Advanced Kalshi Rule Examples

These are complete Kalshi custom strategies (platform: kalshi, strategy: custom) that use the advanced rule keys: composite orders, per-fill exits and rule-level schedules. Every complete example on this page passes the same checks as a Studio save, except the last, Rejected: quotes plus per-fill exits, which shows a combination the validator refuses and the error it gets.

Market tickers, prices, sizes and thresholds are illustrative only. They show how the keys work, not what to trade, and they are not recommendations. Backtest and paper trade any strategy before you deploy it live.

Paired YES and NO quote with a combined cap

Quotes both outcomes one cent above their best bids when the pair's combined cost stays at or under $0.98. The pair rests instead of being re-sent every loop. It is pulled from about 15 minutes before to 15 minutes after an FOMC or CPI release, in the last 10 minutes, and when other traders' bids push the combined best bids above 0.985.

version: 1
platform: kalshi
strategy: custom
strategy_name: CPI paired YES/NO maker with 98c cap
strategy_name_origin: generated
market:
  series_ticker: KXCPI
risk:
  max_position: 2
  max_portfolio_positions: 3
  price_floor: 0.03
  price_ceiling: 0.97
loop:
  interval: 15
edge:
  macro:
    provider: economic_calendar
    event: FOMC,CPI
    fields: [seconds_to_next, seconds_since_previous]
    refresh: 5m
rules:
  - name: announcement_blackout
    when:
      any:
        - {field: edge.macro.seconds_to_next, op: "<=", value: 900}
        - {field: edge.macro.seconds_since_previous, op: "<=", value: 900}
    action: cancel_all
  - name: cancel_near_close
    when:
      all:
        - {field: time_to_expiry, op: "<=", value: "10m"}
    action: cancel_all
  - name: pull_when_outbid
    when:
      all:
        - {field: position_size, op: "==", value: 0}
        - {field: order_count, op: ">", value: 0}
        - {field: paired_best_bid_sum, op: ">", value: 0.985}
    action: cancel_all
  - name: quote_pair
    when:
      all:
        - {field: position_size, op: "==", value: 0}
        - {field: order_count, op: "==", value: 0}
        - {field: paired_best_bid_sum, op: "<", value: 0.965}
    max_combined_price: 0.98
    orders:
      - side: yes
        action: buy
        size: 1
        price: {reference: best_bid, offset: 0.01}
        post_only: true
      - side: no
        action: buy
        size: 1
        price: {reference: best_bid, offset: 0.01}
        post_only: true

What to notice:

• order_count == 0 lets the pair rest and keep its queue place. The three cancel_all rules above it decide when it comes down.
• The quote gate reads the raw bids; max_combined_price re-checks the actual leg prices after rounding and clamping.
• pull_when_outbid can't tell your bids from other traders', so it watches the sum, which includes your own quotes. Pulling needs other bids to push the sum above 0.985.
• If only one leg fills, position_size > 0 stops re-quoting, and the other leg rests until announcement_blackout or cancel_near_close pulls it. This strategy has no sell rule, so a lone filled leg is held until the market settles. That is legging risk; add a sell_all rule if you want a way out.
• risk.max_entries_per_market is left out: each leg uses an entry, so a value of 1 would block every pair. Studio prefills Max contracts per order from the largest leg (1), and Max open contracts as risk.max_position (2), which covers both legs. The validator also checks that the two legs together fit max_position.
• Live, the calendar values refresh every 5 minutes and don't count down in between, so the announcement pull can start and end up to 5 minutes late. Backtests compute them exactly. For a wider live margin, raise the 900-second thresholds.
• Backtests run in the Kalshi maker replay and need economic-calendar data for the whole window.

NO bid one cent under the ask, weekday sessions

A single post-only NO bid one cent under the NO ask, only on weekdays from 13:00 to 24:00 UTC. It exits the whole position at a profit or stop level with sell_all.

version: 1
platform: kalshi
strategy: custom
strategy_name: NYC high temp NO bid one tick under the ask
strategy_name_origin: generated
market:
  series_ticker: KXHIGHNY
risk:
  max_position: 3
  price_floor: 0.05
  price_ceiling: 0.60
  max_entries_per_market: 3
loop:
  interval: 20
trading_schedule:
  timezone: UTC
  trading_hours:
    - days: [mon, tue, wed, thu, fri]
      start: "13:00"
      end: "24:00"
rules:
  - name: take_profit
    when:
      all:
        - {field: no_position_size, op: ">", value: 0}
        - {field: no_best_bid, op: ">=", value: 0.60}
    action: sell_all
  - name: stop_out
    when:
      all:
        - {field: no_position_size, op: ">", value: 0}
        - {field: no_best_bid, op: "<=", value: 0.25}
    action: sell_all
  - name: pull_quote_late
    when:
      all:
        - {field: time_to_expiry, op: "<=", value: "2h"}
    action: cancel_all
  - name: requote_when_ask_moves
    when:
      all:
        - {field: order_count, op: ">", value: 0}
        - {field: spread, op: ">", value: 0.015}
    action: cancel_all
  - name: quote_no_under_ask
    when:
      all:
        - {field: order_count, op: "==", value: 0}
        - {field: no_position_size, op: "<", value: 3}
        - {field: spread, op: ">", value: 0.015}
        - {field: no_best_ask, op: ">=", value: 0.35}
        - {field: no_best_ask, op: "<=", value: 0.50}
    orders:
      - side: no
        action: buy
        size: 1
        price: {reference: best_ask, offset: -0.01}
        post_only: true

What to notice:

• best_ask on a NO leg is the NO ask: 1 minus the best YES bid.
• While the bid rests one cent under the ask, the spread is exactly one cent. If the ask moves up, the spread widens, requote_when_ask_moves cancels, and the next loop bids one cent under the new ask.
• A top-level schedule is allowed with orders. When it closes at 24:00 UTC, the Bot cancels the resting bid. Exits still run at any hour.
• The exits use sell_all, not sell_yes or sell_no, so the strategy backtests the same way in the maker replay. Live, sell_all doesn't cancel the resting bid.

Per-fill 5-cent profit and 10-cent stop

Buys YES in a price band, then manages each fill on its own: sell a fill when the YES bid is 5 cents above what that fill cost, or stop it out when the bid is 10 cents below. This is the full rule order from Rule order and requirements.

version: 1
platform: kalshi
strategy: custom
strategy_name: BTC daily per-fill 5c target and 10c stop
strategy_name_origin: generated
market:
  series_ticker: KXBTCD
risk:
  max_position: 3
  price_floor: 0.20
  price_ceiling: 0.80
  max_entries_per_market: 3
loop:
  interval: 15
rules:
  - name: exit_all_near_close
    when:
      all:
        - {field: position_size, op: ">", value: 0}
        - {field: time_to_expiry, op: "<=", value: "10m"}
    action: sell_all
  - name: stop_each_yes_fill
    when:
      all:
        - {field: yes_position_size, op: ">", value: 0}
    action: sell_yes
    entry_stop_loss_offset: 0.10
  - name: take_profit_each_yes_fill
    when:
      all:
        - {field: yes_position_size, op: ">", value: 0}
    action: sell_yes
    entry_profit_offset: 0.05
  - name: enter_yes
    when:
      all:
        - {field: order_count, op: "==", value: 0}
        - {field: yes_best_ask, op: ">=", value: 0.30}
        - {field: yes_best_ask, op: "<=", value: 0.45}
        - {field: time_to_expiry, op: ">", value: "30m"}
        - {field: yes_position_size, op: "<", value: 3}
    action: buy_yes
    size: 1
  - name: cancel_unfilled_entry
    when:
      all:
        - {field: order_count, op: ">", value: 0}
    action: cancel_all

What to notice:

• Worked numbers: a 0.38 fill has a 0.43 target and a 0.28 stop trigger. A 0.42 fill has a 0.47 target and a 0.32 trigger. Each fill is managed at its own price, not the average.
• buy_yes places a limit at the ask. If it doesn't fill right away, cancel_unfilled_entry pulls it on the next loop, and the entry can fire again at the new ask.
• Each loop sells from at most one fill, capped at Max contracts per order, which Studio prefills as 1 here.
• Backtest with a taker backtest and a flat starting account. See Per-fill exits in backtests.

Entry-price profit table with a time gate

Buys YES between 85 and 88 cents. Each fill gets a target from the table, but only fills made more than 30 minutes before close. Every fill keeps an 8-cent stop.

version: 1
platform: kalshi
strategy: custom
strategy_name: Favorite scalper with price-mapped targets
strategy_name_origin: generated
market:
  series_ticker: KXBTCD
risk:
  max_position: 4
  price_floor: 0.80
  price_ceiling: 0.92
  max_entries_per_market: 4
loop:
  interval: 15
rules:
  - name: stop_each_fill_8c
    when:
      all:
        - {field: yes_position_size, op: ">", value: 0}
    action: sell_yes
    entry_stop_loss_offset: 0.08
  - name: table_targets_for_early_fills
    when:
      all:
        - {field: yes_position_size, op: ">", value: 0}
    action: sell_yes
    entry_profit_targets:
      - {entry_price: 0.85, target_price: 0.95}
      - {entry_price: 0.86, target_price: 0.95}
      - {entry_price: 0.87, target_price: 0.96}
      - {entry_price: 0.88, target_price: 0.97}
    entry_profit_min_time_to_expiry: 30m
  - name: buy_favorite
    when:
      all:
        - {field: order_count, op: "==", value: 0}
        - {field: yes_best_ask, op: ">=", value: 0.85}
        - {field: yes_best_ask, op: "<=", value: 0.88}
        - {field: yes_position_size, op: "<", value: 4}
    action: buy_yes
    size: 1
  - name: cancel_unfilled_entry
    when:
      all:
        - {field: order_count, op: ">", value: 0}
    action: cancel_all

What to notice:

• A fill at a price with no row, such as 0.84 after price improvement or a sub-cent 0.8501, is protected by the stop only.
• The 30-minute gate is judged when the fill happens. A fill at exactly 30 minutes before close is excluded. An earlier fill keeps its target through the final 30 minutes.
• Studio prefills Max open contracts from risk.max_position (4). A deploy with a lower Max open contracts stops the Bot below 4 contracts; a higher one still stops it at 4.

Time-varying EMA-distance entry thresholds

On weekdays, buys YES in the BTC daily series when spot BTC trades far enough above its 26-period, 5-minute EMA. How far is a strict threshold from 09:30 to 11:00 ET and a looser one from 11:00 to 16:00 ET. A short top-level blackout pauses entries around 10:00 ET. Per-fill exits run at all hours.

version: 1
platform: kalshi
strategy: custom
strategy_name: BTC EMA-distance entries with session thresholds
strategy_name_origin: generated
market:
  series_ticker: KXBTCD
risk:
  max_position: 2
  price_floor: 0.05
  price_ceiling: 0.90
  max_entries_per_market: 2
loop:
  interval: 15
edge:
  btc:
    provider: coinbase
    symbol: BTC-USD
    fields: [ema_26_5m_distance_pct]
trading_schedule:
  timezone: America/New_York
  blackouts:
    - days: [mon, tue, wed, thu, fri]
      start: "09:55"
      end: "10:05"
rules:
  - name: stop_each_fill
    when:
      all:
        - {field: yes_position_size, op: ">", value: 0}
    action: sell_yes
    entry_stop_loss_offset: 0.10
  - name: take_profit_each_fill
    when:
      all:
        - {field: yes_position_size, op: ">", value: 0}
    action: sell_yes
    entry_profit_offset: 0.08
  - name: open_strict_threshold
    trading_schedule:
      timezone: America/New_York
      trading_hours:
        - days: [mon, tue, wed, thu, fri]
          start: "09:30"
          end: "11:00"
    when:
      all:
        - {field: edge.btc.ema_26_5m_distance_pct, op: ">=", value: 0.004}
        - {field: order_count, op: "==", value: 0}
        - {field: yes_best_ask, op: "<=", value: 0.85}
        - {field: yes_position_size, op: "==", value: 0}
        - {field: time_to_expiry, op: ">", value: "1h"}
    action: buy_yes
    size: 1
  - name: midday_looser_threshold
    trading_schedule:
      timezone: America/New_York
      trading_hours:
        - days: [mon, tue, wed, thu, fri]
          start: "11:00"
          end: "16:00"
    when:
      all:
        - {field: edge.btc.ema_26_5m_distance_pct, op: ">=", value: 0.0025}
        - {field: order_count, op: "==", value: 0}
        - {field: yes_best_ask, op: "<=", value: 0.85}
        - {field: yes_position_size, op: "==", value: 0}
        - {field: time_to_expiry, op: ">", value: "1h"}
    action: buy_yes
    size: 1
  - name: cancel_unfilled_entry
    when:
      all:
        - {field: order_count, op: ">", value: 0}
    action: cancel_all

What to notice:

• ema_26_5m_distance_pct is a fraction: 0.004 means spot is 0.4% above the EMA. See Edge Data.
• Each entry needs its own conditions, its own window, and the top-level schedule. The 09:55 to 10:05 blackout cuts into the strict window.
• Every rule schedule names its own timezone; nothing is inherited from the top level.
• The Bot cancels pending entry orders at 09:30, 11:00 and 16:00, while the blackout is in force, and on restart. See Orders at window edges.
• If the Coinbase feed can't be fetched (beyond its staleness allowance), the Bot blocks new entries on that market and cancels its resting entry orders there; the stop and take-profit rules keep running. If only the EMA field is unavailable, for example after a late 5-minute candle, conditions on it are false: neither entry rule fires, and the stop and take-profit rules still run.
• yes_best_ask <= 0.85 matters: buy_yes bids at the ask, clamped to price_ceiling, so without it an ask of 0.95 would leave a resting 0.90 bid.

One target table for YES and NO fills

Buys whichever outcome is the 85 to 88 cent favorite, never both. Every YES and NO fill gets an 8-cent stop and one shared target table. Entries are allowed on weekdays only, through a reusable rule-level schedule made of blackouts.

version: 1
platform: kalshi
strategy: custom
strategy_name: Either-side favorite scalper with one shared target table
strategy_name_origin: generated
market:
  series_ticker: KXBTCD
risk:
  max_position: 2
  price_floor: 0.80
  price_ceiling: 0.92
  max_entries_per_market: 4
loop:
  interval: 15
rules:
  - name: stop_yes_fills_8c
    when:
      all:
        - {field: yes_position_size, op: ">", value: 0}
    action: sell_yes
    entry_stop_loss_offset: 0.08
  - name: stop_no_fills_8c
    when:
      all:
        - {field: no_position_size, op: ">", value: 0}
    action: sell_no
    entry_stop_loss_offset: 0.08
  - name: yes_table_targets
    when:
      all:
        - {field: yes_position_size, op: ">", value: 0}
    action: sell_yes
    entry_profit_targets: &favorite_targets
      - {entry_price: 0.85, target_price: 0.93}
      - {entry_price: 0.86, target_price: 0.94}
      - {entry_price: 0.87, target_price: 0.95}
      - {entry_price: 0.88, target_price: 0.96}
  - name: no_table_targets
    when:
      all:
        - {field: no_position_size, op: ">", value: 0}
    action: sell_no
    entry_profit_targets: *favorite_targets
  - name: buy_yes_favorite
    trading_schedule: &weekdays_only
      timezone: UTC
      blackouts:
        - days: [sat, sun]
          start: "00:00"
          end: "24:00"
    when:
      all:
        - {field: order_count, op: "==", value: 0}
        - {field: yes_best_ask, op: ">=", value: 0.85}
        - {field: yes_best_ask, op: "<=", value: 0.88}
        - {field: no_position_size, op: "==", value: 0}
        - {field: yes_position_size, op: "<", value: 2}
        - {field: time_to_expiry, op: ">", value: "30m"}
    action: buy_yes
    size: 1
  - name: buy_no_favorite
    trading_schedule: *weekdays_only
    when:
      all:
        - {field: order_count, op: "==", value: 0}
        - {field: no_best_ask, op: ">=", value: 0.85}
        - {field: no_best_ask, op: "<=", value: 0.88}
        - {field: yes_position_size, op: "==", value: 0}
        - {field: no_position_size, op: "<", value: 2}
        - {field: time_to_expiry, op: ">", value: "30m"}
    action: buy_no
    size: 1
  - name: cancel_unfilled_entry
    when:
      all:
        - {field: order_count, op: ">", value: 0}
    action: cancel_all

What to notice:

• For NO fills, the stop and table read the NO bid and match the NO fill price.
• &favorite_targets and &weekdays_only are YAML anchors; *favorite_targets and *weekdays_only reuse them.
• The opposite-side position checks stop a buy on one side from netting against contracts you hold on the other side. Kalshi nets YES and NO in one market, which would close your oldest lots at that price.

Rejected: quotes plus per-fill exits

Resting composite quotes can't share a strategy with per-fill exits, in any rule order and with a table or an offset (see What can be combined). Use per-fill exits with buy_yes/buy_no entries, or composite quotes with sell_all/cancel_all exits.

# Rejected: resting composite quotes cannot share a strategy with per-fill exits
# error: rules[1].orders: entry-relative exits do not support composite maker orders
version: 1
platform: kalshi
strategy: custom
strategy_name: CPI YES bid improver with per-fill take-profit
strategy_name_origin: generated
market:
  series_ticker: KXCPI
risk:
  max_position: 2
  price_floor: 0.03
  price_ceiling: 0.97
loop:
  interval: 15
rules:
  - name: take_profit_each_yes_fill
    when:
      all:
        - {field: yes_position_size, op: ">", value: 0}
    action: sell_yes
    entry_profit_offset: 0.04
  - name: quote_yes_bid
    when:
      all:
        - {field: position_size, op: "==", value: 0}
        - {field: spread, op: ">=", value: 0.03}
    orders:
      - side: yes
        action: buy
        size: 1
        price: {reference: best_bid, offset: 0.01}
        post_only: true