portfolio-manager

---
name: portfolio-manager
description: Use when the user requests a portfolio review, position analysis, risk assessment, performance evaluation, or rebalancing suggestions for their brokerage account holdings.
category: Portfolio
version: 1.0.0
tools: []
---

# Portfolio Manager

Comprehensive portfolio analysis built on live holdings data: asset allocation, diversification, risk metrics, individual position evaluation, and rebalancing recommendations, generated from actual current positions rather than manually entered data.

## When to Use

Invoke on requests like: "Analyze my portfolio," "Review my current positions," "What's my asset allocation?," "Check my portfolio risk," "Should I rebalance?," "Evaluate my holdings," "Portfolio performance review," "What stocks should I buy or sell?," or any portfolio-level analysis or management request.

## Data Prerequisites

This skill depends on a connected brokerage integration (e.g., Alpaca) to fetch:
- Account equity, cash balance, buying power, and account status.
- Current positions: ticker, quantity, cost basis, market price, market value, unrealized P&L, and weight in the portfolio.
- Historical portfolio performance for time-weighted return and drawdown analysis.

If the integration is not connected, tell the user and ask them to connect it or supply a manual holdings export.

## Workflow

### Step 1: Gather and validate portfolio data
Pull account info and positions. Validate that every position has a real ticker, that market values sum to approximately the stated equity, and handle edge cases (fractional shares, options, crypto).

**Leverage discipline**: compute exposure against equity, not just gross market value — `gross_long_exposure = long_market_value / equity` and `cash_pct = cash / equity`. Explicitly flag margin-funded portfolios when gross exposure is materially above 100% or cash is negative; don't let sector weights look benign by dividing only by gross-long value when the account is levered.

**Cross-workflow handoff**: when the request emphasizes dividend holdings or a forced-review trigger, build normalized input for a dividend-monitoring workflow instead of treating dividend review as a narrative-only section. When account-location notes are requested, build tax-planning input too, and label it degraded if account type or holding-period windows are unavailable.

### Step 2: Enrich position data
For each holding, gather current price, volume/liquidity, 52-week range, and market cap; use web search for sector/industry classification, valuation metrics (P/E, P/B, dividend yield), earnings and financial-health signals, analyst ratings, and recent news. Where available, note the technical picture: trend versus 20/50/200-day moving averages, relative strength, support/resistance, and momentum (RSI, MACD).

### Step 3: Portfolio-level analysis

**Asset allocation** — break the portfolio down by asset class (equities/fixed income/cash/alternatives vs. target for the user's risk profile), by sector (vs. a benchmark like the S&P 500), by market-cap tier (large/mid/small), and by geography (US/international/emerging). Report current vs. target with variance, and list the top 10 holdings by weight:

```
### Current Allocation vs Target
| Asset Class | Current | Target | Variance |
### Top 10 Holdings
| Rank | Symbol | % of Portfolio | Sector |
```

**Diversification** — flag any single position exceeding 10–15% of the portfolio, and any sector exceeding 30–40%. Compute a Herfindahl-Hirschman Index (HHI) for concentration. Estimate correlation between major positions to catch redundant "diversification" (highly correlated pairs add concentration risk without adding names). Optimal position count is roughly 15–30 stocks for an individual portfolio; flag under-diversified (<10) or over-diversified (>50) counts. Report as: Concentration Risk (Low/Medium/High, with top-5 weight and largest single position), Sector Diversification (Excellent/Good/Fair/Poor, with the dominant sector's weight), Position Count assessment, and any correlation concerns.

**Risk** — estimate portfolio beta as the position-weighted average of individual betas; compute maximum drawdown and current drawdown from peak; flag the percentage of the portfolio in high-volatility names (beta > 1.5) and in speculative/unprofitable companies; note leverage usage and single-stock or sector-specific tail-risk exposure. Report as: Overall Risk Profile (Conservative/Moderate/Aggressive), Portfolio Beta vs. market at 1.00, Maximum Drawdown (peak value → trough value, and current drawdown from peak), a table of high-risk positions (symbol, % of portfolio, beta, risk factor), and a Risk Score out of 100.

**Performance** — total unrealized P&L in dollars and percent, best/worst 5 positions by gain/loss, time-weighted returns (YTD, 1/3/5-year annualized) against a relevant benchmark when history is available, and a winners-vs-losers ratio with average gain/loss. Report total portfolio value, total unrealized P&L, cash balance and its % of the portfolio, tables of best and worst performers, and portfolio return vs. benchmark return (alpha) when a benchmark is available.

### Step 4: Individual position analysis
For the top 10–15 positions by weight, evaluate each on: thesis validation (why the position exists, whether the thesis is intact, broken, weakening, or strengthening, and recent company developments), valuation (current metrics vs. historical range and sector peers — undervalued/fair/overvalued), technical health (trend, position vs. moving averages, support/resistance), and position sizing (optimal, overweight, or underweight given conviction and risk). Conclude each with one of: HOLD (well-sized, thesis intact), ADD (underweight given the opportunity, thesis strengthening), TRIM (overweight or valuation stretched), or SELL (thesis broken, better opportunities elsewhere) — each with a one-to-two-sentence rationale. Include position details (shares, avg cost, current price, market value, unrealized P/L), a fundamental snapshot (sector, market cap, P/E, dividend yield, recent developments), and technical status (trend, price vs. 50-day MA, support/resistance).

### Step 5: Rebalancing recommendations
Identify rebalancing triggers first: positions or allocations that have drifted significantly from target weights, overweight positions to trim, underweight areas to add, and tax considerations (capital-gains implications of any trim).

Positions to trim: overweight beyond threshold, run up on stretched valuation, concentrated beyond 15–20% of the portfolio, or thesis broken. Positions to add: underweight sectors/asset classes or high-conviction names currently underweight, and new opportunities that improve diversification. If excess cash exceeds 10% of the portfolio, recommend deployment prioritized by opportunity and allocation gaps.

Prioritize actions: Immediate (risk reduction — trim concentrated positions), High Priority (drift over 10% from target), Medium Priority (drift 5–10%), Low Priority (fine-tuning and opportunistic adjustments). Present the plan as a summary of whether rebalancing is needed, followed by the priority-ranked action list with rationale for each move.

## Output

Produce, in order: an asset-allocation section (with the current-vs-target and top-10-holdings tables), a diversification assessment, a risk assessment with beta/drawdown/risk score, a performance review, per-position analysis for top holdings, and a prioritized rebalancing plan. Label any section that fell back to a degraded data source (e.g., REST fallback instead of the live integration, or missing account-location data) so the user knows which numbers to double-check.