ADDENDUM: Actionable Edge Mapping + Prediction Market Ladder

Wallet Intelligence Desk — MERGED FINAL

Base: Claude Opus 4.6 | Panel additions from: Gemini Pro, GPT-4.1, Grok 4.1 Fast

DeepSeek V3: EXCLUDED (generation failure — truncated/garbled output)

This addendum supplements the original Wallet Intelligence spec. It does NOT replace the existing spec — it adds two critical layers: (1) mapping every edge to a specific trade on a specific platform, and (2) a systematic methodology for scanning prediction market ladders using wallet-derived intelligence.

TOPIC 1: ACTIONABLE EDGE OUTPUT

1. Edge-to-Trade Mapping Table

2. Actionable Output Format

JSON output sent via Telegram and to other desks:

{
  "edge_id": "WALLET-20260315-FOLLOW-001",
  "desk": "wallet_intelligence",
  "timestamp": "2026-03-15T15:22:00Z",
  "edge_description": "Sharp wallet 0xabc...789 (score 82) bought $25K YES on 'Fed cuts in June' at 32c on Polymarket. Cross-platform check: Kalshi same contract at 28c.",
  "signal_type": "sharp_wallet_follow",
  "wallet": {
    "address": "0xabc...789",
    "composite_score": 82,
    "win_rate": 0.68,
    "brier_score": 0.18,
    "total_trades": 342,
    "roi_90d": 0.23,
    "category_specialty": ["politics", "economics"],
    "recent_accuracy_30d": 0.72
  },
  "market": {
    "event": "Fed cuts rate in June 2026",
    "polymarket_id": "0x1234...5678",
    "kalshi_ticker": "KXFED-20260618-CUT25"
  },
  "platforms": {
    "Kalshi": {
      "action": "Buy YES on 'Fed cuts 25bp in June'",
      "current_price": 28,
      "whale_entry_price_poly": 32,
      "cross_platform_edge": 4,
      "edge_after_fees": 2,
      "liquidity_usd": 800,
      "max_suggested_size_usd": 400
    },
    "Polymarket": {
      "action": "Buy YES on 'Fed cuts in June'",
      "current_price": 33,
      "note": "Price has already moved 1c from whale entry (32 → 33). Slippage is real.",
      "liquidity_usd": 2500,
      "max_suggested_size_usd": 500
    }
  },
  "time_sensitivity": "MEDIUM — whale just entered. Price will likely move another 2-5c over the next 1-4 hours as other followers detect the trade. Kalshi may take 6-24 hours to adjust (slower information flow cross-platform).",
  "slippage_estimate": {
    "polymarket_expected_slippage": "1-2c for $500 order",
    "kalshi_expected_slippage": "0-1c for $400 order",
    "fee_kalshi": 1.5,
    "fee_polymarket_gas": 0.3
  },
  "confidence_level": 75,
  "risk_factors": [
    "Wallet may be wrong — 68% win rate means 32% loss rate",
    "Fed cut probability is fundamentally uncertain",
    "Cross-platform divergence may persist (settlement rule differences)",
    "Whale could exit early if trade moves against them"
  ]
}

Telegram summary format:

WALLET EDGE: Sharp Wallet Follow

Wallet 0xabc...789 (score 82, 68% win rate)
Bought $25K YES 'Fed cuts June' @ 32c (Polymarket)

Kalshi: Same contract @ 28c | Cross-platform edge = 4c
Action: Buy YES Kalshi @ 28c (cheaper) | $800 liq
Action: Buy YES Polymarket @ 33c (already moved 1c) | $2500 liq

Wallet stats: 342 trades | 0.18 Brier | +23% ROI (90d)
Confidence: 75% | Window: 1-4 hours before followers push price

3. Platform Priority Ranking

Summary: The Wallet Intelligence desk is primarily a POLYMARKET operation (data source + execution) with KALSHI as the highest-value secondary venue (cross-platform edge). Both platforms are equal-priority for ladder scanning — neither is secondary to the other. Traditional platforms (crypto, stocks) are used when a wallet signal has implications beyond prediction markets.

TOPIC 2: PREDICTION MARKET LADDER STRUCTURE

1. Desk Data Mapping to Prediction Market Contracts

The Wallet Intelligence desk operates ACROSS all prediction market categories. Its data maps to every contract type, because wallets trade all categories. The desk's unique contribution is BEHAVIORAL probability — not model-based, but track-record-based.

2. Sharp Reference Identification

The Wallet Intelligence desk's "sharp reference" is BEHAVIORAL, not model-based. Instead of using options chains or economic models, this desk uses wallet track records as the probability reference.

3. Probability Curve Construction Method

The Wallet Intelligence desk constructs probability estimates differently from other desks — using BEHAVIORAL data rather than statistical models.

Method: Wallet-Weighted Probability Estimate (WWPE)

For a given prediction market event:

1. Identify all sharp wallets with positions on this event. Pull from wallet_trades table.

2. For each wallet, extract their implied probability:

   • If they bought YES at 35c, their implied probability of YES is ≥35% (they wouldn't buy if they thought fair value was below 35c).
   • Adjust upward: sharp wallets typically buy when they see >5c edge. So a 35c entry implies they think fair value is ~40-45c.
   • Weight by: (a) wallet composite score, (b) category-specific accuracy, (c) position size relative to their typical size, (d) recency of the trade.

3. Combine wallet estimates:

   • WWPE = Σ (wallet_i_implied_probability * wallet_i_weight) / Σ (wallet_i_weight)
   • Where weight = composite_score * category_accuracy * size_signal * recency_decay

4. Cross-check against model-based fair value:

   • If WWPE = 45% and the Options Desk BL-derived probability = 42%, they agree. High confidence.
   • If WWPE = 45% and BL says 30%, investigate the divergence. Either wallets know something the options market doesn't, or the wallets are wrong.

5. Apply to the full ladder: If WWPE for the main rung is 45%, construct a probability curve around it:

   • Assume the wallet-implied distribution has the same shape as the model-based distribution (from the relevant desk — Options for stocks, Deribit for crypto, NWS for weather) but SHIFTED so the probability at the main rung matches WWPE.
   • This gives wallet-informed probabilities at every rung of the ladder.

Distribution type by market category:

• BTC/ETH price (short horizon): Log-normal calibrated to Deribit implied volatility and skew.
• Stablecoin peg [PANEL: GPT-4.1]: Empirical distribution based on historical depeg frequency and Curve/DEX pool imbalance data. Update every 5 minutes — stablecoin volatility can spike fast.
• DAO governance votes [PANEL: GPT-4.1]: Binomial distribution based on on-chain voting power alignment of known wallet clusters.
• Airdrop dump probability [PANEL: Gemini Pro]: Empirical — use historical dump rates from prior airdrop cohorts as the base rate, adjusted by confirmed farming wallet consolidation count.

Recalculation frequency:

• On every new sharp wallet trade: immediately recalculate WWPE for that event.
• Every 4 hours: recalculate all WWPEs to account for wallet exits and new entries.
• On wallet exit signal: immediately recalculate.
• Every 60 seconds for crypto price ladders with active Deribit calibration (spot price can move fast).
• Every 5 minutes for stablecoin depeg markets [PANEL: GPT-4.1].
• On each on-chain vote tally shift >5% [PANEL: GPT-4.1].

4. Ladder Scanning Methodology

Step-by-step process:

1. Identify high-interest events: Events where (a) at least one sharp wallet has a position, OR (b) cross-platform divergence >3c exists between Kalshi and Polymarket, OR (c) volume spike detected on either Kalshi or Polymarket, OR (d) airdrop farmer consolidation signal detected [PANEL: Gemini Pro], OR (e) on-chain DAO vote tally is shifting [PANEL: GPT-4.1].

2. For each high-interest event, pull the full ladder from BOTH Kalshi AND Polymarket. Many events have ladder structures on both platforms (e.g., "BTC above $90K", "$95K", "$100K", "$105K" — each is a separate contract on each platform). Neither platform's ladder is secondary — scan both completely.

3. Calculate fair probability at each rung using WWPE (wallet-weighted method) cross-checked against the relevant desk's model-based probability.

4. Check cross-platform pricing at each rung on both Kalshi and Polymarket: For every rung available on both platforms, compare prices. Flag divergences >3c after fees. The cheaper platform is the buy; if the expensive platform allows selling, it's the other leg of an arb.

5. Calculate edge at each rung:

   • wallet_edge = WWPE - contract_price (for YES direction)
   • cross_platform_edge = |kalshi_price - polymarket_price| - total_fees_both_sides
   • combined_edge = max(wallet_edge, cross_platform_edge)

6. Flag tail opportunities: Rungs far from the main line (tail thresholds) are where:

   • Sharp wallets rarely trade (less information → less efficiently priced)
   • Cross-platform divergences are largest (fewer arbitrageurs at tails)
   • Model-based probabilities from other desks are most valuable (e.g., Options Desk BL curve for deep OTM stock thresholds)

7. Filter: Minimum 3c combined edge, minimum $50 liquidity on at least one platform.

8. Rank: Sort by combined_edge * confidence * liquidity_score.

9. Output to all desks: When the Wallet Intelligence desk finds an edge on a sports contract, alert the Sports Desk. When it finds an edge on a crypto contract, alert the Crypto Desk. When it finds a stablecoin depeg signal, alert the Crypto Desk and reduce Polymarket exposure across all open positions (USDC depeg affects all Polymarket settlement). The Wallet desk is a SIGNAL MULTIPLIER for all other desks.

Additional scan trigger — ETF rebalancing window [PANEL: Grok 4.1 Fast]: When known fund-linked wallets are detected exiting a position, immediately scan Kalshi for any equity price ladder on that ETF or stock. If no Kalshi contract exists, route directly to the stock/options broker as the execution venue.

5. Edge Examples with Real Numbers

Example 1: Sharp Wallet Follow + Cross-Platform Divergence on Crypto

• Event: "BTC above $100K by March 31, 2026"
• Wallet signal: Wallet 0xabc (score 82, 71% crypto accuracy) buys $30K YES at 38c on Polymarket.
• Wallet-weighted implied probability: Entry 38c + estimated 5c edge premium = fair value ~43c.
• Cross-platform check: Kalshi "BTC above $100K" (March 31) at 33c. Polymarket ladder also has "$95K" and "$105K" rungs — scan all of them on Kalshi as well.
• Divergence: Polymarket 39c (post-whale movement) vs Kalshi 33c = 6c divergence.
• Crypto Desk model reference: Deribit BL-derived probability at $100K for March 31 = 40%.
• Convergence of signals: Wallet says ~43%, Deribit says 40%, Kalshi at 33c. All agree Kalshi is underpriced.
• Raw edge on Kalshi: 40% - 33% = 7c (conservative, using Deribit as anchor). Wallet-informed: 43% - 33% = 10c.
• Net edge: 7-10c - 1.5c fee = 5.5-8.5c
• Trade: Buy YES on Kalshi "BTC above $100K" at 33c. Edge = 5.5-8.5c.
• Cross-desk alert: Alert Crypto Desk: "Sharp wallet + Deribit + cross-platform divergence all point to BTC $100K being underpriced on Kalshi."

Example 2: Dumb Wallet Fade on Sports

• Event: "Lakers vs Celtics total points Over 228.5" — Polymarket
• Wallet signal: Wallet 0xdef (score 19, 31% win rate, 0.48 Brier, 180 trades, confirmed dumb after Bonferroni correction p<0.005) buys $8K YES on Over 228.5 at 52c.
• Historical: When this wallet buys YES on NBA totals, the NO side wins 63% of the time.
• Sports Desk reference: Pinnacle no-vig total is 227.0. P(total > 228.5) with Normal(227, 14) = P(Z > 0.107) = 45.7%.
• Fade signal probability: Wallet's 63% fade accuracy applied to fair 45.7% → adjusted fair value of NO = 63% * 54.3% + 37% * 45.7% = 51.1%. The dumb wallet signal barely moves the needle here.
• Better rung: Check the tail. "Over 242.5" — dumb wallet also has a separate $3K YES at 18c. Sports Desk model says P(>242.5) with Normal(227, 14) = P(Z > 1.107) = 13.4%. Dumb wallet fade says NO probability is ~87% * 1.10 = ~89%. Fair NO = 89c. Check both Kalshi and Polymarket for this rung — buy whichever has the lower NO price. If Kalshi NO at 82c and Polymarket NO at 80c, buy Polymarket first. Edge = 7-9c net after fees.
• Trade: Buy NO on "Over 242.5 total points" on the cheaper platform. The dumb wallet's tail bet is more fadeable than their main line bet.

Example 3: Wallet Consensus on Economic Event + Cross-Platform

• Event: "Fed cuts rate in September 2026"
• Wallet signals:
  • Wallet A (score 78, econ specialty): $15K YES at 44c on Polymarket
  • Wallet B (score 71, econ specialty): $20K YES at 46c on Polymarket
  • Wallet C (score 80, generalist): $10K YES at 43c on Polymarket
• WWPE calculation:
  • Wallet A: implied ~49c (entry 44c + 5c edge premium), weight = 78 * 0.72 * 1.0 = 56.2
  • Wallet B: implied ~51c, weight = 71 * 0.65 * 1.2 = 55.4
  • Wallet C: implied ~48c, weight = 80 * 0.60 * 0.8 = 38.4
  • WWPE = (4956.2 + 5155.4 + 48*38.4) / (56.2+55.4+38.4) = 49.5%
• CME FedWatch reference: 52% probability of September cut.
• Convergence: Wallets say ~49.5%, FedWatch says 52%. Both significantly above current prices.
• Cross-platform: Polymarket "Fed cuts Sep" now at 47c (moved up from wallets' entries). Kalshi same contract at 40c. Both ladders should be scanned for adjacent rungs (e.g., "cuts 50bp", "cuts 25bp", "no cut").
• Kalshi edge: FedWatch-anchored fair value 52%, Kalshi at 40c. Edge = 12c (10.5c net).
• Trade: Buy YES on Kalshi "Fed cuts in September" at 40c. Triple confirmation: 3 sharp wallets + FedWatch + WWPE all agree.
• Cross-desk alert: Alert Forex Desk (cut = USD weakness → long EUR/USD), Futures Desk (cut = bond rally → long ZN), Stocks Desk (cut = equity bullish → long ES).

Example 4: Stablecoin Depeg as Tradeable Edge [PANEL: GPT-4.1]

• Event: USDT elevated depeg risk detected via Curve pool imbalance
• Signal: USDT reserves in Curve 3pool drop 40% in 2 hours, USDT price on DEXes at $0.996.
• Historical depeg probability: Empirical from prior Curve imbalance events of this magnitude = 22%.
• Ladder scan (both platforms):
  • Kalshi "USDT < $0.98 by EOW": 12c. Edge = 22% - 12% = 10c raw, 8.5c net.
  • Polymarket "USDT depeg below $0.98": 14c (if listed). Edge = 22% - 14% = 8c raw, 6c net.
  • Buy Kalshi first (cheaper). Buy Polymarket as second leg if liquidity allows.
• Risk note: A Polymarket YES position denominated in USDC has compound risk here — if USDT depegs, USDC may follow. Monitor USDC/USDT rate in real time and reduce total Polymarket exposure if USDC itself shows stress.

Example 5: Airdrop Farmer Consolidation Dump [PANEL: Gemini Pro]

• Event: $WORM token airdrop, day of listing
• Signal: System identifies 500 wallets flagged as known airdrop farmers, all sending $WORM allocations to a single new address which deposits 2M $WORM to Binance within 30 minutes of listing.
• Historical dump rate: Prior airdrop cohorts with >300 confirmed farming wallets consolidating same-day show 3%-10% price drop in first 24-48 hours in 78% of cases.
• Trade: Short WORM-PERP on Binance immediately. Also check Polymarket for any "WORM price above listing price by Day 7" contract — if listed, buy NO.
• Kalshi check: Unlikely to have a WORM-specific contract, but if a broader "new token outperforms" contract exists, evaluate for short exposure.
• Exit: Cover position by 48-hour mark regardless of P&L — the initial dump signal loses predictive value after the first farming wave has settled.

IMPORTANT NOTES

• The Wallet Intelligence desk is a SIGNAL MULTIPLIER — its outputs enhance every other desk. Every wallet-derived edge should be routed to the relevant domain desk for cross-validation.
• Both Kalshi and Polymarket have full ladder structures with multiple contracts at different thresholds. Neither platform is secondary to the other. Kalshi and Polymarket are equal-status execution venues. Always scan both completely. Always check the other platform when a volume spike or price move occurs on either one. Cross-platform divergence is bidirectional.
• Cross-platform divergence detection is the quickest win. No wallet analysis needed — just price comparison. Validate this signal first (Phase 0) before investing in wallet tracking infrastructure.
• Dumb wallet fading requires rigorous statistical validation (Bonferroni correction for multiple comparisons, minimum 100 trades, p<0.01). Do NOT fade wallets just because they have a sub-50% win rate — that could be random.
• Settlement rule differences between Kalshi and Polymarket mean that cross-platform "arbitrage" is not always risk-free. Build the settlement_rules table and verify before trading.
• USDC depeg risk affects all Polymarket trades. Monitor USDC/USDT rate (from Crypto Desk shared service) and reduce Polymarket exposure if depeg risk rises. When a stablecoin depeg signal fires, treat all open Polymarket positions as having elevated tail risk. [PANEL: GPT-4.1]
• The 6-month shadow mode is non-negotiable. Do not deploy real capital on wallet signals until the scoring system has been validated on 200+ settled predictions.
• WebSocket reliability for Polymarket is the #1 operational risk. Build the REST polling fallback FIRST and add WebSocket as an enhancement.
• Airdrop farmer wallet identification requires maintaining a separate flagged-wallet database. Update this database continuously — farming wallets are created fresh for each airdrop cycle. [PANEL: Gemini Pro]
• DAO governance tracking requires on-chain data (Etherscan, Snapshot API). The vote tally is live and changes as wallets vote — build a real-time polling loop for active governance markets, not a one-time snapshot. [PANEL: GPT-4.1]
• ETF rebalancing signals are time-constrained. The window between detecting smart-money exit and the rebalancing execution is typically hours to 1-2 days. Automate detection; manual review is too slow. [PANEL: Grok 4.1 Fast]