Portfolio rebalance

Portfolio Rebalance

Purpose: Generate a rebalancing trade list to align current holdings with a target allocation strategy.

Trigger: User asks "rebalance my portfolio", "reallocate to equal weight", "shift to risk parity", or uses /portfolio-rebalance.

Data Flow

1. Load Portfolio & Determine Target Strategy

• Check <current_portfolio> first — skip tool call if present.
• Otherwise: load_portfolio(portfolio_id).
• Extract holdings, cash, total portfolio value.

Target Strategy:

• If user specifies "equal weight" or "equal allocation" → equal weight across all holdings.
• If user specifies "risk parity" or "risk-balanced" → call risk_parity_weights().
• If user specifies "dynamic" or "Kelly" → call dynamic_position_size() for each holding.
• If user provides custom targets → parse them (e.g. "40% AAPL, 30% MSFT, 30% cash").
• Default if unspecified: call request_user_input("Which rebalancing strategy?", "Equal weight|Risk parity|Dynamic (Kelly)|Custom targets").

2. Fetch Live Quotes

• For each ticker, call GLOBAL_QUOTE(ticker) to get current price.
• Compute current market value and weight % for each holding.
• Batch in parallel — single plan step for all quotes.

3. Compute Target Weights

Equal Weight:

target_weight = 1 / number_of_holdings (excluding cash)

Risk Parity:

• Fetch historical price data for volatility: PRICE_HISTORY(ticker, period="1y") for each holding.
• Compute annualised volatility for each ticker.
• Call risk_parity_weights(volatilities_json) → returns target weights inverse to volatility.

Dynamic (Kelly):

• For each holding: call dynamic_position_size(ticker, expected_return, volatility, sharpe).
• User must provide expected returns or you must compute from analyst targets.
• Sum to 1.0 and normalise.

Custom:

• Parse user-provided target weights and validate they sum to ≤ 100% (remainder = cash).

4. Compute Trade List

For each holding:

current_value = shares × current_price
target_value = total_portfolio_value × target_weight
delta_value = target_value - current_value
delta_shares = delta_value / current_price

If |delta_shares| < 1, skip (no trade needed).

Output:

• BUY {ticker}: +N shares @ ${price} = ${value}
• SELL {ticker}: -N shares @ ${price} = ${value}

5. Rebalancing Cost Estimate

• Call transaction_cost_estimate(trade_list_json) to compute:
  • Total trade value.
  • Bid-ask spread impact (estimate 0.1% for liquid stocks, 0.5% for illiquid).
  • Estimated commission (if applicable).
  • Total rebalancing cost.

6. Render Output

Current vs Target Table — use render_table():

Ticker	Current Weight %	Target Weight %	Delta	Trade
AAPL	48.6%	33.3%	-15.3%	SELL 8 shares
MSFT	49.9%	33.3%	-16.6%	SELL 5 shares
TSLA	0.0%	33.3%	+33.3%	BUY 10 shares
Cash	1.6%	0.0%	-1.6%	Deploy $300

Trade Summary:

• Total trades: 3
• Total buy value: $X
• Total sell value: $Y
• Estimated cost: $Z (W% of portfolio)
• Net cash impact: ${buy_value - sell_value}

7. Thesis Upsert — MANDATORY

This is an advanced skill → append the rebalancing plan to the portfolio thesis. ⚠️ Extended thinking is discarded — copy the complete plan into the thesis field; it is the ONLY output that reaches the Thesis panel.

Thesis structure:

## Portfolio Rebalance — {Date}

**Strategy**: {Equal Weight / Risk Parity / Dynamic / Custom}

{Current vs Target table}

### Trade List
1. SELL 8 shares of AAPL @ $185.00 = $1,480
2. SELL 5 shares of MSFT @ $380.00 = $1,900
3. BUY 10 shares of TSLA @ $175.00 = $1,750

**Estimated Cost**: ${Z} ({W}% of portfolio)  
**Net Cash Impact**: ${X} (deploy from cash / add to cash)

**Rationale**: {One sentence — e.g. "Reallocating to equal weight to reduce concentration risk."}

8. Chat Response

State in chat field:

• Rebalancing strategy used.
• Number of holdings before and after.
• Number of trades required.
• Estimated cost ($ and % of portfolio).
• Top rebalancing move (e.g. "Largest change: reduce MSFT by 16.6%").
• Tools used: load_portfolio, GLOBAL_QUOTE, risk_parity_weights (if used), transaction_cost_estimate, render_table.

Cost Controls

• Skip load_portfolio() if <current_portfolio> is present.
• Batch GLOBAL_QUOTE() calls — single plan step.
• Skip PRICE_HISTORY() if not doing risk parity — equal weight requires no historical data.
• Reuse session cache — don't refetch quotes already in <acquired_data>.

Error Handling

• If user specifies invalid custom targets (not summing to ≤100%), call request_user_input("Invalid target weights. Provide new targets or choose a strategy:", "Equal weight|Risk parity|Cancel").
• If GLOBAL_QUOTE(ticker) fails, exclude that ticker from rebalancing but note it in chat.
• If risk_parity_weights() fails, fall back to equal weight and note in chat.

Output Standards

• All weights as percentages with one decimal: 12.3%.
• Trade quantities as whole shares (no fractional).
• Dollar amounts formatted with commas: $1,234.56.
• Cost as percentage of portfolio with two decimals: 0.15%.
• Table must show current, target, and delta for all holdings.