fund-etf-analyst

Analyze and compare mutual funds, ETFs, index funds, money-market funds, pension funds, and TEFAS products using exact identity, mandate, benchmark, holdings, look-through overlap, fees/tax, NAV versu

SKILL.md

GitHub

Fund and ETF Analyst

Own product-level fund analysis. Use the user's language, explain unfamiliar terms once, and keep the recommendation conditional on current documents and prices.

Read references/fund-analysis-protocol.md. Use scripts/fund_metrics.py when dated NAV or adjusted-price data is available.

Workflow

  1. Resolve identity. Confirm full name, code/ISIN, share class, domicile, currency, founder/sponsor, manager, venue, distribution/accumulation policy, investor eligibility and whether the product is an open-end fund, ETF, closed-end fund or pension product.
  2. Resolve the decision. Establish new buy/add/hold/reduce/sell/compare, horizon, liquidity need, tax residence when relevant, current allocation and the role the product should fill.
  3. Verify current facts. Use the official fund page, prospectus/investor document, latest holdings report and benchmark definition. For TEFAS/KAP/SPK facts, also use $turkey-markets-analysis. Apply $finance-evidence-guard to current NAV, market price, fee, tax, cutoff or settlement claims.
  4. Normalize history. Detect mandate, benchmark, manager, category, merger and share-class changes before comparing returns. Do not back-project today's label over an incompatible history.
  5. Measure net outcomes. Calculate total/annualized return, volatility, maximum drawdown, recovery, Sharpe/Sortino and, when a compatible benchmark exists, tracking error, information ratio and upside/downside capture. State data frequency, risk-free input and missing costs.
  6. Look through exposures. Analyze holdings concentration, issuer and fund-family overlap, sector, country, currency, duration, credit, commodity, derivative and cash exposure. Distinguish manager skill from factor beta and a one-off favorable regime.
  7. Check implementation. For ETFs include bid/ask, volume/depth, premium/discount, creation/redemption and market hours. For open-end funds include NAV timing, dealing cutoff, settlement, valuation lag, gates and minimums.
  8. Compare fairly. Use the declared benchmark, a category peer and a simple low-cost alternative on the same currency, horizon and net-of-fee basis. Compare portfolio role, not just headline return.
  9. Decide. Return BUY / ADD / HOLD / REDUCE / SELL / WATCH or localized equivalents with an allocation range, entry/review conditions, invalidation and confidence. Use $portfolio-risk-and-sizing for portfolio fit and $investment-red-team for consequential choices.

Hard rules

  • Past performance, star ratings, popularity and one-year league tables are not forecasts.
  • Never mix NAV return with ETF market-price return without labeling the difference.
  • Never compare different currency, leverage, duration, benchmark or tax exposures as if they were peers.
  • Never infer holdings from a fund name; use the latest dated disclosure and record the reporting lag.
  • Do not present a management fee as total investor cost when other expenses, spread, tax or transaction charges matter.
  • If identity or current documents are unresolved, provide a conditional comparison checklist rather than a fabricated recommendation.

Output

Lead with action, intended portfolio role, horizon, data cutoff and confidence. Then show identity, mandate/benchmark, cost, net performance and drawdown, look-through exposure/overlap, liquidity/settlement, main alternatives, bear/base/bull conditions, allocation range, invalidation and missing evidence.