The question behind this project

How much can an investment lose along the way?

Two investments can finish at the same value after very different ups and downs. We compare historical returns and losses, then explore possible scenarios—not promises of future gains.

Read the analysis

Painting of an investment workspace with price histories, asset allocation, and a chart of losses from previous peaks.

Financial analytics An analytical case study

Comparing investment portfolio returns and risks

How does the allocation change historical risk and model-based scenarios?

Historical prices and simulated scenarios

The finding

Return, drawdown, and simulated outcomes describe different aspects of a portfolio. The observation window and model assumptions connect them.

Analytical work

Connect allocation and adjusted price histories with performance attribution, risk diagnostics, and scenario analysis.

FIG. 01Illustrative example

Two investments with the same final value and different losses along the way

Conceptual paths, not historical returns or a forecast. Equal final returns can conceal very different peak-to-trough losses. Compare drawdown, volatility, and concentration before interpreting simulated future outcomes.

Start with the path, not the endpoint

Cumulative return shows growth over the selected window. Drawdown measures distance below a previous peak, exposing losses that a final return can hide. Compare the portfolio and benchmark over matching dates before interpreting the difference.

Diversification depends on joint movement

Weights alone do not describe risk. Correlations determine how asset movements combine, while historical and parametric tail estimates answer different questions. Compare estimates using the same horizon and confidence level.

Treat the frontier as a model result

Mean–variance optimization uses estimated returns and covariance to select weights. Monte Carlo paths extend assumed portfolio behavior into possible futures. Neither an optimized allocation nor a simulation median demonstrates what a future portfolio will earn.

Decision supported

What to do with this evidence

Compare sensitivity across historical windows and assumptions before interpreting an allocation. Use the outputs as analytical scenarios, with out-of-sample evaluation needed for an allocation claim.

What this analysis cannot establish

The implementation assumes frictionless trading and a fixed 4.5% risk-free rate. This is a model parameter, not a statement about current market rates.
Simulations operate at portfolio level with model-based return assumptions. Cached or refreshed prices can change results; scenarios are not observed future outcomes.

Continue the investigation