QuantNova.jl

Appearance

Scenario Analysis ​

Stress testing and forward projections for portfolios.

Historical Stress Tests ​

Apply historical crisis scenarios:

julia
using QuantNova

# Portfolio state
state = SimulationState(
    timestamp=now(),
    cash=50_000.0,
    positions=Dict(:SPY => 100.0, :AGG => 200.0),
    prices=Dict(:SPY => 450.0, :AGG => 100.0)
)

# Asset class mapping
asset_classes = Dict(:SPY => :equity, :AGG => :bond)

# Apply 2008 crisis
impact = scenario_impact(CRISIS_SCENARIOS[:financial_crisis_2008], state, asset_classes)
println("Portfolio loss: $(impact.pct_change * 100)%")

Built-in Scenarios ​

  • :financial_crisis_2008 - 2008 global financial crisis

  • :covid_crash_2020 - March 2020 COVID crash

  • :dot_com_bust_2000 - 2000-2002 tech bubble burst

  • :black_monday_1987 - Single-day 22.6% crash

  • :rate_shock_2022 - 2022 Fed tightening

  • :stagflation_1970s - 1970s stagflation

Custom Scenarios ​

Create your own stress scenarios:

julia
scenario = StressScenario(
    "Custom Rate Shock",
    "Hypothetical 300bp rate hike",
    Dict(:equity => -0.10, :bond => -0.20, :reit => -0.25),
    90  # duration in days
)

Sensitivity Analysis ​

Analyze sensitivity to specific shocks:

julia
results = sensitivity_analysis(
    state,
    asset_classes,
    :equity;
    shock_range=-0.50:0.10:0.50
)

# Plot or analyze results
for r in results
    println("Shock: $(r.shock*100)% => Value: $(r.portfolio_value)")
end

Monte Carlo Projections ​

Forward-looking simulations:

julia
projection = monte_carlo_projection(
    state,
    Dict(:SPY => 0.08, :AGG => 0.03),  # Expected annual returns
    Dict(:SPY => 0.18, :AGG => 0.05);  # Annual volatilities
    correlation=0.2,
    horizon_days=252,
    n_simulations=10000
)

println("Expected value: $(projection.mean_value)")
println("95% VaR: $(projection.var_95)")
println("Probability of loss: $(projection.prob_loss * 100)%")