A stress test estimates how your portfolio would perform under adverse conditions that have not yet occurred. Unlike backtesting — which examines historical returns — stress testing asks: "What fragilities exist in this portfolio that normal market conditions conceal?"
> Stress Test vs Monte Carlo. A stress test is deterministic: one fixed shock scenario in, one before/after impact out — the same scenario always produces the same result. Monte Carlo is probabilistic: it draws thousands of random return paths to produce an outcome distribution. KlirInvest runs them as two separate engines and this article is about the stress-test engine only.
Why Standard Stress Testing Falls Short
Most stress testing tools apply uniform shocks across a portfolio: "What if equities drop 30%?" This approach misses two critical dynamics:
- Correlations increase during crises. Assets that appear uncorrelated in calm markets often move together under stress. A stress test that ignores this underestimates tail risk.
- Market regimes change the distribution of returns. A portfolio that performs well in a low-volatility regime may behave entirely differently when volatility spikes and correlations shift.
KlirInvest's stress testing uses deterministic holding-level shocks: each scenario defines a fixed shock vector, and the headline equals the sum of holding-level impacts. A separate heuristic uses scenario/regime correlation assumptions to illustrate possible diversification effects. It does not change the headline or establish a measured loss reduction. The engine does not draw random paths.
Three Scenarios That Reveal Hidden Vulnerability
1. Broad Market Crash (−30% Equities)
The baseline resilience test. How much of a broad market decline transmits to your portfolio?
What to examine: If your portfolio drops more than 30% in a −30% market scenario, you carry amplified risk — likely through concentration, leverage, or high-beta exposure. If it drops materially less, your defensive positioning is providing genuine protection.
KlirInvest models this with a deterministic equity shock applied position-by-position, then aggregates the portfolio impact using a scenario- and regime-conditional correlation matrix so that diversification benefit reflects how holdings actually co-move under crash conditions. (The separate Monte Carlo engine, which uses Cholesky-decomposed correlation matrices to draw correlated random paths, is a different tool and is documented in its own article.)
2. Interest Rate Shock (+200 Basis Points)
Rate sensitivity varies dramatically across asset classes. Bonds lose value directly. REITs and utilities reprice. Growth stocks — whose valuations depend on discounting distant cash flows — often suffer disproportionately.
What to examine: Identify the proportion of your portfolio in rate-sensitive holdings. If more than 40% of your portfolio is exposed to duration risk, an unexpected rate move creates outsized impact. KlirInvest surfaces this through the rate-shock stress-test scenario and factor-exposure breakdown — separate from the Risk Score itself, which is a volatility-percentile band.
3. Sector Rotation
What happens when capital flows out of your dominant sector into value, defensives, or international markets? This scenario matters most for portfolios with heavy concentration in any single sector — a pattern frequently observed in self-directed accounts, particularly within technology holdings.
What to examine: If your portfolio exceeds 35% in one sector, model a scenario where that sector underperforms by 20% while others remain flat. The difference between your portfolio return and a balanced benchmark reveals your rotation exposure.
How KlirInvest's Stress-Test Approach Differs
Traditional stress tools apply a uniform percentage shock and then sum the impact as if positions were independent. That understates losses precisely when it matters most, because correlations rise in crises.
KlirInvest's stress-test headline is the sum of deterministic holding-level impacts. Its separate correlation-based adjustment is an illustrative heuristic, not a measurement of how holdings will co-move or a change to the headline loss. Fat-tail return distributions and randomly sampled paths belong to a separate engine — the probabilistic Monte Carlo — and are documented in its own article; the stress test itself is a single, reproducible scenario, not a probability simulation.
What Stress Tests Are Used For (Educational)
Stress tests do not predict the future. They are typically used in educational and institutional risk frameworks to:
- Examine the structural impact of adverse scenarios before they occur
- Identify whether hedging conversations are warranted given concentration and exposure
- Compare a portfolio's modelled drawdown to a self-defined risk capacity, so any mismatch is visible during calm periods rather than during volatility
The most useful time to stress test is when conditions feel calm. That is precisely when the results are least contaminated by recency bias.
See how your portfolio scores — run your free analysis at [KlirInvest](/free-tools?tab=analysis).