Where the claims come from
The load-bearing ideas on this site are old and well-documented: diversification, low costs, compounding, the failure rate of market timing, and the gap between investment returns and investor returns. None of them are this site’s inventions. They trace to decades of academic work and to publicly available market history, and the articles try to present them at the strength the evidence supports — no stronger.
When an article states a number — a historical return, a drawdown depth, an expense-ratio threshold — it is one of three kinds:
- A computed figure, produced by scripts in this site’s own repository against public data, and reproducible.
- A public, checkable fact, like a fund’s listed expense ratio or a well-known historical episode.
- A deliberately rough illustration, flagged as such in the text (“somewhere between fifty and a few hundred dollars”).
What you will not find is a precise-sounding number invented to make a paragraph feel authoritative.
What gets simplified, on purpose
This is a beginner’s site, and it simplifies. Tax detail is the clearest example: rules differ by country and change yearly, so articles describe the shape of tax-advantaged investing and leave your country’s specifics to your country’s sources. Similarly, the site rounds where rounding doesn’t change the decision. The test for any simplification is: would the reader have made a different choice with the full detail? If yes, the detail stays in.
Where the chart data comes from
The portfolio charts (growth, drawdowns, rolling returns) run on annual US asset returns from the dataset publicly maintained by Professor Aswath Damodaran at NYU Stern: S&P 500 total return, 10-year Treasuries, 3-month T-bills, Baa corporate bonds, real estate, and gold, from 1928 onward, with the same dataset’s US CPI series as the inflation deflator. Real returns are computed year by year as (1 + nominal) / (1 + inflation) − 1, and portfolios are combined with annual rebalancing.
The limits are disclosed as plainly as the numbers: annual data understate true intra-year drawdowns; the long series are US-dollar, US-market history and other markets differ; nothing is net of taxes or fees. Where a classic portfolio recipe calls for an asset the data layer lacks, the nearest available series stands in (10-year Treasuries for long-term, 3-month T-bills for short-term, small caps for small-cap value), recipes requiring commodities are not listed, and every proxied portfolio is labeled as such on its card. Chinese asset series are not yet included — until a verifiable public source is wired in, the site leaves that gap open rather than estimating it. The build and validation scripts, source links, and year-by-year spot checks live in the repository (docs/data/ASSET-RETURNS.md) and are reproducible by anyone.
Withdrawal-rate rules
The withdrawal-rates and retirement-spending charts share one auditable contract: annual real-return series; withdrawal taken at the start of each year; the withdrawal fixed at the starting capital times the rate (constant in real terms, which is equivalent to a nominal withdrawal raised with inflation each year); a selectable horizon of 20/30/40 years (default 30); and a balance at or below zero after a withdrawal counts as depletion in that year. The “historical safe withdrawal rate” is defined as the highest rate at which every feasible starting year survived the full selected horizon. Annual granularity and the 1928-onward dollar series make this site’s results (about 3.5% for the classic 60/40) more conservative than Bengen’s monthly-data “4% rule”. That gap is a difference of definitions, not an error on either side.
Metric definitions
The dashboard’s strategy cards and screener table compute six metrics from the same monthly NAV series, live at build time:
- CAGR: geometric annualized return: (ending NAV / starting NAV)^(12 / months) − 1.
- Volatility: standard deviation of monthly returns × √12.
- Max drawdown: deepest decline from a previous NAV peak.
- Sharpe: CAGR divided by volatility.
- Sortino: CAGR divided by downside deviation. Note the definition used here is stricter than the classic one: downside deviation is the standard deviation of the negative months only (× √12), not the root-mean-square of below-target returns over all months. It produces somewhat lower ratios; compare within this site, not against numbers computed elsewhere.
- UPI (Martin ratio): CAGR divided by the Ulcer Index, where the Ulcer Index is the square root of the mean squared underwater depth, measured monthly.
All three ratio metrics use a 0% risk-free rate for cross-strategy comparability, so their absolute values run higher than cash-benchmarked versions. Strategies with fewer than 36 months of history get a “short sample” tag, and ratio comparisons across different start dates deserve extra skepticism. Backtests and live strategies share these tables; backtests skew optimistic, for the reasons this page lists elsewhere.
Corrections
When something is wrong, it gets fixed. Typos and awkward phrasing are corrected silently. Anything that could have influenced a decision — a wrong number, a misstated rule — gets corrected with a dated note in the article, and this page keeps the running list. There is deliberately no comment box: accuracy here is a publishing discipline, not a conversation, and this page outranks everything else on the site the moment they disagree.
The relationship to the Chinese edition
The Chinese site is the origin and remains the larger of the two, with a full four-level curriculum, calculators, and a strategy dashboard. The English edition is curated from it, not translated: articles are rewritten from scratch in English, examples are re-chosen for a global audience, and China-specific material is omitted. Over time the two editions will converge in coverage. They already share one spine: principles, actions, the edges of risk, and nothing for sale.