Citadel Sector Pods AI Trading Team¶
This AI team is inspired by the “pod” setup used by big multi-manager hedge funds like Ken Griffin’s Citadel. Instead of asking one AI to understand the whole market, five specialist agents each cover one part of it and pitch their best sector ETF. A risk manager pushes back, and only the portfolio manager places trades. It shows how specialist agents can check each other before money moves.
How it works¶
Five sector agents each study one area: technology and communications, financials, healthcare, energy, and consumer. Each pitches its strongest idea from the ETF universe.
Risk manager agent reads all five pitches and challenges them: crowded trades, big drops, macro risk, and sudden reversals.
Portfolio manager agent builds a portfolio across at least three sectors and is the only agent allowed to place orders. The team repeats this once a day.
The copyable example now defaults to GPT-6 Luna with medium reasoning. A leveraged version holds leveraged sector ETFs instead. Earlier BotSpot paper observations used Gemini; changing the model does not change those historical results.
Run it on BotSpot¶
Run this team on BotSpot without installing anything. BotSpot runs LumiBot in the cloud, backtests it, and connects it to your broker.
Backtest tear sheet¶
GPT-6 Luna, January 5 to 16, 2026, Yahoo daily prices, $100,000 start. The team rotated across sector ETFs such as XLC, XLE, XLF, XLI, XLV, and XLB and ended at $101,680 (+2%) while SPY rose 1%. Cash stayed near zero, lowest $33 in the daily stats. Earlier BotSpot paper observations used Gemini and are separate from these Luna backtest results and the current Luna example.
Open the full tear sheet. A short backtest shows the team works as written. It is not a promise of future returns.
The code¶
Regular sector ETFs:
"""Citadel / Surveyor-inspired sector-pod AI trading team example.
Regular sector ETF data-on variant. Uses LumiBot's default built-in tools,
including FRED/ALFRED macro tools when FRED_API_KEY is supplied, Alpaca News
when ALPACA_NEWS_API_KEY / ALPACA_NEWS_API_SECRET are supplied, SEC tools,
market state, account state, and order tools. It preserves the sector-pod debate
structure but requires a diversified final allocation.
"""
import os
from lumibot.credentials import IS_BACKTESTING
from lumibot.entities import Asset, TradingFee
from lumibot.strategies.strategy import Strategy
from lumibot.traders import Trader
ORDER_READINESS_RULE = (
"Immediately before every buy or sell order, call account_portfolio, "
"account_positions, and market_last_price for the exact ordered symbol in "
"this same agent run, then call orders_submit_order. LumiBot rejects blind "
"orders with ORDER_READINESS_REQUIRED when those readiness calls are missing."
)
DATA_USAGE_RULE = (
"Use official LumiBot built-in tools by their real names: get_fred_snapshot, "
"get_fred_latest, get_fred_series, and list_fred_series for macro evidence; "
"alpaca_news for recent market, sector, rates, and ETF-proxy headlines; SEC "
"tools are available and optional when company or sector fundamentals matter. "
"Keep tool use bounded: at most one FRED snapshot or short series request and "
"one Alpaca News call per agent run; set Alpaca News limit <= 5; do not paginate "
"or repeatedly re-check the same evidence. Do not rely on a custom public CSV FRED helper."
)
DIVERSIFICATION_RULE = (
"Preserve the sector-pod process and convert it into a diversified portfolio. "
"Hold at least three sector ETFs in normal conditions. Do not buy a single "
"sector ETF with nearly all capital unless the only non-sector allocation is "
"cash-like SHV due to explicit risk-off evidence."
)
class AITradingTeamCitadelSectorPodsStrategy(Strategy):
parameters = {
"universe": [
"XLK", "XLF", "XLV", "XLE", "XLY", "XLI", "XLP", "XLU", "XLB", "XLRE", "XLC", "SHV",
],
"min_positions": 3,
}
def initialize(self):
self.sleeptime = "1D"
model = os.environ.get("AI_TRADING_TEAM_MODEL", "openai/gpt-6-luna")
self.agents.create(
name="technology_pod",
model=model,
allow_trading=False,
system_prompt=(
"Rank technology and communications sector ETFs. First call alpaca_news for XLK/XLC/QQQ/SMH-relevant headlines and call get_fred_snapshot or get_fred_latest for rates, growth, and liquidity context. "
"Use exact symbols from the universe only. " + DATA_USAGE_RULE
),
)
self.agents.create(
name="financials_pod",
model=model,
allow_trading=False,
system_prompt=(
"Rank financial and rate-sensitive sector ETFs. First call get_fred_snapshot or get_fred_latest for yield curve, credit, liquidity, and policy-rate context, then call alpaca_news for financial-sector headlines. "
"Use exact symbols from the universe only. " + DATA_USAGE_RULE
),
)
self.agents.create(
name="healthcare_pod",
model=model,
allow_trading=False,
system_prompt=(
"Rank healthcare and defensive growth sector ETFs. Call alpaca_news for healthcare/biotech/defensive-growth headlines and use FRED tools for macro risk context. "
"Use exact symbols from the universe only. " + DATA_USAGE_RULE
),
)
self.agents.create(
name="energy_pod",
model=model,
allow_trading=False,
system_prompt=(
"Rank energy and commodity-sensitive sector ETFs. Call alpaca_news for oil/energy headlines and get_fred_snapshot/get_fred_latest for inflation, rates, dollar, and growth context. "
"Use exact symbols from the universe only. " + DATA_USAGE_RULE
),
)
self.agents.create(
name="consumer_pod",
model=model,
allow_trading=False,
system_prompt=(
"Rank consumer discretionary, staples, and housing-sensitive sector ETFs. Call alpaca_news for consumer/housing headlines and FRED tools for inflation, income, rates, and growth context. "
"Use exact symbols from the universe only. " + DATA_USAGE_RULE
),
)
self.agents.create(
name="risk_manager",
model=model,
allow_trading=False,
system_prompt=(
"Compare the pod picks. Challenge crowding, factor exposure, drawdown risk, reversal risk, and macro contradictions. "
"Check whether pods actually used get_fred_snapshot/get_fred_latest and alpaca_news. Do not re-call tools unless pod evidence is entirely absent; if you must, make only one short FRED call and one Alpaca News call with limit <= 5. "
+ DATA_USAGE_RULE + " " + DIVERSIFICATION_RULE
),
)
self.agents.create(
name="portfolio_manager",
model=model,
allow_trading=True,
system_prompt=(
"Allocate across the best sector ETFs from the pod process. Build a diversified portfolio of at least three positions in normal conditions, using variable weights based on pod conviction and risk-manager objections. "
"Do not make a one-sector all-in trade. Reconcile any override of pod advice or risk-manager warnings. "
+ DATA_USAGE_RULE + " " + DIVERSIFICATION_RULE + " " + ORDER_READINESS_RULE
),
)
def on_trading_iteration(self):
context = {
"date": self.get_datetime().date().isoformat(),
"universe": self.parameters["universe"],
"min_positions": self.parameters["min_positions"],
"data_expectation": "Use get_fred_snapshot/get_fred_latest and alpaca_news commonly; smoke tests inspect agent_detail for actual tool calls.",
"portfolio_expectation": "Portfolio manager should normally hold at least three sector ETFs and avoid single-sector all-in behavior.",
"data_tool_validation_run_id": "2026-07-08-fresh-alpaca-news-fred-smoke-v1",
}
technology = self.agents["technology_pod"].run(
task_prompt="Call alpaca_news and FRED tools, then rank technology/communications sector opportunities.",
context=context,
)
financials = self.agents["financials_pod"].run(
task_prompt="Call get_fred_snapshot or get_fred_latest and alpaca_news, then rank financial/rate-sensitive sector opportunities.",
context=context,
)
healthcare = self.agents["healthcare_pod"].run(
task_prompt="Call alpaca_news and FRED tools, then rank healthcare/defensive-growth sector opportunities.",
context=context,
)
energy = self.agents["energy_pod"].run(
task_prompt="Call alpaca_news and FRED tools, then rank energy/commodity-sensitive sector opportunities.",
context=context,
)
consumer = self.agents["consumer_pod"].run(
task_prompt="Call alpaca_news and FRED tools, then rank consumer/housing-sensitive sector opportunities.",
context=context,
)
risk = self.agents["risk_manager"].run(
task_prompt="Compare all pod picks, challenge concentration/crowding/macro risks, and recommend a diversified 3+ sector allocation.",
context={**context, "technology": technology.summary, "financials": financials.summary, "healthcare": healthcare.summary, "energy": energy.summary, "consumer": consumer.summary},
)
self.agents["portfolio_manager"].run(
task_prompt=(
"Rebalance into the best diversified 3+ sector ETF portfolio. Do not sell everything into one strongest ETF. "
"Before each order, call account_portfolio, account_positions, and market_last_price for the exact ordered symbol in this same run. "
"Explain which pod advice you accepted or rejected and cite FRED/Alpaca evidence used."
),
context={**context, "technology": technology.summary, "financials": financials.summary, "healthcare": healthcare.summary, "energy": energy.summary, "consumer": consumer.summary, "risk": risk.summary},
)
if __name__ == "__main__":
quote_asset = Asset("USD", Asset.AssetType.FOREX)
params = AITradingTeamCitadelSectorPodsStrategy.parameters
if IS_BACKTESTING:
trading_fee = TradingFee(percent_fee=0.001)
AITradingTeamCitadelSectorPodsStrategy.backtest(
datasource_class=None,
benchmark_asset=Asset("SPY", Asset.AssetType.STOCK),
buy_trading_fees=[trading_fee],
sell_trading_fees=[trading_fee],
quote_asset=quote_asset,
parameters=params,
)
else:
trader = Trader()
strategy = AITradingTeamCitadelSectorPodsStrategy(
quote_asset=quote_asset,
parameters=params,
)
trader.add_strategy(strategy)
trader.run_all()
Leveraged sector ETFs:
"""Citadel / Surveyor-inspired sector-pod AI trading team example.
Leveraged sector ETF data-on variant. Uses LumiBot's default built-in tools,
including FRED/ALFRED macro tools when FRED_API_KEY is supplied, Alpaca News
when ALPACA_NEWS_API_KEY / ALPACA_NEWS_API_SECRET are supplied, SEC tools,
market state, account state, and order tools. The trade universe is 2x/3x
leveraged sector and broad ETFs plus SHV/cash as a rare escape hatch.
"""
import os
from lumibot.credentials import IS_BACKTESTING
from lumibot.entities import Asset, TradingFee
from lumibot.strategies.strategy import Strategy
from lumibot.traders import Trader
ORDER_READINESS_RULE = (
"Immediately before every buy or sell order, call account_portfolio, "
"account_positions, and market_last_price for the exact ordered symbol in "
"this same agent run, then call orders_submit_order. LumiBot rejects blind "
"orders with ORDER_READINESS_REQUIRED when those readiness calls are missing."
)
DATA_USAGE_RULE = (
"Use official LumiBot built-in tools by their real names: get_fred_snapshot, "
"get_fred_latest, get_fred_series, and list_fred_series for macro evidence; "
"alpaca_news for recent market, sector, rates, and ETF-proxy headlines; SEC "
"tools are available and optional when company or sector fundamentals matter. "
"Keep tool use bounded: at most one FRED snapshot or short series request and "
"one Alpaca News call per agent run; set Alpaca News limit <= 5; do not paginate "
"or repeatedly re-check the same evidence. Do not rely on a custom public CSV FRED helper."
)
LEVERAGE_RULE = (
"This is a leveraged ETF strategy. Use only symbols from the leveraged universe plus SHV/cash-like exposure. "
"Stay biased toward diversified 3x exposure when evidence supports risk-taking; use 2x as a risk-down choice; "
"use inverse leveraged ETFs only with explicit downside or hedge evidence. Hold at least three positions in normal conditions, "
"and do not make a single-sector all-in bet. SHV/cash may count as one position when the model refuses every reasonable leveraged setup."
)
class AITradingTeamCitadelSectorPodsStrategy(Strategy):
parameters = {
"universe": [
"TECL", "TECS", "SOXL", "SOXS", "FNGU", "FNGD", "WEBL", "WEBS",
"FAS", "FAZ", "DPST", "WDRW", "CURE", "RXD", "LABU", "LABD",
"ERX", "ERY", "GUSH", "DRIP", "WANT", "NEED", "RETL", "DUSL",
"SIJ", "MATL", "UTSL", "SDP", "DRN", "DRV", "NAIL", "UPRO",
"SPXU", "SSO", "SDS", "TQQQ", "SQQQ", "QLD", "QID", "TNA",
"TZA", "SHV",
],
"min_positions": 3,
}
def initialize(self):
self.sleeptime = "1D"
model = os.environ.get("AI_TRADING_TEAM_MODEL", "openai/gpt-6-luna")
self.agents.create(
name="technology_pod",
model=model,
allow_trading=False,
system_prompt=(
"Rank leveraged technology, communications, internet, AI, and semiconductor ETFs. First call alpaca_news for TECL/SOXL/FNGU/WEBL proxy headlines and call get_fred_snapshot or get_fred_latest for rates, growth, and liquidity context. "
"Use exact symbols from the universe only. " + DATA_USAGE_RULE + " " + LEVERAGE_RULE
),
)
self.agents.create(
name="financials_pod",
model=model,
allow_trading=False,
system_prompt=(
"Rank leveraged financial, bank, and rate-sensitive ETFs. First call get_fred_snapshot or get_fred_latest for yield curve, credit, liquidity, and policy-rate context, then call alpaca_news for financial-sector headlines. "
"Use exact symbols from the universe only. " + DATA_USAGE_RULE + " " + LEVERAGE_RULE
),
)
self.agents.create(
name="healthcare_pod",
model=model,
allow_trading=False,
system_prompt=(
"Rank leveraged healthcare, biotech, and defensive-growth ETFs. Call alpaca_news for healthcare/biotech/defensive-growth headlines and use FRED tools for macro risk context. "
"Use exact symbols from the universe only. " + DATA_USAGE_RULE + " " + LEVERAGE_RULE
),
)
self.agents.create(
name="energy_pod",
model=model,
allow_trading=False,
system_prompt=(
"Rank leveraged energy, oil, commodity, and materials ETFs. Call alpaca_news for oil/energy/materials headlines and get_fred_snapshot/get_fred_latest for inflation, rates, dollar, and growth context. "
"Use exact symbols from the universe only. " + DATA_USAGE_RULE + " " + LEVERAGE_RULE
),
)
self.agents.create(
name="consumer_pod",
model=model,
allow_trading=False,
system_prompt=(
"Rank leveraged consumer discretionary, staples, retail, housing, industrials, utilities, and real estate ETFs. Call alpaca_news for consumer/housing/defensive-sector headlines and FRED tools for inflation, income, rates, and growth context. "
"Use exact symbols from the universe only. " + DATA_USAGE_RULE + " " + LEVERAGE_RULE
),
)
self.agents.create(
name="risk_manager",
model=model,
allow_trading=False,
system_prompt=(
"Compare the pod picks. Challenge crowding, factor exposure, drawdown risk, reversal risk, macro contradictions, inverse ETF decay, and single-sector concentration. "
"Check whether pods actually used get_fred_snapshot/get_fred_latest and alpaca_news. Do not re-call tools unless pod evidence is entirely absent; if you must, make only one short FRED call and one Alpaca News call with limit <= 5. "
+ DATA_USAGE_RULE + " " + LEVERAGE_RULE
),
)
self.agents.create(
name="portfolio_manager",
model=model,
allow_trading=True,
system_prompt=(
"Allocate across the best leveraged ETFs from the pod process. Build a diversified portfolio of at least three positions in normal conditions, using variable weights based on pod conviction and risk-manager objections. "
"Favor 3x ETFs for high-conviction sector sleeves, use 2x when conviction or drawdown risk is lower, and use inverse ETFs only as a carefully justified hedge or downside exposure. "
"Do not make a one-sector all-in trade. Reconcile any override of pod advice or risk-manager warnings. "
+ DATA_USAGE_RULE + " " + LEVERAGE_RULE + " " + ORDER_READINESS_RULE
),
)
def on_trading_iteration(self):
context = {
"date": self.get_datetime().date().isoformat(),
"universe": self.parameters["universe"],
"min_positions": self.parameters["min_positions"],
"data_expectation": "Use get_fred_snapshot/get_fred_latest and alpaca_news commonly; smoke tests inspect agent_detail for actual tool calls.",
"leverage_expectation": "Use only leveraged ETFs plus SHV/cash escape, hold at least three positions, and avoid single-sector all-in behavior.",
"data_tool_validation_run_id": "2026-07-08-fresh-alpaca-news-fred-smoke-v1",
}
technology = self.agents["technology_pod"].run(
task_prompt="Call alpaca_news and FRED tools, then rank leveraged technology/communications/semiconductor opportunities.",
context=context,
)
financials = self.agents["financials_pod"].run(
task_prompt="Call get_fred_snapshot or get_fred_latest and alpaca_news, then rank leveraged financial/rate-sensitive opportunities.",
context=context,
)
healthcare = self.agents["healthcare_pod"].run(
task_prompt="Call alpaca_news and FRED tools, then rank leveraged healthcare/biotech/defensive-growth opportunities.",
context=context,
)
energy = self.agents["energy_pod"].run(
task_prompt="Call alpaca_news and FRED tools, then rank leveraged energy/commodity/materials opportunities.",
context=context,
)
consumer = self.agents["consumer_pod"].run(
task_prompt="Call alpaca_news and FRED tools, then rank leveraged consumer/housing/defensive/real-estate opportunities.",
context=context,
)
risk = self.agents["risk_manager"].run(
task_prompt="Compare all pod picks, challenge concentration/crowding/macro/inverse-decay risks, and recommend a diversified 3+ leveraged ETF allocation.",
context={**context, "technology": technology.summary, "financials": financials.summary, "healthcare": healthcare.summary, "energy": energy.summary, "consumer": consumer.summary},
)
self.agents["portfolio_manager"].run(
task_prompt=(
"Rebalance into the best diversified 3+ leveraged ETF portfolio. Bias toward 3x ETFs, use 2x only as a risk-down choice, and use inverse ETFs only with explicit hedge/downside evidence. "
"Do not sell everything into one strongest ETF. Before each order, call account_portfolio, account_positions, and market_last_price for the exact ordered symbol in this same run. "
"Explain which pod advice you accepted or rejected and cite FRED/Alpaca evidence used."
),
context={**context, "technology": technology.summary, "financials": financials.summary, "healthcare": healthcare.summary, "energy": energy.summary, "consumer": consumer.summary, "risk": risk.summary},
)
if __name__ == "__main__":
quote_asset = Asset("USD", Asset.AssetType.FOREX)
params = AITradingTeamCitadelSectorPodsStrategy.parameters
if IS_BACKTESTING:
trading_fee = TradingFee(percent_fee=0.001)
AITradingTeamCitadelSectorPodsStrategy.backtest(
datasource_class=None,
benchmark_asset=Asset("SPY", Asset.AssetType.STOCK),
buy_trading_fees=[trading_fee],
sell_trading_fees=[trading_fee],
quote_asset=quote_asset,
parameters=params,
)
else:
trader = Trader()
strategy = AITradingTeamCitadelSectorPodsStrategy(
quote_asset=quote_asset,
parameters=params,
)
trader.add_strategy(strategy)
trader.run_all()
Run it yourself¶
pip install lumibot
python -m lumibot.example_strategies.ai_trading_team_citadel_sector_pods
Add OPENAI_API_KEY for GPT-6 Luna and your broker keys (for example ALPACA_API_KEY, ALPACA_API_SECRET, and ALPACA_IS_PAPER=true) to your .env file. This file trades by default. Set IS_BACKTESTING=true in your environment to backtest it instead. The default is openai/gpt-6-luna. Set AI_TRADING_TEAM_MODEL only when deliberately choosing a different model, with its matching provider key.
Good to know¶
Inspired by public descriptions of multi-manager hedge funds. Not affiliated with or endorsed by Citadel or its people, and not a copy of any real strategy.
The picture shows the five sector agents side by side because none of them depends on another. The code calls them one after another.
See AI Trading Bot Examples for more AI trading bots and Backtest, paper, or live: choose the runner for backtest and live runs.