed was three protocol versions, each one revealing a deeper truth about what floor-backed liquidity requires.

V1: The experiment (Blast, 2024)

The first Baseline token launched on Blast using Uniswap V3 as the underlying AMM. The thesis: if we manage concentrated liquidity positions intelligently enough, we can enforce a floor and grow it over time.

It worked, partially. The floor held. The BLV grew. But V3's concentrated liquidity ranges meant value accumulated slowly. Capital got stuck in ranges that weren't actively serving the market.

V2: Different liquidity, same curve

V2 iterated on liquidity management with new range strategies and the Afterburner, a randomized leveraged buyback-and-burn. Capital efficiency improved. But no matter how sophisticated the liquidity management layer, the fundamental behavior of constant product doesn't change. It was still designed to maintain balanced exposure between two external assets, not to form capital and grow a floor price.

V3: Confirmation (Base)

V3 was the cleanest implementation of Baseline on a constant product curve. Running it in production confirmed what we'd been circling: the curve was the bottleneck. Not the parameters. Not the ranges. Not the chain.

The constant product formula harvests fees from volatility, stays balanced, and treats all flow symmetrically. We needed a curve that does the opposite: accumulate value asymmetrically, price based on circulating supply, and budget reserves to buy back every token in existence.

Mercury: a new curve for a new era

Mercury is not an optimization of x*y=k. It's a different equation for a different purpose.

y = K · (x/c)² + BLV · c

The critical variable is c: circulating supply. For the first time, the curve knows how many tokens are held outside the pool and prices accordingly.

Price scales with distribution. What makes a token valuable under Mercury is the ratio of tokens being held versus tokens that have been sold back. When the entire supply is sold back into the pool, Mercury uses all available reserves to buy out every remaining token. Each token's value decomposes into a Baseline Value that holds constant regardless of circulation, and a premium that scales quadratically with demand. Strong demand means trading at multiples above BLV. Full unwind means the price returns to the floor. Mercury adapts across the entire token lifecycle. The pool becomes an asset that improves itself over time, not a liability that needs constant management.

What is a Baseline Token

The token owns and manages its liquidity on-chain, 24/7.
Fees flow back into growing the floor, funding staking rewards, and strengthening reserves. The floor goes up because the mechanism captures trading activity and redirects it into backing.
Zero-liquidation lending. BLV can never decrease, so holders borrow against it at 0% interest with no liquidation risk and no oracles. Capital unlocked without selling.

The data

We replayed $GAME's entire trade history, over 980,000 trades on Aerodrome, through a simulated Baseline pool. Same trades, same timing, different curve.

Price: +59.6% higher from identical trade flow.
Liquidity growth: +267.7% more reserves backing each circulating token.
Supply control: +200.5% more tokens pulled out of circulation.
Backing: +21% growth in guaranteed floor price.
Fees: +81.1% more collected in trading fees.

Other simulation reports: https://sim.baseline.markets/

Why this matters now

Mercury fixes the curve. And when you fix the curve, you fix a lot of things in crypto.

Every version taught us something. V1 proved the floor could hold. V2 proved liquidity management alone wasn't enough. V3 proved the curve was the bottleneck. Mercury is the answer.