This section covers the measurement, modelling and use of macro betas — the slope coefficients that quantify how much an asset's return moves per unit of movement in a macroeconomic factor (rates, inflation, the dollar, oil, the growth/credit cycle). Where intermarket analysis asks "which markets move together?", this section asks the sharper, quantified question: "per unit of move in factor X, how much does this stock/sector/portfolio move, and how reliable is that number?" The core tension that runs through every child node is that a macro beta looks like a structural property of an asset but behaves like a regime-dependent, noisily-estimated variable. The framework is mainstream and well established; the individual point estimates are fragile. The whole section is therefore as much about not fooling yourself with a single backward-looking slope as it is about computing one.

What this section is — the core ideas

A macro beta generalizes CAPM beta: instead of regressing a stock's return on the market, you regress it on a macro factor (or a factor-mimicking basket). Elasticity is the same slope put in percentage-per-percentage terms — closely analogous to bond modified duration (percentage price change per unit change in the discount rate), which is why long-duration growth equities fell hardest as long rates rose in 2022 (T. Rowe Price; FDIC RMS manual). The academically correct form (Chen, Roll & Ross 1986; the BIRR model) regresses returns on factor surprises — the unanticipated component — because markets price what is already known; the practitioner shorthand of regressing on raw factor levels/returns is more intuitive but conflates the expected and unexpected parts and overstates predictability. Conflating surprise-betas with level-betas is the single most common conceptual error in the section — the child nodes flag it repeatedly.

When it matters vs. when it doesn't

Macro sensitivities dominate returns in macro-driven regimes — aggressive tightening cycles, inflation shocks, oil supply shocks, sharp dollar moves, risk-off flights to quality — when cross-asset correlations rise and "everything trades on the macro." They matter much less in calm, low-dispersion regimes where idiosyncratic and bottom-up stock-specific drivers dominate. The honest caveat is that the betas are most useful exactly when they are least stable: a relationship estimated over a calm decade can weaken, strengthen, or flip sign in a crisis. So the section's value is conditional — these tools are for bounding and hedging exposure and for regime-aware positioning, not for standalone return forecasting.

Map of the sub-topics (point to the children, don't duplicate)

The branch is organized definition → specific factors → estimation → instability → models → application → Delvantic build:

• What a Macro Beta / Elasticity Is (001) — the foundational definition: the regression-slope form, surprise-vs-level betas, the duration/elasticity analogy. Start here.
• Sensitivity to Specific Macro Factors (002) — the per-factor mechanics and stylized signs: equity duration / interest-rate beta, dollar (DXY) beta, oil & commodities, inflation & breakevens, credit spreads, and the 10-year & real yields. Each documents the transmission channel and how the sign/magnitude shifts by regime.
• Sector & Factor Macro Sensitivities (003) — how the 11 GICS sectors and the style factors (value, growth, momentum, quality, low-vol, size, dividend) tend to respond, via three channels: the discount-rate/duration channel, the earnings-cyclicality channel, and the inflation channel (SSGA; Bloomberg).
• Estimating Sensitivities (004) — the measurement toolkit: rolling-window regression (the workhorse, with its bias–variance window trade-off), multivariate macro regression, beta stability & look-back choice, and orthogonalizing correlated factors (because oil, the dollar and inflation move together, naive univariate betas mislead).
• Why Macro Betas Are Regime-Dependent / Unstable (005) — the central cautionary node. Uses the stock–bond correlation (itself a macro beta) as the canonical case: positive on average through ~1970–1999, persistently negative ~2000–2022, then positive again in 2022 — sign-flips driven by whether growth shocks or inflation shocks dominate (AlphaArchitect; AQR; NBER).
• Macro Factor Models (Barra, BIRR, Statistical/PCA) (006) — the modelling machinery: macroeconomic models (specify factors, regress for exposures), fundamental models (specify exposures, regress for factor returns — the institutional standard, Barra/MSCI, Axioma/SimCorp), and statistical/PCA models. Risk-and-attribution tools, not alpha tools.
• Applying Macro Sensitivities (007) — the decision layer: positioning for a macro view (tilting toward/away from factor exposures), hedging unwanted macro exposure (neutralizing an incidental tilt, trading factor risk for basis + estimation risk), and portfolio macro-exposure decomposition ("what am I actually long?").
• Building This in Delvantic (008) — the engineering blueprint for computing per-stock macro betas from the existing universe price history and surfacing them as a read-only overlay.

Adoption, debate & evidence

The factor-sensitivity framework is uncontested and institutional: it underpins CAPM, APT, the Chen-Roll-Ross macro model, and every commercial multi-factor risk model; fundamental models (Barra/MSCI, Axioma) are the de-facto standard for equity risk (MSCI; SimCorp). Two honest debates recur across the children. First, the pricing of macro factors is genuinely contested — Chen, Roll & Ross (1986) found industrial production, unexpected inflation and the default spread were priced, but Shanken & Weinstein (2006) showed the result was fragile to reasonable specification changes, robust mainly for the industrial-production factor. Second, that betas are time-varying and regime-dependent is well established, but whether you can profitably exploit that time-variation with a conditioning model is much weaker — Lewellen & Nagel (2006) showed the conditional CAPM still fails to explain anomalies. Separate the strong claim (betas are unstable — supported) from the weak one (you can trade the instability — contested). And note Connor's (1995, Financial Analysts Journal) finding that macroeconomic factor models have the lowest explanatory power of the three types (fundamental > statistical > macroeconomic for U.S. equities) — economic interpretability is bought at the cost of fit.

Strengths & limitations (section-level)

Strengths — turns vague macro narrative into a quantified, sizable exposure; separates the macro bet from the stock-specific bet; betas compose cleanly across a portfolio (they add up, weighted), enabling exposure mapping, hedging, risk attribution and scenario stress-testing.

Limitations / the dominant failure modes — (1) instability: the slope drifts and can reverse across regimes; (2) estimation fragility: sensitive to window length, factor choice, frequency and outliers, and corrupted by multicollinear factors; (3) the level-vs-surprise confusion; (4) linearity: betas are local first-order slopes that understate non-linear tail moves. The single most damaging misuse — recurring in node 005 — is treating a beta estimated entirely within one regime as a structural constant (e.g. sizing a "bond hedge" on 2000–2020 data going into the 2022 inflation regime).

Sources

• AnalystPrep (CFA L2) — Macroeconomic, Fundamental & Statistical Factor Models. https://analystprep.com/study-notes/cfa-level-2/describe-and-compare-macroeconomic-factor-models-fundamental-factor-models-and-statistical-factor-models/
• MSCI — Hedging Macro Risk in Equity Portfolios. https://www.msci.com/research-and-insights/blog-post/hedging-macro-risk-in-equity-portfolios
• SimCorp/Axioma — When macro factors speak, investors should listen. https://www.simcorp.com/resources/insights/industry-articles/2024/When-macro-factors-speak-investors-should-listen
• T. Rowe Price — Measuring equity market sensitivity to interest rate changes (duration/elasticity analogy). https://www.troweprice.com/financial-intermediary/au/en/thinking/articles/2024/q3/integrated-equity-newsletter-q2-2024-apac.html
• State Street Global Advisors — How macroeconomic variables impact sector performance. https://www.ssga.com/us/en/individual/insights/how-macroeconomic-variables-impact-sector-performance
• Chen, Roll & Ross (1986), Economic Forces and the Stock Market; Shanken & Weinstein (2006), Economic forces… revisited (fragility critique).
• Lewellen & Nagel (2006), The Conditional CAPM Does Not Explain Asset-Pricing Anomalies, J. Financial Economics 82:289-314. https://faculty.tuck.dartmouth.edu/images/uploads/faculty/jonathan-lewellen/Capm.pdf
• Connor (1995), The Three Types of Factor Models: A Comparison of Their Explanatory Power, Financial Analysts Journal 51(3) — fundamental > statistical > macroeconomic explanatory power. https://rpc.cfainstitute.org/research/financial-analysts-journal/1995/the-three-types-of-factor-models-a-comparison-of-their-explanatory-power
• AlphaArchitect — Implications of Regime-Shifting Stock-Bond Correlation; AQR — A Changing Stock-Bond Correlation.

Flagged disputes (carried from the children): the pricing of macro factors is genuinely contested (CRR vs Shanken-Weinstein); betas' instability is established but exploiting it profitably is not (Lewellen-Nagel); all specific correlation/beta figures cited in the children are sample-, window- and source-dependent, not universal constants.