Skip to main content
Where does the money come from? Where does it go? How does Skepsis stay solvent?

The Simple Version

Money In: Traders pay for shares
Money Out: Winners get paid

The Protocol: Manages the flow, takes a small fee
That’s it. No complex token economics. No ponzinomics. Just predictions and payouts.

The Money Flow

Step 1: Market Creation

A market creator deposits initial liquidity:
Creator deposits: $10,000 USDC
This becomes: The prize pool
This liquidity enables trading from day one.

Step 2: Trading

Traders buy shares in ranges they believe will win:
Alice buys: $100 on range $96K-$97K (gets 400 shares)
Bob buys: $200 on range $97K-$98K (gets 500 shares)  
Charlie buys: $150 on range $95K-$96K (gets 800 shares)

Pool grows: $10,000 + $100 + $200 + $150 = $10,450

Step 3: Resolution

The outcome is revealed, winners claim:
Result: BTC = $97,200 (Bob's range wins!)

Bob's 500 shares Γ— $1 = $500 payout
Alice: $0 (wrong range)
Charlie: $0 (wrong range)

Bob's profit: $500 - $200 = $300 βœ…

Step 4: Remaining Balance

After all claims:
Pool started: $10,450
Bob claimed: $500
Pool remaining: $9,950

This stays in the system for:
- LP withdrawal
- Protocol fees
- Next market seeding

Where Winners’ Money Comes From

Source 1: Losing Bets

The primary source. When you lose, your bet goes to:
  1. Pay winners
  2. Protocol fees
  3. LP returns
Losers paid: $250 (Alice + Charlie)
Winner received: $500 (Bob)
Gap of $250 from: Initial liquidity

Source 2: Initial Liquidity

When losers don’t cover winners, the initial liquidity bridges the gap.
Initial LP: $10,000
Total bets: $450 (Alice $100, Bob $200, Charlie $150)
Pool: $10,450

If winners claim $500:
Losers contribute: $250
LP contributes: $250
Winner receives: $500

Source 3: LMSR Pricing Premium

The LMSR algorithm builds in a small edge for the market maker:
Sum of all range probabilities = 100%
Sum of implied prices = slightly > $1 (spread)

This spread = market maker's edge = system stays solvent

The Solvency Guarantee

Critical question: Can the market always pay winners? Answer: Yes, by mathematical design.

How LMSR Guarantees Solvency

Max possible payout = Total shares Γ— $1/share
Pool balance = Initial liquidity + All bets - Payouts

LMSR ensures: Pool balance β‰₯ Max possible payout (always)

The Alpha Parameter

The alpha (Ξ±) in LMSR is calibrated to ensure:
Max LP loss = 50% of initial liquidity

If creator deposits $10,000:
Worst case for LP: Lose $5,000
Best case for LP: Profit from fees and losers
This is a known, bounded risk.

Extreme Scenarios

What if everyone bets on the winning range?
All traders bet on Range C
Range C probability β†’ very high
Range C odds β†’ very low (close to 1x)

Result: Everyone wins... but wins very little
Pool still covers all payouts
The system self-balances.

Fee Structure

Trading Fees

A small fee on each transaction:
Fee: 0.3% (30 basis points)

On a $100 bet:
Fee: $0.30
Net bet: $99.70

Where Fees Go

Total fees collected
β”œβ”€β”€ Protocol treasury (30%)
β”œβ”€β”€ Market creator (30%)  
└── Liquidity providers (40%)

Why Fees Exist

  1. Protocol sustainability β€” Development, infrastructure, audits
  2. Creator incentives β€” Reward for creating good markets
  3. LP incentives β€” Return for providing liquidity

For Market Creators

The Investment

Create market with $10,000 liquidity
Risk: Lose up to 50% ($5,000) in worst case
Return: Share of trading fees + any remaining liquidity

The Return

Market generates $50,000 in trading volume
Fees at 0.3%: $150
Creator share (30%): $45

Plus: Return of remaining liquidity at market close

Risk/Reward

OutcomeCreator Result
Low volume, balanced betting~$0 profit, get liquidity back
High volume, balanced bettingFee profit + most liquidity back
Low volume, unbalanced bettingPossible small loss
High volume, unbalanced bettingFee profit may offset LP loss

For Traders

Your Money Flow

You bet $100:
β”œβ”€β”€ $99.70 goes to buy shares
└── $0.30 goes to fees

If you win:
You receive: Shares Γ— $1/share = deterministic payout

If you lose:
Your $99.70 pays other winners

Expected Value

EV = (Your probability estimate Γ— Payout) - Cost

Positive EV = Good bet
Negative EV = Bad bet

Example:
Market says: 25% (4x odds)
You think: 35%
EV = (0.35 Γ— $400) - $100 = +$40 βœ…

Economic Incentives Summary

For Traders

  • Bet early β†’ Better odds (prices haven’t moved)
  • Bet against crowd β†’ Higher potential returns
  • Be right β†’ Profit

For Market Creators

  • Create interesting markets β†’ More volume β†’ More fees
  • Set appropriate ranges β†’ Better UX β†’ More bets
  • Bootstrap liquidity β†’ Attract early traders

For the Protocol

  • More markets β†’ More fees
  • Better UX β†’ More users
  • Solvency β†’ Trust β†’ Growth

Comparison: Where Money Flows

Traditional Betting (Casino/Sportsbook)

You bet $100
House edge: 5-10%
House always profits
Zero-sum for bettors

Prediction Markets (Polymarket-style)

You bet $100 against other users
Platform takes ~2% fee
Winners take from losers
Platform always profits

Skepsis

You bet $100 against the market state
LMSR prices your bet fairly
If you have edge, you profit
Platform takes 0.3% fee
LPs take bounded risk for fee share
Key difference: Skepsis rewards accurate predictions, not just being on the winning side of a 50/50.

FAQ

”What if no one bets on the winning range?”

The initial liquidity pays winners. This is by designβ€”LPs take this risk in exchange for fees.

”Can the market run out of money?”

No. LMSR mathematically guarantees solvency. The alpha parameter ensures max payout ≀ pool balance.

”Why would anyone provide liquidity?”

Expected return = Fees earned - LP losses With balanced betting and sufficient volume, this is positive.

”Is this sustainable?”

Yes. The 0.3% fee covers:
  • Protocol costs
  • Creator rewards
  • LP incentives
No external token subsidies needed.

Next Steps

Trust & ResolutionHow outcomes are determined fairlyTrust & Resolution
LMSR ExplainedThe math behind solvencyLMSR Explained