Source the dataset
Published analysis should identify the data provider, exact series, dates, access date, and any transformations. If a source cannot be verified, we do not present its numbers as historical facts.
Separate observation from model
Actual market data, synthetic proxies, hypothetical scenarios, and illustrative artwork must be labeled separately. Model output is not observed investment performance.
Make the assumptions visible
State starting capital, contributions, timing, currency, rebalancing, fees, taxes, inflation treatment, and any smoothing or lag. Comparisons need matching periods and clearly described differences.
Explain dividends and reinvestment
Identify whether a series is price-only or total return. Specify dividend treatment and reinvestment timing. Do not silently compare a price index with a reinvested total-return portfolio.
Show the limitations
Discuss drawdowns, hindsight, survivorship and selection bias, missing data, and proxy limitations. Historical results are not predictions and do not establish suitability for an individual.
Experiment 001: the private equity proxy
“$10,000 in 1986: S&P 500 vs. Private Equity” compares actual S&P data with a synthetic, smoothed and lagged S&P-return-based proxy. The proxy is illustrative, not actual private equity index data. It does not capture real private equity cash flows, leverage, fees, illiquidity, valuation practices, or manager dispersion. No proxy endpoint is presented here as historical private equity performance.
What is available on this site today
The homepage and experiment-card charts are fictional illustrations. They are not sourced market returns. Experiment 001 directs to our Instagram profile. Its source series, dividend treatment, exact smoothing parameters, and reproducible backtest have not been supplied for this site, so no factual return figures or completed analysis are published here. Other experiment concepts are clearly marked coming soon.