Order-book liquidity often looks cheaper because market makers recycle capital and hedge elsewhere. AMM liquidity pre-pays depth by locking collateral in a pool with bounded but real subsidy risk.
This closing chapter quantifies the trade-off—essential for operators, LPs, and serious traders—and connects capital math to maker models versus pool models before Module 03.
Two ways to buy depth
A CLOB market maker posts capital on both sides, turns inventory over, and hedges on correlated markets. Capital velocity is high; loss is limited by skill and hedges. An AMM operator or LP locks capital in a pool with parameter b. Capital is stationary; fees compensate partially; worst-case loss is bounded by LMSR math but can still hurt.
Informative prices need real aggregation with skin in the game. Capital efficiency asks: how much skin per dollar of honest, tradable liquidity?
Subsidy as a line item
For LMSR-style markets, worst-case market-maker loss scales roughly with b (times a log term on outcome count). To offer “one percent slippage on $5k trades,” you might need a large b—implying a large contingency reserve. That reserve is dollars locked per dollar of daily volume—a ratio operators hate and traders love.
Sketch: operator chooses b = 500 for a smooth binary curve; worst-case subsidy band is on the order of hundreds to low thousands before fees. Supporting $5k trades with ~1% impact may still need six-figure TVL in practice because LPs demand fee yield versus drawdown. Locked capital ÷ daily volume is the efficiency ratio.
Return on liquidity
Rough mental model: return ≈ (fees × volume) − adverse selection losses − opportunity cost of capital.
AMMs with thin volume fail when fees do not cover subsidy opportunity cost. CLOB makers exit when adverse selection dominates spread capture. A marquee election hybrid might sustain high fees and manageable adverse selection; a long-tail prop with big b and $3k/day volume bleeds lockup; a meme spike with small b can crush LPs.
Worked example: is this listing worth it?
Assume $200k effective capital locked (pool plus b subsidy), $3k/day volume, 1% pool fee. Fee income might be on the order of $900/month; opportunity cost at 5% annual on $200k is roughly $833/month; one bad headline adverse-selection hit can erase a quarter. Economics are fragile—sunset the listing, cut b, or add a maker program. Scale platforms push volume to marquee books; regulated venues pay makers on hits rather than idle pools.
Comparison snapshot
| Metric | CLOB | AMM |
|---|---|---|
| Capital per $1m daily volume | Lower (turnover) | Higher (pool) |
| Cold-start cost | High (pay makers) | Medium (set b) |
| Scalability to headline events | Excellent if makers committed | Expensive without hybrid |
Regulated book vs pool capital story
Designated makers post capital that recycles; depth scales via rebates and competition. LPs and protocol subsidy post capital that often sits stationary; depth scales via b, TVL, or added book. The same 70% display on a book may imply large notional near touch from recycled hedges; on a $70k TVL pool it may imply a curve price—only the first is likely a crowd price.
Why hybrids are capital hacks
Hybrids keep moderate b for backstop only, deploy maker rebates on high-volume contracts, and migrate capital from pool to paid CLOB as maturity proves. Week one might be b = 80 and $40k TVL without makers; week six $2m/day with makers at touch and cut pool risk; week twelve marquee behaves like a book with pool only on Sunday gaps—that path is improving efficiency over time, not a static design pick.
Opportunity cost in plain terms
Locked pool capital could sit in Treasuries, staking, or other markets. Operators implicitly pay that foregone yield while waiting for volume. A contract doing $500/day with $300k locked is a museum piece, not a business—sunset it even if the percent looks exciting on Twitter.
Cross-margin and advanced topics
Large operators eventually cross-margin correlated events so maker capital stretches further. That is beyond this module, but it explains why mature books look “cheaper” in capital terms than raw TVL suggests. AMMs rarely get the same cross-margin benefits; their capital sits in the pool until withdrawn.
LP vs professional maker
Retail DeFi LPs earn fees pro-rata and may not hedge. Professional book makers pull quotes in chaos. “Pool LP” is not synonymous with liquidity without APY and drawdown history.
Trader implications
When pool TVL is $20k and headline says 70%, capital efficiency is terrible—odds are not deep; your $3k order is the market. When a regulated book shows $2m at best bid/ask, efficiency is good for retail takers up to limits. The forecast you trade is not the same quality even if resolution text matches.
Sustainable venues optimize subsidy per contract, sunset illiquid AMM listings, and pay makers only where volume justifies. Unsustainable: infinite listing with fixed b and no volume.
APY marketing versus drawdown reality
DeFi interfaces highlight fee APY; prediction pools should also show worst-case resolution drawdown and peak inventory skew. A pool can earn fees all month then lose on one correlated election sweep. Traders treating LP shares like savings products misunderstand the risk; operators who hide drawdown history misunderstand incentives.
What traders should cite publicly
When you publish analysis, pair engine type with liquidity evidence: “CLOB mid 54¢ (2¢ spread, $40k at touch)” vs “pool 54% ($180k TVL, 1.2% impact on $2k).” Capital efficiency is the bridge between mechanics and credibility—without it, Module 03’s stories about manipulation and arbitrage lack a budget line.
Sunsetting and listing discipline
Capital efficiency improves when operators delist dead contracts and recycle subsidy to marquees. Traders should notice delistings—sometimes the last trade on a zombie market is an illusion of liquidity. Sustainable catalogs are smaller than maximal catalogs.
Module 02 synthesis and handoff
You now have the full stack: CLOB mechanics, AMM and LMSR, trade flow and liquidity, choice framework, hybrids, and capital. Module 03 asks who trades and why—game theory, manipulation costs, and arbitrage that keep these engines honest. Capital efficiency sets manipulation budget: moving a deep Senate book costs more than moving a $15k TVL pool five points. Cross-market arb fails when capital cannot flow—pools and books then diverge for weeks, not seconds.
Closing metaphor
Capital efficiency is the difference between a price that prints and a price that means something to a crowd. Underfund liquidity and you buy the print; overfund it and the operator buys it for you.
CLOB depth is rented—makers can leave. AMM depth is prepaid—capital sits until someone withdraws it. Traders consume whichever is cheaper all-in for their ticket; operators fund whichever ROI survives audit. A market that looks deep on a banner but fails the TVL-to-volume test is consuming someone else's locked capital—often yours if you are the next taker.
Summary sentence
Ask how many dollars lock per dollar traded before you call a market deep—engines differ, economics do not.
Key ideas
AMM depth is pre-funded; CLOB depth is rented from makers. b and TVL translate into locked capital. Low volume plus large b is bad economics and fragile odds. Hybrids and maker programs exist to fix capital efficiency at scale.
You finished Module 02 — Market Mechanisms & Trading Infrastructure.
Carry capital literacy into Module 03: manipulation and arbitrage are cheaper when liquidity is mislabeled as deep. The mechanics module ends here; the strategic module begins with who actually trades these markets and why.