Where the numbers come from, how a Compass rating gets assigned, how a white paper scenario gets built, and what a claim has to clear before it becomes a published letter.
The Stock Market Compass and Forex & Commodity Compass rate markets and instruments on relative, not absolute, terms. A "cheap" rating is not a prediction that a market will rise — it is a statement that, relative to its own history and to comparable peer markets, current pricing sits below where that history would suggest fair value belongs.
The core inputs behind a Compass rating are:
Price-to-earnings and, where relevant, price-to-book, measured against that same market's own 10-year average and against a basket of comparable peer economies (matched roughly by development stage and market structure, not just region).
Total market capitalisation as a share of GDP, used as a structural sanity check — flagged as a signal, not treated as gospel, since some economies (Hong Kong, for instance) run structurally high ratios for reasons unrelated to overvaluation, and that distinction is stated explicitly whenever it applies.
Currency stability, central bank policy stance, and sovereign debt trajectory — a statistically cheap market with a currency crisis underway is not rated the same as an equally cheap market with a stable one.
Where a rating conflicts with consensus sentiment (a market everyone is bullish on rated as expensive, for instance), that conflict is stated directly rather than softened — the Compass exists specifically to say the unfashionable thing when the numbers say so.
White Paper 5 (The Theseus Trajectory) is the clearest example of the method. Rather than a single-point forecast, it builds a growth-rate range with an explicit base case (15%, with a 12% conservative floor and a 17-20% optimistic ceiling), then names the Six Filters — distribution, governance, repricing, cognition, the human factor, and ethical/environmental cost — as the specific mechanisms that could cause reality to diverge from the model, rather than leaving "things could go wrong" as an unstated caveat.
The scenario fan (four named futures ranked by probability, not by preference) exists so that a reader can see which future the model considers most likely without the paper collapsing into a single confident prediction. White Paper 5, Part B takes this one step further: it is an explicit scenario exercise simulating a 2028 checkpoint using only trends and announced plans visible in 2026 — labelled throughout as a simulation, not a forecast dressed up as a report from the future. That relabeling happened after an earlier draft blurred the line, and the site's Letter 127 documents that correction in full rather than quietly fixing it.
Investment Letters draw on public company disclosures, government and multilateral data (the same IMF/World Bank/BIS sources listed above), and NGE's own prior letters where a thesis builds on earlier work. Every specific figure cited traces to a named source in the letter's closing citation block — we do not publish a number without being able to say where it came from.
The Trajectory series has a different discipline: no fixed publishing schedule, and entries appear only when there is something genuinely worth recording against one of the sixteen named forces in the $1Q Thesis. Silence between entries is deliberate, not neglect.
A number, date, or directional call has to be specific enough that it could later be proven wrong. This is the standard behind the Track Record page — claims vague enough to never be falsifiable don't meet the bar.
Where a statistic comes from advocacy-funded research (an industry association, an NGO-commissioned study), that funding relationship is named in the same sentence as the figure, not hidden in a footnote.
If a specific number can't be traced to a primary source we can check, it's left out and that omission is stated directly rather than silently dropped — this happened explicitly in Letter 132, where several commonly-repeated cost figures were excluded because they didn't hold up under our own verification.
This is a small research operation, not an institutional data desk with a Bloomberg terminal and a compliance department. Worth stating plainly what that means in practice:
No real-time data feed. Figures cited in letters and white papers reflect the most recent published data available at the time of writing, not live market prices. Compass issues carry an explicit date range for exactly this reason.
No proprietary survey data. Everything cited traces to public sources — IMF, World Bank, BIS, FRED, ECB, company disclosures, or peer-reviewed research. NGE does not run its own primary surveys or have access to non-public data.
Questions about how a specific figure was derived are welcome — write to us directly and we'll point to the underlying source.