Original studies from the simulator data
These studies answer three questions the interactive tools often raise: how much start dates matter, how a purchase schedule changes the result, and why a strong ending value can hide a difficult holding experience. Each article publishes its complete sample, assumptions, and calculation code. The results describe the selected historical data, not an investable recommendation.
1. How holding periods change S&P 500 outcomes
Every eligible 1-, 3-, 5-, and 10-year window in a frozen 2010–2025 snapshot. Compare the minimum, median, maximum, and number of losing windows, with explicit treatment of overlapping observations.
2. Invest now or spread twelve purchases?
$12,000 upfront versus twelve $1,000 purchases, tested from all 181 eligible starting months. Both schedules use the same capital and valuation date. See where each schedule did better and what excluding interest on waiting cash changes.
3. Strong growth can coexist with deep drawdowns
A common 2021–2025 window for S&P 500, NVIDIA, Tesla, and Palantir. Ending growth, peak-to-trough declines, and recovery dates are calculated separately. The study explains why monthly data can miss the worst losses.
How to check a result
Each article links to the same price CSV and snapshot manifest, its own full results CSV, and a Python script that runs offline. You can also replay a selected interval in Single mode. Use a nonzero initial deposit, such as $10,000, and zero monthly contributions to measure an asset's endpoint growth.
Correction to the previous findings
On September 7, 2026, this collection replaced the earlier six-asset summary. That page used unsupported endpoint figures, called S&P price growth “total return,” included retired Korean series, and instructed readers to enter two zero deposits even though the calculator rejects that input. Those figures are withdrawn. The new studies retain exact input data and do not claim uniform total-return treatment across stocks and the index.
Start with the data methodology for the important difference between the main simulator's monthly samples and Quant Lab's separate dataset. Missing dividends, survivorship bias, and sampling frequency can change the interpretation even when the arithmetic is correct.