Portfolio Simulator

Project thousands of possible futures for your portfolio with a Monte Carlo simulation built on historical returns, volatility, and correlations.

For educational and informational use only — not investment advice or a performance guarantee. Past returns do not predict future results. See Terms §17.

Portfolio & Assumptions
Total weight: 0%
Median Terminal Value

—

P10 — Worst Reasonable

—

10% of outcomes ended below this
P90 — Best Reasonable

—

10% of outcomes ended above this
Probability of Loss

—

odds of ending below total invested
Median Annualized Return

—

on total contributions
Projected Value Over Time

Shaded bands are probability cones: the inner band spans the 25th–75th percentile of simulated outcomes, the outer band the 10th–90th. The solid line is the median (50th percentile) path.

Distribution of Terminal Values
Risk & Inputs
Estimated asset stats (annualized, from history):
Asset Wt Return Vol

What is a Monte Carlo Portfolio Simulation?

A Monte Carlo simulation projects how a portfolio might grow by running thousands of randomized "what-if" scenarios instead of assuming a single fixed rate of return. Markets are uncertain — a 7% average return can hide years of +25% and −30% along the way. By drawing random monthly returns from the historical behavior of your chosen assets, the simulator produces a range of plausible outcomes and tells you how likely each one is.

How this simulator works

  1. Historical returns: For each ticker we pull daily closing prices from our market-data warehouse (up to the last several years) and compute daily returns.
  2. Mean & covariance: We estimate each asset's average return and volatility, plus the correlation between assets — diversification only helps when assets don't move in lockstep.
  3. Correlated random draws: Using a Cholesky decomposition of the covariance matrix, every simulated month draws a set of returns that preserves those historical correlations.
  4. Path simulation: Each of the (up to) 10,000 simulations steps month-by-month across your time horizon, applying returns, adding your monthly contribution, and rebalancing back to target weights at your chosen frequency.
  5. Statistics: We summarize the thousands of ending values into percentile cones, the probability of finishing below what you put in, a distribution of final balances, and a range of annualized returns.

Reading the results

  • Probability cones show the spread of outcomes over time. A wide cone means high uncertainty; contributions and diversification narrow it.
  • P10 / P90 are the "worst reasonable" and "best reasonable" cases — there's roughly an 80% chance the true outcome lands between them.
  • Probability of loss is the share of simulations that ended below your total contributions. Longer horizons and regular contributions usually reduce it.
  • Rebalancing sells winners and buys laggards to hold your target mix, which can reduce risk; "never" lets winners run and drift the allocation.

Important limitations

This model assumes future returns resemble the historical sample and are drawn from a normal distribution — real markets have fatter tails, regime changes, and crashes that simple Monte Carlo understates. It ignores taxes, trading costs, and fees. Estimates from short price histories are noisy. Treat the output as a planning aid, not a forecast or a guarantee.