QCS Python changelog

0.7.3 (2025-09-08)

Improvements

• Optimizer
  • Add scr to supported constraint_type
  • Add cost function in snapshot.optimize(): min_scr
  • Method check_portfolio_constraints supports subportfolio constraints
  • Allow use of np.nan in lieu of -1e9 in portfolio constraint for subportfolio ones
  • Turnover constraint at position level now defaults to relative calculation

0.7.2 (2025-03-06)

New features

• Methods for logging in and out using auth0 through QCS web portal
  • Add qcs.login()
  • Add qcs.logout()
• Update examples

0.7.1 (2025-01-31)

Improvements

• Method top_factors from snapshot
  • Add replication_method (forward, backward, stepwise, enhanced_stepwise, adaptive_stepwise) and epsilon
  • Add trace (true/false) in order to output the full sequence of steps taken during replication

0.7.0 (2024-12-02)

Improvements

• Improve documentation on https://qcs-python-docs.quarisma.net/
• Improve error message when optimizer fails
• Allow no local_db in Context
• Update examples
• Improve build and internal tools process

0.6.0 (2024-05-31)

New features

• Methods for calculating returns
  • Add snapshot.get_historical_returns(context=context, window="1m")
  • Add snapshot.get_market_implied_returns(context=context, implied_perf_exponent=3)
• Are constraints feasible
  • Add snapshot.are_constraints_feasible() method
• Check portfolio constraints
  • constraints_met, breached_constraints = snapshot.check_portfolio_constraints(portfolio_constraints, context)
• Efficient frontier
  • Add snapshot.efficient_frontier() method
• Top factors and multiregression beta indicator
  • Add snapshot.top_factors() method
  • Add example_multiregression_beta.py

Improvements

• Various refactoring and improvements

0.5.0 (2024-03-25)

New features

• Constraints builder
  • Add PositionConstraint class with 3 parameters:
    • type: Literal["lock", "unlock", "long_only", "long_short", "short_only", "turnover"]
    • filter: function
    • value - used only for type=turnover
  • Add snapshot.apply_position_constraints() method

0.4.0 (2024-03-13)

New features

• Add snapshot.get_covariance() for covariance analysis
  • Support for rectangular covariance as well
• Add robust mean variance optimizer
  • Add more cost functions and inputs in snapshot.optimize(): min_tracking_error, robust_mean_variance, markowitz, risk_parity, min_variance, max_diversification
  • Position class now supports properties: locked, turnover, long_short, expected_return
  • Improved validation errors
  • Added basic support for optimizer_constraints List input.
  • Supported constraint_type : std, std=, std.bm, var, cvar, mdd, cmdd, turnover, min_return
  • Added an example file example_optimize_robust_mean_variance.py
• Add Black-Litterman returns
  • Add snapshot.bl_returns() method
  • Positions in snapshot can now be given as size or allocation. This works everywhere (risk analysis, optimizer etc.).
    • Error is thrown if snapshot contains some positions with size and some with allocation
  • AllocationSnapshot now has new properties:
    • reference_market which must be another AllocationSnapshot. Currency must be identical to AllocationSnapshot.
    • pricing_context which can contain risk_aversion or risk_free_rate.
  • Updated examples/example_black_litterman_returns.py
• Add asset, history and snapshot import helpers
  • qcs.import_assets_from_csv()
  • qcs.import_histories_from_csv()
  • qcs.import_snapshot_from_csv()
  • Use these helpers in examples

Improvements

• Throw error when importing unsorted histories
• Various refactoring

0.3.0 (2024-01-29)

Improvements

• Use compressed history format in requests
• Use RTNEW with custom pricer formulas when sending request