Capstone Lab: The Macro Briefing

Section by section

The regime section classifies the economy on the two axes and states the evidence, with the trend and target estimates explicit because they are judgements rather than observations. It should name the quadrant and the direction of travel, since transitions are what reprice portfolios — and it should say what would move the economy into the adjacent quadrant, because that is the risk the allocation is exposed to rather than a general worry. A briefing that names the quadrant without naming what would change it has produced a label rather than an analysis. The path section takes the curve and extracts what it implies, using the forward arithmetic: the rate priced for next year, the rate priced five years out, and the implied terminal rate if the market is right. Then it compares that with the reaction-function prescription under the briefing’s own forecast, which is how the document states where it disagrees with the market. That disagreement, expressed as a number in basis points, is the tradeable content — and if there is no disagreement, the honest conclusion is that the path is priced and the position should come from somewhere else. The stance section converts the policy rate into a real rate and describes what is being delivered through conditions rather than through the rate: credit spreads, the currency, equity prices, lending standards. Then the fiscal and supply section, which says what the issuance of duration is doing to the term premium and over what horizon — and the duration section, which states the book’s rate exposure in years and prices a 100 basis point move through it. Finally the allocation: which assets are paid in this quadrant, which are being carried in case of a change, and the falsifier — a specific, observable condition that ends the call. The briefing, in numbers — Regime: growth 0.5% against a 2% trend; core 2.9% against a 2% target · Implied path: the curve prices one cut over twelve months; the rule prescribes a mild hold ← · Real policy rate: about 1.0%, mild rather than restrictive, against an unobservable neutral rate · Book rate exposure: 4.5 years, so a 100bp rise costs about 4.5% before convexity and coupons · The falsifier: core below 2.5% with the labour market still tight — that would move the regime A briefing is not a forecast. The deliverable is a description of the state, the exposures it pays, and the condition under which the description stops being true. Presenting a point forecast of growth or inflation as the output is the most common failure of the genre, and it is the one least useful to a portfolio.

How this capstone is graded

On four things, none of which is being right about the next data release. First, whether every claim is a number with a source or a derivation: a real rate of one percent, a duration of 4.5 years, an implied path of one cut, rather than adjectives about caution. Second, whether the market’s view is stated before the disagreement: the whole document rests on what is priced, because a view that agrees with the price cannot produce a return. Third, whether the allocation names the assets it holds for the current quadrant and the assets it holds for the adjacent one, which is the regime lesson applied to a book rather than a slogan. Fourth, whether the falsifier is observable, dated and specific enough to act on — the same standard a position invalidation has to meet. The second half is calibration, exactly as in the company capstone. Before filing, record the two quantities the briefing depends on most — core inflation in two quarters and the unemployment rate — with explicit confidence levels. The record of those forecasts against the outcomes is the only way to know whether your confidence means anything, and the pattern of errors is the most valuable output of the subject: a systematic underestimate of services inflation, or a habit of assuming the market is wrong about the path, is a fact about the analyst rather than about the economy. The subject closes by being self-contained. The frameworks came from the beginning: growth and the identity from the first lesson, the reaction function from the second, real returns from the third, the curve from the fourth, inflation from the fifth, expectations and transmission from the middle, the credit channel and the fiscal arithmetic from the cases. What the capstone adds is the assembly — and the discipline of saying, in writing and in advance, what would make the whole thing wrong. • Every claim a number with a source; no adjectives standing in for analysis. • The market’s priced view first, then the disagreement in basis points. • Hold for the current quadrant and for the adjacent one, explicitly. • The falsifier observable, dated and specific. • File your key forecasts with confidence levels, because calibration is the transferable skill. This is the same structure the equity research note used: a chain of arithmetic ending in a sized exposure with a stated way of being wrong. The subjects differ in their inputs, not in their discipline.

The forward arithmetic, worked

The section the briefing depends on most gets the briefest instructions, so the mechanics deserve a page of their own. The curve does not publish a path; it publishes yields at maturities, and a path is extracted from them by comparing two points. Take a one-year yield of 3.75% and a two-year yield of 3.50%. Money invested for two years at the two-year rate should equal money invested for one year and then reinvested, so the second year is priced at roughly 2 × 3.50 − 3.75, which is 3.25%. That figure — the one-year rate one year forward — is what the market expects policy plus a term premium to deliver in the second year, and it is the raw material of the briefing. The conversion to policy is a division by twenty-five basis points, because a cut is a quarter of a point. If the one-year forward is 3.25% against a policy rate of 4.00%, the market is pricing about three quarter-point reductions across that year. Stated that way the number is checkable against the reaction function: if the briefing’s own payroll and inflation forecasts would have the committee cutting twice, then the document is short the market’s path by roughly 25 basis points, and that figure *is* the position. Running the same comparison across a whole curve — one year forward two years out, three years out — produces the implied terminal rate, which is where the market thinks the cycle stops. Two cautions belong with the calculation. The first is that forwards contain a term premium, so a forward above or below the expected policy rate does not simply mean the market disagrees with the committee; part of the gap is the compensation investors demand for holding duration, and in periods of heavy issuance that component can be substantial. The second is that the arithmetic assumes a no-arbitrage relationship that holds only in expectation, so a path derived from the curve is a market forecast rather than a fact about the future. Both caveats are stated rather than resolved, because the briefing’s job is to be explicit about what the price implies, not to pretend it is right. The payoff of doing it in numbers is that the rest of the document becomes arithmetic. Once the path is extracted, the real rate is the policy rate less expected inflation over the same horizon; the rate exposure of the book converts it into a portfolio number; and the falsifier can be written against an observable — the next two inflation prints, the unemployment rate, the terminal rate the curve implies. That chain is what the grading section means by every claim being a number with a derivation, and it is what separates a briefing that can be reviewed from one that can only be agreed with. • Extract a forward by comparing two maturities: 1y at 3.75 and 2y at 3.50 implies about 3.25% for the second year. • Divide the gap from the policy rate by 25bp to get the number of cuts or hikes priced. • Run the same comparison across maturities to get the implied terminal rate. • Forwards contain a term premium, so a gap from policy is not purely a policy disagreement. • The derived path is a market forecast: state it, then disagree with it in basis points. The position is the difference between your prescription and the priced path, expressed in basis points. If the calculation shows no difference, the honest briefing says so and finds its exposure somewhere else.

Sources

• Regime-based asset allocation and expectation investingMR12 on quadrants; Rappaport & Mauboussin, “Expectations Investing” for the reverse-engineering discipline
• The stance of policy and financial conditionsMR2 and MR9 on the reaction function and conditions; FOMC materials, September 2026
• Duration as the common measure of rate exposureMR13 on duration and convexity
• Fiscal arithmetic and the term premiumMR11 on debt dynamics and the supply of duration