File size: 1,572 Bytes
d4c51f2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
data_dir: data

discover:
  use_leaderboard: true
  leaderboard_windows: [all, 30d]
  leaderboard_top_n: 150
  use_market_holders: true
  markets_to_scan: 25
  holders_per_market: 25
  seed_wallets: []

analyze:
  max_wallets: 100
  activity_page_limit: 500
  max_pages_per_wallet: 15
  request_sleep_sec: 0.15

# Exploration thresholds. RELAXED vs the shipped defaults because capped history
# (~3500 recent events) + the win-rate/resolution heuristic make these metrics
# unreliable — so ranking leans on authoritative leaderboard profit + copyability
# + repeatability, which ARE trustworthy. Tighten before trusting for real money.
screen:
  min_total_pnl_usd: 10000       # authoritative profit; leaderboard whales exceed easily
  min_roi: 0.03                  # reconstructed ROI is noisy -> low bar
  min_win_rate: 0.50             # heuristic-based -> modest bar
  min_resolved_markets: 8        # heuristic under-counts resolved markets
  min_active_days: 20            # truncated history shrinks the observed span
  max_drawdown_pct: 1.5          # reconstructed drawdown is noisy -> loose
  min_monthly_profit_ratio: 0.40
  min_repeatability: 0.30        # observed max ~0.43 on the sample
  max_rebate_share: 0.35         # reject market-maker / rebate-driven profit
  top_n_shortlist: 25

backtest:
  our_bankroll_usd: 800
  fee_bps: 0
  slippage_bps: 150
  copy_delay_sec: 60
  sizing:
    max_frac_per_trade: 0.05
    max_frac_per_market: 0.15
    min_frac_per_trade: 0.005
    w_stake: 0.6
    w_edge: 0.4
    kelly_cap: 0.5
    assumed_max_edge: 0.15