Macro ReviewHow to read
Outlook · Playbook & ranges · Playbook & ranges

Stocks, bonds and oil in the bonds-with-stocks, oil-against pattern for 178 sessions

What has this been followed by, and what do the site's models say each asset is worth?

Every number carries its own dateUpdated 17 Nov 2026
Market patternBonds with stocks, oil against178 sessions since 5 Mar 2026 · the daily sign ruleBonds with stocks, oil against
S&P 500, 13 weeks after a start+6.0%median of 7 starts, today's included · any week +4.3%
Gold, 13 weeks after a start+14.0%median of 7 starts, today's included · any week +2.6%
Assets with a target2of 6 on the board · the rest have no fair value there
The , S&P 5004,338−42% from the index · · no target
10-year Treasury yield, fair value4.12%par 5.12% · 100bp below it · 16 Nov 2026

In brief

Descriptions of history and levels, each with its count, model and what would change it
  • Stocks, bonds and oil have moved in the bonds-with-stocks, oil-against pattern for 178 sessions, since 5 Mar 2026 on the daily .→ Exhibit 1
  • After the 7 starts of the bonds-with-stocks, oil-against pattern since 2016, today's included, the S&P 500's median 13-week return was +6.0%.→ Exhibit 2
  • The correlation pattern is used because its states went with different later S&P 500 returns (p 0.026); the 's did not.→ Exhibit 12
  • 4 of 6 assets carry no target: only the 10-year and credit have a fair value on the board.→ Exhibit 6
  • The S&P 500 carries no target: the at the 4,338 (−42%) · · no target.→ Exhibit 6
What would change this view
What would change this viewTriggerTodayDecided by
A pattern formsboth legs beyond 2 standard errors+0.5 / −2.0 (20 Nov 2026)not crossedeach Friday's close
The Committee's path moves off the the 2026 or 2027 median funds rate ≠ 4.1%4.1% · 4.1% (16 Sep 2026)pendingFOMC projections, 9 Dec 2026
The premium turns positive > 0bp−39bp (Sep 2026)not crossedthe Board's file for Oct 2026 (no published date)
The reasoning behind each point
  • State
    Stocks, bonds and oil have moved in the bonds-with-stocks, oil-against pattern for 178 sessions, since 5 Mar 2026 on the daily sign rule.

    The daily sign rule reads the state from the signs of the 63-day stock–bond and stock–oil correlations and names it by its legs. This page reads that sign pattern, not the four regimes of Correlations & regime, which are the site's regime; the regime model's patterns describe how assets move together, and are not what this page conditions on either.

  • Playbook
    After the 7 starts of the bonds-with-stocks, oil-against pattern since 2016, today's included, the S&P 500's median 13-week return was +6.0%.

    Log returns over the 13 weeks after each start; the current run's own start (6 Mar 2026) is one of the 7, because its 13 weeks are complete. Interquartile range +5.2 to +7.9%, against +4.3% for any week (618 weeks). Cells of fewer than 5 episodes are greyed: both against stocks (3). Seven episodes cannot tell +6% from +4%: a description of history, not a forecast.

  • Classifier
    The correlation pattern is used because its states went with different later S&P 500 returns (p 0.026); the regime model's did not.

    Kruskal–Wallis across states on point-in-time labels, 13-week forward S&P 500 log returns on 39 non-overlapping windows, circular shifts of the labels as the null: the pattern's states give p = 0.0256; the regime model's p = 1.00 on 38. That p is the smallest the test can give: 39 windows allow 39 arrangements, so 1/39 is its floor, and the observed split beat every shifted one. What it supports is that the states are not arbitrary labels of what followed — not a return for any one state: the bonds-with-stocks, oil-against state holds 4 of those windows. The pattern is read as a description of how assets move together, not a signal.

  • No target
    4 of 6 assets carry no target: only the 10-year and credit have a fair value on the board.

    The 10-year: range 3.10 to 4.62 (the SEP path at the grid's term premia, 2015–19 ACM average (−0.29pp) to today +50bp (+1.23pp); the five published paths at today's premium span 3.94–4.17%); it changes on “The Committee's path moves off the SEP path” (FOMC projections, 9 Dec 2026). Credit: range 94 to 200 (the fair spread plus the premium's 10th to 90th percentiles since 1973 (−38 to +69bp)); it changes on “The premium turns positive” (the Board's file for Oct 2026 (no published date)).

  • Equities
    The S&P 500 carries no target: the model's value at the yield-gap premium 4,338 (−42%) · record unproven · no target.

    Record unproven: the gap's 10-year slope is +0.038 per point with a Hodrick t of 1.61, short of the 1.96 the rule asks (4 non-overlapping windows); out of sample it loses to the historical mean at every horizon (R²_OS −0.05 at one year, 39 windows; −0.38 at 10 years, 2 windows); the record cannot tell it apart from the earnings yield less the real yield at any horizon (encompassing p 0.44–0.73; a test with little power on this few windows, so this is not evidence that it adds nothing). The record, and what it can and cannot tell apart on so few ten-year windows, is on the Valuation history page.

Next: Fair value board — each value's model and record

Source: Yahoo Finance; U.S. Treasury; New York Fed; Federal Reserve Board; EIA; FactSet; Robert Shiller; Macro Review calculationscross.playbook-read

Stocks move with bonds and against oil: what followed that before

The two correlations that name today's pattern, and the returns over the 4, 13 and 26 weeks after each earlier start of it, with the count printed once and the any-week median beside it

Exhibit 1 · Stock–bond and stock–oil correlations over time, against zero

63-day correlation ratios of daily returns, 16 Nov 2023 to 16 Nov 2026: stock–bond above zero and stock–oil below it is the bonds-with-stocks, oil-against state
About
About this exhibit
Lines of the 63-day stock–bond and stock–oil correlations of daily returns over the last three years (range toggles to the whole record), against zero
The two correlations the daily sign rule reads
63-day correlations of daily returns, 16 Nov 2023 to 16 Nov 2026: stock–bond above zero and stock–oil below it is the bonds-with-stocks, oil-against state of the daily sign rule.
Source: Yahoo Finance; Macro Review calculations · data to 16 Nov 2026cross.playbook-2

Exhibit 2 · Each asset's return after starts of the current pattern, against any week

Median log return, %, after the 7 starts of the sign rule's “bonds with stocks, oil against” pattern since 18 Nov 2016; interquartile range beneath
About
Asset4 weeks aftern = 7 starts13 weeks aftern = 7 starts26 weeks aftern = 7 startsAny week, 13 weeks
S&P 500+1.6%+0.6% to +2.9%+6.0%+5.2% to +7.9%+10.1%+6.5% to +12.7%+4.3%
Nasdaq-100+1.1%+0.5% to +4.2%+8.3%+6.2% to +13.2%+9.4%+5.2% to +14.3%+5.1%
Semiconductors+2.1%+0.6% to +10.1%+20.7%+14.7% to +24.9%+30.6%+17.4% to +33.0%+7.4%
Russell 2000+1.6%−1.1% to +5.2%+7.3%+2.2% to +10.6%+15.5%+6.7% to +16.8%+3.4%
Treasuries 20y++1.3%−1.9% to +2.0%+3.3%+1.7% to +5.1%+2.5%+0.9% to +6.7%+0.3%
Treasuries 7–10y−0.1%−0.7% to +1.6%+2.3%+1.2% to +3.4%+3.7%+2.1% to +4.1%+0.3%
Gold+2.9%+0.7% to +6.7%+14.0%+5.0% to +15.9%+24.8%+18.2% to +32.6%+2.6%
WTI crude (USO)−0.9%−3.3% to +9.4%−1.0%−6.8% to +0.3%+4.9%−5.6% to +19.7%+1.8%
US dollar bullish (UUP)+0.8%+0.1% to +1.6%+2.4%+0.4% to +2.5%+2.2%+1.3% to +3.3%+1.1%
High yield+0.7%±0.0% to +1.2%+1.5%+1.3% to +1.7%+3.9%+3.0% to +5.2%+1.3%
Bitcoin (USD)−0.7%−3.0% to +8.3%−11.2%−13.6% to −4.4%−13.6%−37.3% to +10.6%+8.5%

7 episodes cannot tell +6% from +4%: read the medians as what happened, not as a base rate.

About this exhibit
Table of each asset's median log return (interquartile range beneath) at each horizon after each start of the current sign-rule pattern, against the any-week median
What followed earlier starts of today's pattern (Bonds with stocks, oil against)
Median log return after each of the 7 starts of the bonds with stocks, oil against pattern since 18 Nov 2016 on the weekly sign rule, today's included; the interquartile range beneath.
A start is a run of the pattern lasting ≥ 3 weeks; the current run's own start on 6 Mar 2026 is one of them, because its weeks are complete. The last column is the median over every week of the sample. Cells resting on fewer than 5 starts are greyed. Descriptions of history, not forecasts.
Source: Yahoo Finance; Macro Review calculations · data to 13 Nov 2026cross.playbook-3

Two assets carry a target; the S&P 500's value is a model level, not a target

The S&P 500's value by basis against the index, every asset's value and range against today, and each target with its price, distance and the test that would change it

Exhibit 3 · S&P 500 valuation-model ranges on each premium basis, against the latest price

% from the 16 Nov 2026 price (zero line); circle: centre; bar: the spread of inputs on each premium basis
About this exhibit
For each premium basis (the implied premium at the site's growth, the yield-gap premium, the full-history median, the yield gap in months when stocks and bonds moved apart, all inputs drawn): the valuation model's range and centre for the S&P 500, in % from the latest price, the price at zero
The S&P 500 against the valuation model's ranges
S&P 500, % from today's price: each premium basis's range and centre on the valuation model, today at zero. The model's value at the yield-gap premium 4,338 (−42%) · record unproven · no target.
Source: FactSet; Robert Shiller; U.S. Treasury; New York Fed; Yahoo Finance; Macro Review calculations · data to 16 Nov 2026cross.playbook-9

Exhibit 6 · Each asset's target or model level against the latest price

The latest price, the target or the model's estimate, the gap (%, bp), the range and the test that changes it, as of 13 Nov 2026
About
AssetTodayTarget, or the model's levelDistanceRangeWhat would change itA year with no view
S&P 5007,41716 Nov 20264,338model's level, not a target−41.5%No target — the model's value at the yield-gap premium 4,338 (−42%) · record unprovenThe gap's record →−21.4% to +27.4%
10-year Treasury5.12%16 Nov 20264.115%target · One further +25bp (the SEP path)−101bp3.10%–4.62%assumed valuesThe Committee's path moves off the SEP pathFOMC projections, 9 Dec 2026−110 to +110bp
Corporate credit (GZ spread)93bpSep 2026132bptarget · the spread issuers' default risk predicts+39bp94bp–200bppast missesThe premium turns positivethe Board's file for Oct 2026 (no published date)–
Gold (GLD)$355.4513 Nov 2026$288.81model's level, not a target−18.7%No target — record unproven: too few windows to test itGold's model in full →−22.6% to +30.3%
Crude oil (WTI)414.4commercial crude stocks, not a price (the model's input) · 6 Nov 2026––No target — no reliable model: fails its stability test−39.2% to +65.1%
The dollar–––No target — no model: no foreign interest rates to model it on−8.4% to +9.2%
About this exhibit
Table of the assets with a target or a model level: the latest price, the target or the model's level, the distance, the range and its kind, the test that would change it, and the volatility model's one-year range
The targets, and the assets without one
Today's price, the target or the model's level, the distance, the range and the test that would change it.
A target here is the board's fair value, printed only with its range and what the range is, its horizon, its model and the dated test that would change it. An asset with no reliable model, or a model whose record is unproven, prints that and no target; its model level is greyed, not a target. The last column is the volatility model's 52-week 10th–90th range with no view (13 Nov 2026), for reading a distance against a year's typical move.
Source: Yahoo Finance; U.S. Treasury; New York Fed; Federal Reserve Board; EIA; FactSet; Robert Shiller; Macro Review calculationscross.playbook-6

What the policy paths do, and what every pattern was followed by

The rate paths the site prices, through the fair 10-year and the valuation formula; and every pattern's 13-week medians side by side

Exhibit 7 · 10-year yield and S&P 500 under each policy path, valuation formula

10-year moved to its fair value on the policy-rules grid; S&P 500 % from the 16 Nov 2026 price, all else held; the model's value, %
Range: named paths, no odds
Path10-year move10-yearS&P 500, % from todayThe model's value at that 10-year, % from today
Hold pathThe ten-year at the hold path's fair value−106bp4.065%+29.6%−29.7%
SEP pathThe ten-year at the SEP path's fair value−101bp4.115%+27.8%−30.4%
Above the dotsThe ten-year at the fair value of a hike path above the dots (+50bp, then hold)−96bp4.165%+26.0%−31.0%
Term premium upThe ten-year up by the term premium's last three monthsnot priced
About this exhibit
Table of the policy paths (hold, SEP, above the dots): the 10-year's move to its fair value and level, the S&P 500's move from today on the valuation identity, and the model's value at that 10-year as % from today; the term-premium path is not priced
What the policy paths do to the 10-year and the S&P 500
Each path moves the 10-year to its fair value on the policy-rules grid; the S&P 500 is the valuation identity at that 10-year with everything else held (model-conditional, copied from Scenarios, 16 Nov 2026). Today: 10-year 5.12%; the model's value −41.5% from today's price. 4 named branches (Hollow deal, Attrition, Escalation, Genuine deal) are not priced through the identity: no ten-year pin: oil is not modelled through the identity — no elasticity of index earnings or of the premium to it exists among the site's estimates, and none is invented here. The spread across the paths is a : named paths, not a probability.
Source: U.S. Treasury; New York Fed; Federal Reserve Board; FactSet; Robert Shiller; Macro Review calculationscross.playbook-5

Exhibit 8 · Each asset's 13-week return after starts of each pattern

Median log return, 13 weeks after sign rule starts since 18 Nov 2016, %; starts counted under each pattern; any week last; under 5 starts: thin
AssetBonds against stocks, oil with11 startsBoth against stocks3 starts · thinBoth with stocks11 startsBonds with stocks, oil against (today)7 startsAny week
S&P 500+5.8%+5.4%+4.4%+6.0%+4.3%
Nasdaq-100+8.2%+7.2%+1.8%+8.3%+5.1%
Semiconductors+9.6%+2.9%+0.3%+20.7%+7.4%
Russell 2000+5.3%+6.8%+1.5%+7.3%+3.4%
Treasuries 20y++0.1%−4.2%−1.5%+3.3%+0.3%
Treasuries 7–10y−0.5%−0.9%+0.7%+2.3%+0.3%
Gold+2.0%−5.6%+5.7%+14.0%+2.6%
WTI crude (USO)+8.3%+9.1%+5.9%−1.0%+1.8%
US dollar bullish (UUP)+1.1%+4.9%−0.3%+2.4%+1.1%
High yield+2.0%+1.9%+0.9%+1.5%+1.3%
Bitcoin (USD)+33.1%−28.8%−2.4%−11.2%+8.5%
About this exhibit
Table of each asset's median 13-week log return after starts of each of the sign rule's states (start counts under each name, the current state marked), with any week last
Every sign-rule state, 13 weeks after its starts
The median 13-week log return after the starts of each state since 18 Nov 2016, the number of starts under each state's name; today's state is marked. A state with fewer than 5 starts is flagged thin: its medians rest on a handful of observations.
Source: Yahoo Finance; Macro Review calculationscross.playbook-4

What equity funds did in each spell of the regime

Each fund's change over every spell of today's regime, and the median change in the year after each regime began against any week's, with the count beside it

Exhibit 9 · Equity funds' change in each spell of the regime "Only the dollar hedges"

% change from each spell's first week to its last; the current spell, since 20 Mar 2026, to the latest close. 6 spells since 1998.
SpellWeeksS&P 500
SPY
Technology
XLK
Semiconductors
SMH
Growth
SPYG
Value
SPYV
Consumer staples
XLP
23 Oct 2020 to 25 Dec 202010+7.2%+9.2%+17.6%+6.3%+8.5%+2.9%
26 Feb 2021 to 26 Nov 202140+21.8%+27.7%+24.3%+29.0%+14.1%+15.8%
15 Apr 2022 to 12 May 202357−4.4%+4.0%+4.4%−8.7%−0.8%+0.4%
14 Jul 2023 to 8 Nov 202470+35.4%+36.0%+66.3%+43.6%+26.6%+12.6%
20 Dec 2024 to 28 Mar 202515−5.7%−12.2%−12.2%−9.8%−1.1%+1.9%
20 Mar 2026 to 16 Nov 2026 (current)35+14.7%+22.2%+20.1%+13.4%+16.4%+16.1%
Source: the regime vote (Correlations & regime); Yahoo Finance adjusted closes; Macro Review calculations · prices to 16 Nov 2026cross.playbook-12

Exhibit 10 · Equity funds' median change in the 52 weeks after each regime began

Regime that beganS&P 500TechnologySemiconductorsGrowthValueConsumer staples
Nothing hedges+14.7%
n = 10
+27.3%
n = 10
+32.8%
n = 9
+21.7%
n = 9
+9.7%
n = 9
+7.0%
n = 10
Bonds hedge+5.4%
n = 11
+6.0%
n = 10
+1.5%
n = 10
+6.0%
n = 10
+3.5%
n = 10
+9.4%
n = 10
Bonds and the dollar hedge+11.2%
n = 2
+24.8%
n = 2
+12.6%
n = 2
+17.2%
n = 2
+7.0%
n = 2
+12.7%
n = 2
Only the dollar hedges (today's)+16.6%
n = 5
+23.3%
n = 5
+46.2%
n = 5
+20.1%
n = 5
+15.9%
n = 5
+7.4%
n = 5
Any week+13.3%
n = 1456
+15.0%
n = 1405
+15.8%
n = 1329
+13.9%
n = 1312
+11.6%
n = 1312
+8.7%
n = 1405
About this exhibit
Median % change over the 52 weeks after each start of a regime since 1998, the number of starts beneath; any week's median on the last row.
Source: the regime vote (Correlations & regime); Yahoo Finance adjusted closes; Macro Review calculations · prices to 16 Nov 2026cross.playbook-13

What has moved against the S&P 500, pattern by pattern

Each asset's hedge ratio to the S&P 500 over all weeks since 2014 and within each correlation pattern, and the 's value today — what happened, not a forecast

Exhibit 11 · Each asset's hedge ratio to the S&P 500, by correlation pattern

Weekly beta ratios, 672 weeks 3 Jan 2014 – 13 Nov 2026: all weeks and each pattern with 52 weeks or more; the latest conditional beta
About
AssetAll weeksBonds with stocks, oil against (now)81 weeksBonds against stocks, oil with428 weeksBoth with stocks126 weeksBoth against stocks37 weeksEngine, conditionalβ = ρ σ_asset ÷ σ_SPY
Treasuries 20y+TLT−0.08ρ −0.09+0.30ρ +0.29−0.18ρ −0.23+0.09ρ +0.09< 52 weeks−0.02ρ −0.05 · σ 0.79/1.59
Treasuries 7–10yIEF−0.03ρ −0.09+0.17ρ +0.33−0.09ρ −0.26+0.07ρ +0.14< 52 weeks−0.03ρ −0.09 · σ 0.45/1.59
GoldGLD+0.11ρ +0.12+0.33ρ +0.22+0.12ρ +0.14+0.09ρ +0.09< 52 weeks+0.32ρ +0.30 · σ 1.70/1.59
US dollar bullish (UUP)UUP−0.11ρ −0.25−0.19ρ −0.45−0.08ρ −0.20−0.14ρ −0.30< 52 weeks−0.03ρ −0.12 · σ 0.40/1.59
About this exhibit
Table of each asset's weekly beta (with correlation) to the S&P 500 over all weeks since 2014 and within each correlation pattern with enough weeks, plus the latest conditional beta from the correlation model
Hedge ratios to the S&P 500
β = Cov(asset, SPY) ÷ Var(SPY) on weekly log returns 3 Jan 2014 – 13 Nov 2026 (672 weeks), over all weeks and within each correlation quadrant with at least 52 weeks (current: Bonds with stocks, oil against); the last column is the regime model's conditional value on 13 Nov 2026 from the DCC correlation and the GARCH standard deviations of that day (daily returns). A negative ratio means the asset moved against the index.
Source: Regressions and the regime engine computed here on Yahoo Finance prices (licensed).cross.factors-11

How far can each object be trusted?

The test that picks the correlation pattern, and each target model's record

Exhibit 12 · Each classification tested against later S&P 500 returns, and target model records

p-value of a test that states separate later 13-week S&P 500 returns (0.026 is the lowest it can give), point-in-time states; then each target's record
About
ClassificationpWindowsUsed here
The sign rule's stateswindows by state: Bonds against stocks, oil with 22 · Both against stocks 2 · Both with stocks 11 · Bonds with stocks, oil against 40.025639the playbook conditions on it
The engine's patterns1.0038a description of co-movement, not conditioned on
  • S&P 500

    Record unproven on this board's rule, on the Valuation history page's record of the gap (real-time data): the gap's 10-year slope is +0.038 per point with a Hodrick t of 1.61, short of the 1.96 the rule asks (4 non-overlapping windows); out of sample it loses to the historical mean at every horizon (R²_OS −0.05 at one year, 39 windows; −0.38 at 10 years, 2 windows); the record cannot tell it apart from the earnings yield less the real yield at any horizon (encompassing p 0.44–0.73; a test with little power on this few windows, so this is not evidence that it adds nothing). Its benchmark — the earnings yield less the history's real yield, the 10-year less trailing ten-year CPI inflation, not the TIPS yield — does no better on the same months (out-of-sample R² −0.18 at 10 years, 2 windows), though on the long record it scores +0.26 since 1926 (9 windows).

  • 10-year Treasury

    None out of sample: an identity — a path's average short rate plus a premium — so what can be wrong is an input. The ACM model that supplies the premium fits today's par 10-year to 0.9bp (5.111% fitted, 13 Nov 2026).

  • Corporate credit (GZ spread)

    In sample (the Board re-estimates the split on the whole history with each update): as a 12-month target the fair spread missed by 75bp root-mean-square against 85bp for today's spread, a ratio of 0.89, over 633 months since 1973. After the 61 months with the premium at or below today's, it rose over the next 12 months in 87% (median +25bp).

  • Gold (GLD)

    Model level — fails the rule. In monthly changes, Feb 2006 – Oct 2026 (249 months, R² 0.31): real yield −7.53% per point (t = −4.6); dollar −1.04% per 1% (t = −4.6); breakeven −1.84% per point (t = −1.1); drift +0.84% a month. As a 12-month target the level's error is 1.34 times the random walk's over 178 origins, and the gap between the price and the level has shown no tendency to close (ADF p 0.71). Since Jan 2022 the model leaves +57% unexplained.

  • Crude oil (WTI)

    Slope +0.95% of price per 1% of stocks (t = 1.69, R² 0.02); +1.43% (t = 0.9) to Apr 2007 and −1.56% (t = −8.3) since Apr 2007; as a 52-week target, error 2.00 times the random walk's over 1729 weeks.

  • The dollar

    No source the site reads carries foreign interest rates, so a rate-differential model has no inputs.

About this exhibit
A table of a test of whether each classification's states separate later 13-week S&P 500 returns (p-value and windows, for the sign rule and the correlation model), then each target model's record as text
The test that picks the classifier, and each target's record
Does a classification's state separate what the S&P 500 did over the next 13 weeks? Point-in-time states, non-overlapping windows, 38 circular shifts of the states as the null.
Source: Yahoo Finance; U.S. Treasury; New York Fed; Federal Reserve Board; EIA; FactSet; Robert Shiller; Macro Review calculationscross.playbook-7

The playbook as bars

The median after a start of today's pattern against any week, asset by asset

Exhibit 13 · Each asset's 13-week return after current-pattern starts, against any week

Median log return, %, over the 13 weeks after each of the 7 starts since 18 Nov 2016, the latest included (bars), against any week (diamonds)
About
About this exhibit
Bars of each asset's median 13-week log return after the starts of the current sign-rule pattern, against the any-week median (diamonds)
13 weeks after a start of today's state, against any week
Median log return over the 13 weeks after each of the 7 starts since 18 Nov 2016, today's included (bars), against any week (diamonds).
Source: Yahoo Finance; Macro Review calculations · data to 13 Nov 2026cross.playbook-1

Appendix

What is copied from where, what the playbook conditions on and why, and what a target is on this site89 checks, all pass
About this page

Model-conditional, and not a forecast or advice: what followed earlier episodes of today's sign pattern, named by its legs, with the count printed (the site's regime is the four-regime map), and what the site's models say each asset is worth — each target printed only with its range, horizon, model and the dated test that would change it; the S&P 500's value in one register everywhere.

Exhibit 14 · What is copied, and what each object claims

  • Model-conditional

    Model-conditional, and a description of history: the playbook prints what followed earlier episodes of today's market state, and each target is the level a stated model implies on stated inputs, printed with its range, horizon, model and the dated test that would change it — not a forecast, not a position and not advice.

  • The playbook

    The regime page's playbook, copied: every run of the sign rule's state lasting 3 weeks or more since 18 Nov 2016 (505 weeks to 13 Nov 2026), and for each asset the median and interquartile range of the log return over the 4, 13 and 26 weeks after each run began, with the number of runs that had a full window. The unconditional median is over every week of the sample. A cell of fewer than 5 runs is flagged thin.

  • Episodes

    One definition of an episode, used wherever the site counts them: a run of the sign rule's state lasting 3 weeks or more, counted from its first week, on the weekly sign rule from 18 Nov 2016 to 13 Nov 2026. The separation test samples every 13th week instead, so its counts are windows, not episodes, and are printed as windows.

  • Why the sign rule

    Why the sign rule. The regime page tests whether a classification's states separate what followed: 13-week S&P 500 log returns on non-overlapping windows, states labelled point in time (no week labelled by a model that had seen it), Kruskal–Wallis with every one of the 38 distinct circular shifts of the labels as the null, so p cannot fall below 0.0256. The sign rule's quadrants give p = 0.0256 on 39 windows; the regime model's patterns p = 1.00 on 38. The test is why the playbook conditions on the sign rule rather than on the regime model. The site's regime is the four-regime map (the author's decision of 18 Oct 2026 (D-2026-10-18-TB8)); this page and Asset outlook still read the sign rule's states, named by their legs, until they are rewired to it.

  • Targets

    A target on this site is the fair value of the board — the level a stated model implies on stated inputs — printed only with its range, the range's kind, its horizon, the dated test that would change it and the model. An asset with no model, a model that failed its stability test, or a graded model below grade A (gold's, and the S&P 500's since its gap has a published record to be graded on) prints that and no target, with its reason and a link to the record; no other measure is put in its place. A range's kind says what it is: an input range is the spread of estimated inputs carried through the model; an assumption or scenario range is declared values or named paths, not a probability; a backtest error is the rule's out-of-sample miss over its record. The targets are copies of the fair-value board.

  • Policy paths

    The rate paths the site prices, copied from Scenarios: each path's 10-year is its fair value on the policy-rules grid, and the S&P 500 under it is the valuation identity with everything but the 10-year held. The spread across them is a scenario range — named paths, not a probability. 4 named branches are not priced through the identity: no ten-year pin: oil is not modelled through the identity — no elasticity of index earnings or of the premium to it exists among the site's estimates, and none is invented here. Term premium up is listed and not priced: no decision by the author.

  • Range with no view

    The range with no view: the cone's 10th–90th percentile over 52 weeks from the Range of outcomes page (zero drift, 13 Nov 2026), in percent for prices and in basis points for the 10-year yield — an input range, the spread of the volatility model's simulated shocks. It is printed beside each target so the target's distance can be read against how far the market typically moves in a year.

  • Copies

    Every number on this page belongs to another page and is copied through the site's manifest with its path, then re-read from disk at each update and compared.

Source: Yahoo Finance; U.S. Treasury; New York Fed; Federal Reserve Board; EIA; FactSet; Robert Shiller; Macro Review calculationscross.playbook-8

Before you read

Macro Review is a personal research site, still being built.

Nothing on it is investment advice, a recommendation or a solicitation.

Check every figure for yourself. Each names its publisher — confirm it there before you rely on it.

Figures can be revised, delayed or wrong.

Cones, fans, scenarios, fair values and “what followed” tables are conditional on the model each page names: they describe, they do not predict.

No warranty is given; use is at your own risk.