6 minutes. You have just joined the Structural Forces and Geopolitical Insights team in the CIO Office of a large Canadian pension investor β an illustrative ~$300 billion profile. Your job is not to forecast events. It is to convert structural forces into parameters a strategic asset allocation model can actually use β and to size your own conviction correctly.
How it works: you will make 2 predictions over ten years. For each one, you choose your confidence level. At the end, you see what it was worth β in dollars.
Every number below is published. Your calls in this simulation attach to these books β not to an abstract index.
| Asset class | Net AUM | 1-yr | 5-yr | 10-yr |
|---|---|---|---|---|
| Public Market Equities | $85B | 18.0% | 11.0% | 11.5% |
| Fixed Income | $66B | 2.5% | 3.0% | 3.2% |
| Private Equity | $39B | 6.0% | 12.5% | 12.0% |
| Credit Investments | $33B | 4.0% | 10.5% | 11.0% |
| Infrastructure | $33B | 9.5% | 14.5% | 12.5% |
| Real Estate | $27B | (6.0)% | (0.5)% | 3.0% |
| Natural Resources | $17B | 3.0% | 8.0% | 8.5% |
| Total profile | $300B | 6.0% | β | 9.0% |
This determines which signals you will read early β and which ones you will systematically miss.
These become the themes in your Ten-Year Thematic Outlook. Pick two β you will be asked to make a call on each.
A structural view that does not name an asset class is a newsletter. Choose where your recommendation lands. Real net AUM at March 31, 2026.
Your Outlook was published, debated, partially implemented. The structural forces did what structural forces do β they moved on their own schedule, not yours.
Not "were you right." Whether the mechanism you described is the mechanism that operated β and whether you sized it correctly.
The engagement version runs on your asset mix, your country exposures and your Reference Portfolio β not a published summary table.
Charter: no invented figures. Every empirical value is sourced to a citable public series. Model outputs are labelled projected.
Baseline (illustrative profile). Net AUM and one-, five- and ten-year returns by asset class are rounded, representative proportions for a large Canadian pension investor β not a specific institution. The structural indicator values in this tool remain real and sourced.
Each structural force is assigned an illustrative sensitivity per asset class, expressed in basis points of ten-year annualised return per unit of force intensity. These coefficients are modelling assumptions, not measurements. They are stated openly so they can be argued with. The engagement version replaces them with coefficients estimated on the client's own return series.
Each force resolves to CONFIRMED, PARTIAL or EARLY based on whether the underlying Canadian public series has in fact moved in the direction the mechanism requires. Household debt and fertility resolve as confirmed because the published series are unambiguous; AI diffusion resolves as early because the mechanism is visible while the measured productivity effect is not yet.
Score = Ξ£ (verdict weight Γ conviction alignment) β timing penalty. High conviction on a confirmed mechanism scores full credit; high conviction on an early mechanism is penalised roughly as heavily as being directionally wrong, which is the intended lesson.
Structural indicators β Statistics Canada (fertility, household debt to disposable income, male suicide, educational attainment) Β· OSFI (mortgage renewal volumes) Β· Bank of Canada Β· CMHC Β· FRED. Country axes β Statistics Canada, INSEE, U.S. CDC, IMF (2024). Asset-mix proportions are illustrative, representative of a large Canadian pension. Theoretical frame: Turchin, Todd, Case & Deaton, Putnam, Piketty, Reeves, Alesina, Knack & Keefer.
Understanding the simulation: theoretical frame, forces tracked, calibration philosophy and method.
Structural foresight is not about forecasting events. It identifies the slow forces β demography, debt, cohesion, geopolitical realignment β that move asset values over a decade, and converts them into parameters a CIO office can actually use.
An investment committee rarely lacks short-term information. What escapes it are the long-lead variables: those whose effects are certain but whose timing is not. A pension plan reasons over thirty or forty years of liabilities; its assets must be positioned accordingly.
The difficulty is not being right on direction β most of these forces are visible. It is sizing conviction correctly: putting weight where the evidence justifies it, and holding back where the mechanism has not yet resolved.
| Author / concept | What it contributes, in one line |
|---|---|
| Turchin | A society destabilises when the populace is immiserated AND too many elites compete for too few positions of power ("elite overproduction"). |
| Immiseration | The "impoverishment" side of Turchin: wages that no longer track the cost of living, a standard of living sliding slowly downward. |
| Piketty | When capital compounds faster than the economy grows, wealth concentrates mechanically at the top. Inequality as a structural force, not an accident. |
| Todd | Demography and family structure (fertility, literacy) are the most forecastable variables in social science. |
| Case & Deaton | "Deaths of despair" β suicide, overdose, alcohol among the less-credentialled β signal social decay before political ruptures. |
| Putnam | "Social capital" β ties, associations, trust β erodes; when it collapses, institutions weaken. |
| Acemoglu | What separates a rich country from a poor one is its institutions: rule of law, property rights, rules that hold. |
| Knack & Keefer | Measured that idea: institutional quality predicts growth and investment. The basis of country risk. |
| Reeves | Documents the structural falling-behind of men (education, employment, social role) as a long-term political signal. |
| Force | Mechanism |
|---|---|
| Household debt | Household leverage at G7 highs; a dated renewal cliff converting into political and credit pressure. |
| Demographic contraction | Sub-replacement fertility feeding into the labour force, the tax base and domestic demand. |
| Elite overproduction | More credentialled aspirants than positions, generating counter-elites and institutional volatility. |
| Trade fragmentation | Supply chains reorganising along political lines rather than cost lines. |
| Tariffs Γ productivity | Tariff walls protect incumbents and suppress the productivity growth that justifies multiples. |
| AI diffusion | AI capability accelerates; measurable productivity stays flat β a question of timing as much as direction. |
| Climate costs | Insurer retreat and physical-risk repricing hitting real assets before anything else. |
The simulation does not score "were you right." It scores whether the mechanism described is the one that operated, and whether it was sized in proportion to the evidence. Only one of the four outcomes earns anything.
| Outcome | What it is worth |
|---|---|
| Right, sized correctly | The mechanism held and you put real conviction behind it. The only outcome that creates value. |
| Right, but only flagged | You saw it coming and hedged. The fund captured none of it. Correct and worthless are the same line item. |
| Right on mechanism, wrong on timing | You were right, and three years early. Over ten years, that prices like being wrong. |
| Wrong, high conviction | You moved basis points on a chain that broke. It costs money and credibility. |
Asset-mix proportions and returns are rounded and representative of a large Canadian pension investor β not a specific institution. The structural indicators themselves remain real and sourced.
Each force is assigned an illustrative sensitivity per asset class, in basis points of ten-year annualised return. These coefficients are modelling assumptions, not measurements. They are stated openly so a committee can challenge them β and, under engagement, replace them with coefficients estimated on the client's own series.
Each force resolves to CONFIRMED, PARTIAL or EARLY, according to whether the underlying public series in fact moved in the direction required.
Score = sum of (verdict weight Γ conviction alignment) minus a timing penalty. High conviction on a confirmed mechanism earns full credit; the same conviction on an early mechanism is penalised almost as heavily as a directional error.
| Indicator | Value and source |
|---|---|
| Fertility | 1.25 children per woman (2024), a record low, below the 2.1 replacement threshold β Statistics Canada. |
| Household debt | 179.6% of disposable income in Q1 2026; sixth consecutive quarterly increase β Statistics Canada. |
| Mortgage wall | 3.1M mortgages (52% of the total) renewing by end-2027; 1.3M for the first time since 2021β22 β OSFI. |
BEAVER.WATCH holds to a non-negotiable rule: no invented figures. Every empirical value is sourced to a citable public series; contested claims are flagged; model outputs are explicitly labelled as projected.