A token can launch with deep markets and tight spreads, then become expensive to exit a few months later. Nothing about the project needs to have changed for this to happen.

Early depth is usually purchased

Liquidity does not appear because a token exists. Someone has to place capital on both sides of the market and accept the risk of holding it.

New projects pay for that, either through emissions to liquidity providers or through arrangements with market makers who are compensated for quoting continuously.

The depth is real while it lasts, but it is a service being bought rather than a property the token has acquired.

Providers are exposed to the token they support

Supplying a pool means holding an inventory of the new token. If it falls, the position absorbs that decline directly.

Automated market maker mechanics compound the effect, because the pool sells the rising asset and accumulates the falling one without any decision being made.

Compensation has to exceed that exposure. When emissions taper or the token becomes more volatile, the arithmetic stops working and capital withdraws.

Thin markets become self-reinforcing

As depth falls, the same order size moves the price further, so traders arrive to worse execution and volume declines.

Lower volume means less fee income for whoever remains, which weakens the case for staying and thins the book further.

The spiral is why a token can go from actively traded to effectively illiquid over a period during which no bad news was published.

Fragmentation across venues makes it worse

Listing on many chains and exchanges looks like reach, but it divides the same limited capital into several shallow pools instead of one deep market.

Each venue then quotes worse prices than a single consolidated market would, and arbitrage between them extracts value from providers on both sides.

What durable liquidity depends on

Markets stay deep where there is genuine two-sided demand: people who want to hold the asset for a reason and people who need to sell it for a reason.

Protocols that own their liquidity outright rather than renting it remove the withdrawal risk, though they take on the price exposure themselves instead.

Either way, the useful question about a new listing is who is being paid to make the market and what happens to the book when that payment stops.