About Simfolio

Built by Investors, for Investors.

Simfolio uses academic research to create portfolio management tools, providing investors with the infrastructure to utilize them systematically.

Simfolio is self-funded, independent, and built with one goal: provide the infrastructure to access better portfolio management tools.

For more details, please see the Methodology page.

About Simfolio

Simfolio was built by a small team of investors who value the idea of turning research-grade methodologies into practical workflows to backtest, forecast, and optimize portfolios.

Simfolio is built around two core workflows:

Backtesting: Understand how a portfolio would have performed historically.

125 years of historical data
100+ metrics, statistics and diagnostics
Custom factor regression
Principal component analysis
Walk-forward optimization
Frontier forecasting methodology
AI MCP integration
Upload custom time-series data
Downloadable reports

Forecasting: For any given portfolio, understand the possible future paths based on a forecasting methodology optimized for distributional accuracy.

Up to 100-year portfolio forecasts
Full predictive distributions
Model selected from 200+ methodologies
Rolling-origin out-of-sample validation
Validated across 80 multi-asset portfolios
Dynamic multiscale stochastic volatility
Empirical fat tails and return dependence
Outperformed a niave model at every tested horizon
Forecast-based portfolio optimization
Target-probability and downside-risk analysis

Additional tools added to the current suite aim to enhance the usefulness of the two core workflows.

Core Problem #1

The Historical Replay

Many portfolio simulation tools are scenario randomizers that simply replay history, resulting in the significant oversimplification of how financial markets behave.

Simfolio's forecasting model directly targets those weaknesses and was derived by optimizing for out-of-sample distributional accuracy.

Core Problem #2

Lookahead Bias in Traditional Portfolio Optimization

A popularized approach to portfolio optimization is to optimize asset weights once, based on full in-sample data and assuming those weights would have been known at the start of any backtest.

Simfolio offers a robust solution via walk-forward portfolio optimization, in which the model can only optimize based on data it has at the time of optimization. Weights are then rolled forward until the next optimization, ensuring the backtest reflects true out-of-sample performance.

Core Problem #3

The Failure of Market Timing

Many arbitrary, tactical or signal-based strategies assume the effectiveness of market timing, indirectly asking investors to assume which asset class, signal, or regime will lead next. Across nearly all live tactical funds and large technical-rule universes, that promise usually fails to beat simple static and periodically rebalanced portfolios once data snooping, transaction costs, taxes, and out-of-sample validation are taken seriously.

Simfolio does not make that timing bet part of the core tools offered in-house; tactical or signal-based investors can upload any daily time series data to integrate any strategy as a custom ticker.