Chapter 15 · What Liquidity Means
How can a $100,000 sale crush a billion-dollar token?
- Capabilities practiced
- Financial LiteracyOnchain LiteracyResearch
- Hands on
- Pick a small-cap token, calculate the price impact of selling 5% of circulating supply, and show your work.
- AI Lab
- Have AI explain the difference between market cap and liquidity. Make it calculate from real pool data and check whether it treats FDV as realizable value.
A Real-World Problem
A market website reports an $800 million capitalization.
You own 0.5% of the token supply. The interface helpfully values your holdings at $4 million.
You decide to exit, enter the full amount, and confirm.
The transaction succeeds. You receive $610,000.
Where did the other $3.39 million go? No theft, contract exploit, or incorrect slippage setting occurred. That $3.39 million never existed.
Market capitalization multiplies the last traded price by circulating supply. It means: if everyone could sell at the last price, the total would be $800 million. Everyone being able to sell at that price has never been a valid assumption.
The question is specific: before pressing sell, how can you determine what you will actually receive?
Thought Experiment
A town's houses total $100 million in value. Its only pawnshop has $500,000 in its cash drawer today.
How much is the town worth?
$100 million is reasonable if houses sell gradually over time. $500,000 is also correct if everyone needs cash this afternoon: that is all the cash available.
These differ by 200 times while answering different questions: estimated value at the last price versus cash immediately accessible.
Replace the pawnshop with an on-chain pool and residents with token holders. On-chain rules make the situation even stricter:
The money on the pool's other side is the maximum cash this token can release now. Not a penny more.
Whether you sell a million or a hundred million tokens, $500,000 is the ceiling. Larger sales approach it but never reach it: constant product requires infinite tokens to drain every dollar.
Now turn this into a calculation.
Your Decision
You hold a small-cap token with a paper value of $400,000 and want to exit entirely.
What Happens
Build intuition with the simulator. Set pool depth to minimum and trade size to maximum, then observe the average execution deviation.
No market maker sets this quote. The x·y=k curve determines the price. A shallower pool and a larger purchase move your price further from the market price; arbitrageurs and liquidity providers capture that difference.
Notice halving depth can more than double the price movement: the relationship is nonlinear. And no input size reaches the ceiling on assets available for withdrawal.
For a constant-product pool, both have a simple expression:
Let x = token reserves in units, not dollars
y = dollar reserves on the other side
D = tokens sold into the pool
r = D ÷ x (sale size relative to token reserves)
After the sale:
Price as a fraction of the original = 1 ÷ (1 + r)²
Dollars received = y × r ÷ (1 + r)These two expressions serve the entire chapter. Note the square: price responds to order size through a squared relationship. That creates the nonlinearity.
This mathematical table does not expire:
| Sale relative to token reserves | Remaining price | Price decline | Share of dollar reserves received |
|---|---|---|---|
| 10% (r = 0.1) | 82.6% | 17.4% | 9.1% |
| 25% (r = 0.25) | 64.0% | 36.0% | 20.0% |
| 50% (r = 0.5) | 44.4% | 55.6% | 33.3% |
| 100% (r = 1) | 25.0% | 75.0% | 50.0% |
| 200% (r = 2) | 11.1% | 88.9% | 66.7% |
| 500% (r = 5) | 2.8% | 97.2% | 83.3% |
| 1000% (r = 10) | 0.8% | 99.2% | 90.9% |
The last column matters most: you can never withdraw all the money on the other side. Even selling ten times the token reserve releases only 90.9%.
How does a $100,000 sale crush a billion-dollar token?
| Token A | Token B | |
|---|---|---|
| Circulating supply | 1 billion | 1 billion |
| Last price | $1 | $1 |
| Market capitalization | $1 billion | $1 billion |
| Pool token reserves | 5 million | 50,000 |
| Pool dollar reserves | $5 million | $50,000 |
| Decline after selling $100,000 nominal | 3.9% | 88.9% |
| Cash from selling 5% of circulation | $4.55 million, nominally $50 million | $49,900, nominally $50 million |
The website makes them look identical. Their only difference is the two reserve rows the website does not show.
Token B illustrates the title: $50,000 of pool capital supports a billion-dollar headline. This is common among newly issued tokens.
Build a Model
Convert this into four actions. This is the one procedure to take away, reusable for any token in three minutes.
- Find pool token reserves x
- Find circulating supply S
- Calculate r = sale units ÷ x
- Read the decline and cash proceeds from the table
The details:
| Step | Action | Common mistake |
|---|---|---|
| 1. Find x | Open the pool address in an explorer and read token units | Confusing x with TVL, which values both sides in dollars |
| 2. Find S | Use circulating supply, not total or maximum supply | Total supply exaggerates the sellable fraction |
| 3. Calculate r | Token units to sell ÷ x | Dividing dollars by token units mixes dimensions |
| 4. Read results | Use the table or calculate 1 ÷ (1+r)² | Assuming execution at the quoted market price |
Two real-world variations: with multiple pools and chains, adding x across pools is an optimistic approximation and actual results are worse; with an order book, the formula does not apply directly, but the action remains: sum buy orders down to your price limit to find available money, your y.
Separate three commonly confused numbers:
| Number | Answers | Does not answer | Used by |
|---|---|---|---|
| Market Cap | Estimated circulating value at the last price | Whether you can sell | Rankings, media, size comparisons |
| FDV | Value at current prices if every token circulated | Tokens may not yet exist in the market | Promotion and valuation comparisons |
| Liquidity / Depth | Cash accessible now and its cost | Long-term value | People entering or exiting positions |
The first two are estimates; the last is a fact. Chapter 19 examines FDV. Remember: FDV prices tokens not yet in circulation and is furthest from cash you can recover.
What It Is Called
The ability to buy or sell an asset without materially changing its price.
Both limited price change and actual tradability are necessary. A house can be valuable but illiquid; a $100 note has modest value and excellent liquidity.
The volume a market can absorb within a price range.
For an AMM, this is pool reserves; for an order book, cumulative posted quantities. It is liquidity's directly measurable form, including the x you can use in calculations.
The gap between the best buy and sell quotes: the immediate cost of buying and selling back.
Spread measures small-trade cost; depth measures large-trade cost. They can diverge: a 0.1% spread with only $50,000 in a pool looks cheap for small trades but cannot absorb large ones.
The difference between average execution and the price seen when ordering.
Distinguish price impact: slippage concerns your average execution, while impact concerns the pool's new final price. The latter is larger because execution travels along the curve and averages between its endpoints.
Total traded value over a period.
It is often a proxy for liquidity, but volume describes the past; depth exists now. Self-trading can manufacture volume without adding depth. Divide 24-hour volume by pool depth; an implausibly large ratio deserves investigation.
Last traded price × circulating supply.
It is an estimate for comparing asset scale, not recoverable cash. Confusing these is the chapter's primary target.
Hands-on
Choose a token worth tens to hundreds of millions. Huge tokens give less dramatic results; tiny ones lack data. No funds or wallet connection are required.
Find its main pool. From the website's trading pairs, choose the highest-volume pair and obtain its pool contract address.
Count all its pools and chains too.
Find x and y. Open the pool address and inspect its asset balances.
- x = token units
- y = dollar value of the other asset, converting ETH or similar assets at current prices if necessary
Record the time; reserves change every minute.
Find S. Use circulating supply, not total supply. Confusing them can make conclusions overly optimistic.
Calculate r and consult the table.
Sale units D = S × 5%
r = D ÷ x
Remaining price fraction = 1 ÷ (1 + r)²
Actual cash = y × r ÷ (1 + r)
Nominal value = D × current priceRecord the calculation and a conclusion.
| Value | |
|---|---|
| Pool token reserve x | |
| Pool capital y | |
| Circulating supply S | |
| 5% of supply = D | |
| r = D ÷ x | |
| Price decline | |
| Nominal value | |
| Actual cash | |
| Actual cash ÷ nominal value |
Repeat for 1%. Replace 5% and observe the decline.
A roughly proportional reduction means r remains small and splitting helps. Little improvement, such as 99% falling to 89%, means depth is inadequate for the position. Splitting cannot rescue it; reconsider the position itself.
The final ratio is the output. Remember its scale; market-cap headlines will look different afterward.
AI Lab
Use two stages: concepts first, then real data. Most errors occur in the second.
Stage 1:
Use a numerical example to distinguish Market Cap, FDV, and Liquidity.
Explain why the first two are not cash holders can recover.
Stage 2:
Token [name], pool [address], chain [name].
1. Find current reserves on both sides, with sources and timestamps.
2. List circulating and total supply separately.
3. Calculate price declines and dollar proceeds from selling 1%, 5%, and 10%
of circulating supply.
4. Compare nominal position value and actual proceeds, including their ratio.
5. State the formula and substituted numbers for every step.
Write "not found" for unavailable data. Do not estimate or use remembered data.Then repeat the previous Lab's four actions and compare every line.
The intended trap is that a model may use market cap × 5% as recoverable cash or FDV to value a position. Both sound plausible; manual calculations reveal the mistake.
Models also learn many claims of adequate token liquidity. Ask for evidence and often find volume, rather than depth, behind them.
Verify these points yourself after using AI
- Can you verify its reserves at the cited explorer address?
- Did it use circulating or total supply? The mistake can distort conclusions severalfold.
- Did it treat market cap or FDV as cash available for sale?
- Does its price decline follow the squared relationship when sale size changes?
- Does its liquidity claim rely on volume or depth? If volume, how does it rule out wash trading?
- Did it add pools across chains without describing that as an optimistic approximation?
- Does every number have an observation time? Reserves change continuously.
Real-World Case
Repeated self-trading with the same funds can inflate daily volume while leaving depth unchanged.
Divide volume by depth. Healthy ratios often range from single digits to tens. A pool holding tens of thousands but reporting millions of daily volume needs explanation.
Volume is promotional material; depth is your exit route.
Some token liquidity comes from agreements between projects and one or two market makers. Expiry, termination, or institutional trouble can remove depth rapidly.
Prices may look normal beforehand: removing depth creates no trade; it makes the next trade expensive. Recheck depth regularly, not only upon entry.
During severe declines, LPs and market makers withdrew or reduced quotes together. Depth shrank within hours: more sellers faced less buying capital. An order normally causing a 5% drop might cause 30% that day.
Research implication: calm-market depth systematically understates stressed exit costs. Apply a conservative haircut and test the conclusion again.
Projects repeatedly release a small supply fraction and fund pools for only part of it. A tiny pool sets price, while capitalization and FDV extend it across supply.
Hundreds of thousands in reserves can support hundreds of millions in headline valuation. The arithmetic can be correct while its interpretation is wrong.
Chapters 19 and 20 revisit the structure through token supply.
Change One Variable
Depth appears five times larger, but needs discounting.
Assets start on one chain, bridging takes time and money, and each sale needs a separate transaction. Arbitrage also spreads your first sale's decline to other chains: you compete with your next order.
Calculate recoverable cash on each chain separately, total it, and discount it. Fragmented liquidity cannot simply be added.
FDV exceeds market cap by more than 30 times.
The future unlock schedule is the risk: the remaining 97% arrives later. Today's tiny pool cannot absorb supply many times larger.
Divide each unlock by current x and read the impact table. Chapter 19 builds a full unlock curve.
Depth rises and your exit cost immediately falls.
Ask: whose money is it, and why is it there? Subsidized money leaves when incentives stop. Project-owned funds can be withdrawn anytime unless locked, which you can check on-chain.
Depth provided by others can be reclaimed by them. Chapter 17 studies subsidy lifetimes.
Single-order impact falls substantially.
Two risks appear: time exposure, including price changes and others anticipating visible sales; and the need for arbitrageurs to restore reserves, funded by actual external buyers. Without buying demand, recovery slows each time.
Splitting reduces impact but cannot create buyers. Calculate liquidity before entry, not only at exit.
Takeaway Questions
Make this executable: who supplies liquidity, how much, and under which conditions will it leave? Check balances on-chain and incentives or locks in their arrangements. Without all three answers, market cap is of little use to you.
Insufficient depth hurts those exiting last. Earlier sellers get better prices and later ones progressively worse. This queue explains why everyone recognizes liquidity risk only afterward.
If a token goes to zero, sellers drain the other asset and leave LPs holding the worthless token. Liquidity disappears with price, often faster. Both belong to the same feedback loop.
Subsidies stop, LPs leave, depth falls, slippage rises, and traders leave. Liquidity arrives early for subsidies and departs early too. Ask this separately for every liquidity-dependent product.
Chapter Check
Market Cap is an estimate from multiplication, assuming all circulating tokens can sell at the last price.
Liquidity is an observable fact: actual money currently available in pools or order books to absorb your sale.
Capitalization is a quoted valuation; liquidity is recoverable cash. The former appears on websites; the latter requires investigation.
Constant-product price is a reserve ratio, and sales enlarge one reserve while shrinking the other: numerator and denominator move in opposite directions.
The resulting 1 ÷ (1+r)² relationship contains the square creating nonlinearity.
Initially a full pool absorbs sales with little movement. Later, more tokens compete for less cash, making each additional sale harder. The path is a curve rather than a straight line.
It demands explanation: a ratio of 250 means every reserve dollar turns over 250 times daily.
Possibilities include wash trading, major depth elsewhere on other chains or centralized exchanges, or extremely frequent arbitrage.
None allows volume to substitute for exit-depth calculations. Inspect reserves and perform the four actions.
No. It only indicates cheap small trades.
Spread measures the nearest price level; depth measures cumulative quantity. Tight quotes can contain only a few thousand dollars each.
Under stress, market makers may keep spreads tight while withdrawing quantity. Quotes remain visible but executable size disappears. Check depth first, then spread.
At minimum:
- List all pools, chains, reserves x and y, and observation times.
- Run the four actions for sales of 1%, 5%, and 10% of circulating supply.
- Compare volume with depth and investigate unusually high ratios.
- Trace liquidity ownership to dispersed LPs, projects, or market makers, and inspect locks.
- Check upcoming unlocks over 12 months relative to current reserves.
- Apply a haircut and repeat, because exit-day depth is often worse.
Finally, state uncertainty clearly. An honest missing-data finding is more useful than an invented conclusion.
Key idea
Market cap is an estimate; liquidity is the money you can actually get back.