Asset Mapping Rationale

Macro Terminal's asset mapping layer answers one specific question: once a macro regime has been determined, what directional bias does it imply for each asset, which indicators drive that bias, how strong the evidence is, and under what conditions the bias should be judged invalid.

This is not a list of market calls. The mapping layer outputs conditional rules: the same asset takes a different bias under a different regime, and every bias must be bound to three things — observable driver indicators, verifiable evidence sources, and a falsifiable invalidation condition. Missing any one of them, a mapping is only narrative.

1. Method Framework: Why Bias Must Be Regime-Conditional

The problem with the conventional "macro view → asset recommendation" chain is that it hard-codes a point-in-time judgment into an intrinsic property of the asset: "gold hedges inflation", "small caps benefit from rate cuts" — statements that cannot be falsified. The mapping layer replaces that chain with three constraints:

Bias binds to the regime, not to the asset. A single card carries an independent value across the mapping matrix's seven defined regimes — stagflation risk, growth scare, goldilocks, overheating, Reflation Watch, disinflation, and soft landing — plus a neutral fallback. Soft landing is not a state the pillar logic defines on its own; it is a coverage state reached only by promoting the neutral baseline through a positive-evidence gate (see the macro-regime methodology page for the gate's conditions). "Gold is bullish under an easy-liquidity regime" is a legitimate methodological statement; "gold is bullish" is not.

Bias must resolve to indicators. Each card carries its own confirming indicators. If every asset shared the same regime-level drivers, the evidence on all 18 cards would be near-identical, which adds no information.

Bias must carry its own invalidation conditions. State which indicators moving in which direction overturn the premise. That supplies exit discipline and makes the boundary of applicability explicit.

2. Evidence Grading

How much a bias can be trusted depends not only on internal logic but on the nature of the data supporting it. Each card is tagged with one evidence basis type:

Each card also carries a pending indicator count: the number of confirming inputs that are known but not yet connected. The larger the gap, the less suitable the card is as a standalone basis for a decision. Grading exists to stop judgments of different quality from being used interchangeably — two biases pointing the same way but graded differently should not carry the same risk budget.

3. Crowding: Measures Fragility, Does Not Generate Direction

Most cards list "positioning is excessively crowded" among their invalidation conditions. Crowding here has a precise definition.

Single-contract crowding = the robust z-score of that contract's CFTC Managed Money — Leveraged Funds, for financial contracts — net position expressed as a percent of open interest (net % OI):

z = (current net % OI − rolling median) / (MAD × 1.4826)

The window is 260 weekly observations (5 years), it excludes the current point, and the result is winsorized to ±4. Median and MAD are used instead of mean and standard deviation so that a single positioning blow-off cannot inflate the yardstick every later reading is measured against; dividing by open interest is what makes different contracts and different periods comparable at all. Managed Money covers CTAs, commodity funds and trend-following hedge funds — the most representative trend-following flow in futures markets. Large positive values are long crowding, negative values are short crowding, and readings near zero impose no constraint.

Group crowding exists for exactly two baskets — agriculture and softs — and each is the equal-weight mean of its members' signed z. Because members keep their sign, a group reading has a direction: positive means the speculative crowd across the basket is net long, negative net short, and the magnitude is how many member sigmas that lean is worth. The composite is not re-standardized — a mean of z-scores is already on the z axis, so running a second z over it would measure "is this composite unusual for itself" rather than "how far is this crowd leaning", which is a different question. One consequence is worth stating plainly: averaging three correlated members compresses dispersion, so the shared crowding bands bite slightly later on these two legs than on a single-contract leg.

A retired pair of series sits behind them. Until 2026-08 the two group composites averaged their members' |z| instead, which made them magnitude gauges: they said how stretched the basket was, never which way, and a crowded short registered identically to a crowded long. Those series (cftc-ags-crowd, cftc-softs-crowd) are frozen read-only — their history is still charted on their own indicator pages — but no card reads them any more, and nothing writes to them.

Energy crowding is a single WTI contract series rather than a composite, and metals and livestock have no composite crowding series at all — those families are read member by member. Signal labels then combine the z-score with price trend, open interest (OI) change and commercial hedging behavior to classify the reading as long crowding / short crowding / weak trend / positioning watch / positioning confirmed.

Core principle: crowding measures fragility only. High crowding says only that a reverse catalyst would likely produce larger moves; it cannot decide long or short on its own. That is enforced in the wiring rather than left to prose: a crowding input can neither support nor neutralize a card's bias, and it does not change the card's conviction either. What it does instead is carry its own published label. The correct use is cross-confirmation: price trend and momentum, OI change (expansion supports trend quality, contraction shows the move is driven by existing positions), commercial hedging direction, and the macro drivers.

The fragility label

Every card that has a CFTC contract behind it publishes a signed fragility label for each crowded leg. The label has three parts and no score:

• Side — which way the crowd is positioned, read from the sign of the leg's signed net position: long crowding (z above the band) or short crowding (z below it). Every crowding leg on every card now carries a side, the two family composites included: they average their members' signed z (§3 above), so their sign is the basket's crowd side rather than a distance from norm. That was not true before 2026-08 — the retired |z| composites published a tier and nothing else, because a negative reading there was the least crowded five-year state and not a crowded short, and inferring a side from it would have inverted the meaning. A leg whose series genuinely has no side still degrades that way by design; there simply is no such leg on a card today.
• Tier — crowded once |z| reaches the shared signal band, extreme crowding once it reaches twice that band. Both bounds are inclusive. Two tiers, not a continuum: the label is a warning, and a warning that varies smoothly is a number pretending to be a warning.
• Alignment, read against the card's own published direction — the part that carries the information:
• Aligned: the crowd is on the same side as this card's call. The thesis is the consensus trade. A short-crowded leg on a bearish card is the squeeze case; a long-crowded leg on a bullish card is the profit-taking case. Either way, an unwind hits this view directly, and that is the reading that should shorten a holding period, cut size or raise the bar for entry.
• Opposed: the crowd is on the other side. An unwind is a tailwind here, not a threat. Positioning is still stretched — the market can still move violently — but the violence points the card's way.
• No direction: the card publishes no direction at all (no signal, or conflicting evidence), or — reserved, and unused on any card today — the leg itself has no side to compare. The crowd is recorded; there is nothing to align it against.

Alignment is not simply "long crowd ⇒ bullish card". That identity holds only when the card's payoff axis is the same as the futures contract's price axis, and each leg declares its mapping explicitly rather than inheriting it. The curve steepener is the case where it inverts: the crowding leg is net 10-year note futures, but the card pays off on the 2s10s slope. A crowded long in 10s unwinding means selling 10s, which steepens the curve — a tailwind. A crowded short unwinding is short covering, which bids 10s, flattens the curve, and is the squeeze that hits this card. So on the steepener a crowded short is the aligned, dangerous side — the opposite of the duration card, which reads the very same series.

Where multiple legs are labelled, the card lists them most actionable first: aligned before opposed before no-direction, then extreme before crowded. A merely crowded leg pointed at the card's own thesis is worth the first glance more than an extreme one pointed the other way.

The label never changes the bias and never changes the conviction. That is deliberate, and it is a change from earlier versions of this methodology, which discounted conviction by one notch when a crowding leg fired. Two reasons for the change. First, conviction is meant to answer one question — how good is the evidence for this call — and folding a positioning fact into it made the number answer two questions at once and neither of them cleanly. Second, and decisively, a one-notch discount cannot express alignment: it renders "the crowd is on my side" and "the crowd is on the other side" identically, when those are close to opposite facts. A signed label can say which one it is; a scalar penalty cannot. So conviction went back to being a pure evidence statement, and fragility became its own channel.

Practical reading: conviction sizes the confidence, the fragility label sizes the path. A high-conviction card with an aligned extreme-crowding label is a good call in a crowded trade — the destination may well be right and the route violent. That is a real and common situation, and it deserves to be shown as two facts rather than averaged into one.

4. Asset Grouping

The 18 assets are sorted into five groups by shared drivers. Assets in the same group share primary driver variables, their biases tend to point the same way, and their risk budgets must be aggregated.

Gold and the dollar sit in different groups but are two outlets of the same variable set along the transmission chain: real rates and dollar strength constrain both at once, so the two biases must be checked as a pair.

5. Asset-by-Asset Mapping

The bias values below correspond to the Reflation Watch regime. Under a different regime each card re-derives its bias by its own rules; the card structure is unchanged.

Front-End Rates / SOFR

Start with whether the policy corridor (EFFR / SOFR / IORB) is stable, then check whether short-end funding prices are being disturbed by fiscal supply or the liquidity plumbing. A sustained rise in SOFR relative to IORB shows secured funding is tightening, which requires a parallel check of reserves, the Treasury General Account (TGA) and overnight reverse repo. A brief spike may be nothing more than a month-end or quarter-end effect; only sustained widening accompanied by falling reserves or a rising TGA shows the constraint is transmitting. 2Y / SOFR positioning turning excessively crowded raises reversal or timing risk on this trade; it does not by itself invalidate the macro thesis.

Duration

The 10Y nominal yield is the discount-rate anchor for global asset pricing, but it must be decomposed into three components. A rising real rate directly suppresses long-duration assets, gold and high-multiple growth. A rising breakeven with the real rate unchanged carries a stronger read-through for commodities and cyclicals. A rising term premium usually traces to fiscal supply, inflation uncertainty, balance-sheet runoff or duration supply-demand, and is the least friendly of the three for long bonds. The invalidation condition requires both legs to hold at once, because either one alone admits the opposite interpretation.

TIPS Real Duration

The duration card is a nominal exposure: its payoff mixes the real yield, inflation compensation and term premium in one number. This card splits the real-yield leg out and prices it on its own, so the two rates cards can disagree without either being wrong — nominal is approximately real plus breakeven, and the breakeven leg reconciles them.

Two disclosures are load-bearing and are not optional caveats.

A breakeven is not an investable return. Inflation compensation enters this card as evidence for decomposing the nominal yield, never as a payoff. A rising breakeven with the nominal yield unchanged mechanically implies a falling real yield, which is why it reads as support for the bias; it does not mean a breakeven position is being expressed.

In a liquidity crisis, the breakeven measures the wrong thing. The 2008 and March 2020 breakeven collapses were driven principally by the TIPS liquidity premium — the market's willingness to hold an off-the-run, less liquid inflation-linked bond — not by a genuine collapse in expected inflation. In those windows measured real yields spike for a reason that has nothing to do with the macro thesis, and this card's read is contaminated for as long as the dislocation lasts. Cross-check against the funding and liquidity indicators before treating such a move as information.

There is also a structural asymmetry against the other rates cards worth stating in the open: TIPS carry no CFTC contract, so this card has no crowding leg and can never display the signed fragility label that front-end rates, duration and the steepener all display. Its fragility check is therefore weaker than theirs by construction, not by oversight — a reader gets no positioning warning here, and the absence of a label must not be read as an absence of crowding. The compensating discipline is deliberate: no strong-evidence rule is registered for the card, so its coverage grade tops out at "Adequate" and its published conviction can never reach "High". Initial conviction is capped at low/medium across all eight regimes and should be revisited only once a quarterly forward-return check for the card exists.

Under stagflation risk the card is deliberately neutral while nominal duration is strongly bearish. That is not a contradiction: both legs of the identity push the nominal yield up, but the real leg alone is genuinely two-way — the growth shock argues for easing and a falling real yield, the inflation shock argues for tightening and a rising one, and which dominates is a policy-reaction question this card cannot settle from its own evidence.

Curve Steepener

The change in shape does not constitute a direction; the source does. A steepening curve must be broken out by source — a falling 2Y is cut expectations or weakening growth (bull steepening), a rising 10Y is term premium, fiscal supply or reflation (bear steepening), and the two carry opposite implications for equity and credit. 10Y-3M sits closer to the traditional recession-probability framework; a rapid steepening there requires a simultaneous check of initial jobless claims, credit spreads and financial conditions.

S&P 500

The bias on broad equity is driven by the risk budget, not by earnings forecasts. Growth, employment, housing and consumption supply the earnings backdrop; real rates, bank lending standards, financial conditions and credit spreads set valuation and funding pressure; volatility and futures positioning judge whether risk is already crowded. High yield spreads are the high-beta read on risk appetite, and widening typically leads the contraction in equity risk budgets. S&P futures positioning turning excessively crowded raises reversal or timing risk; it does not by itself invalidate the thesis.

Growth / Nasdaq

Growth is the purest long-duration asset inside equity, more sensitive to real rates than to near-term earnings: the constraint on the S&P comes mostly from the risk budget, the constraint on growth mostly from the discount rate. Services and shelter inflation are the core of inflation stickiness and the focus of the policy reaction function; as long as services / shelter CPI is not falling, adding to high-multiple growth exposure is premature.

Small Cap / Cyclical

This card relies on growth diffusion, not on the level of growth. An upward revision to the GDP nowcast shows high-frequency data supporting growth resilience. A recovery in manufacturing orders normally precedes improvement in industrial production. Building permits lead starts; permits stabilizing while starts lag is an early-repair signal, while both weakening together shows rate pressure is still transmitting. The card has macro evidence only and no price confirmation, which makes it the most exposed to "right thesis, wrong timing".

Gold

Gold's pricing framework is the dual constraint of "real rates + dollar"; inflation expectations enter only indirectly, through the real-rate path. Work in order: use core CPI and median CPI to confirm whether underlying inflation is improving, then use breakevens to fix the real-rate path, then check whether the dollar and metals positioning resonate. Pay particular attention to divergence between the two — both rising together is the clearest suppressive combination; when one rises and the other falls, mark confidence down rather than picking whichever supports the existing position. Metals positioning turning excessively crowded while OI does not confirm raises reversal or timing risk; it does not by itself invalidate the thesis.

Copper / Industrial Metals

Copper carries a cyclical character and a financial character at the same time. On the cyclical side, read the China credit proxy against global leading indicators: money supply improving while activity indicators stay flat shows credit transmission is still weak; only a joint recovery genuinely supports the China chain and the commodity chain. Copper's positioning reading separates a move supported by the cycle from a move driven by chasing flows: alignment with the macro drivers is confirmation, extreme positioning with macro not following is a warning.

Crude Oil

Crude must be read as price and inventories together: price rising with inventories drawing points to demand resilience or supply tightness, and trend quality is higher; price rising while inventories build is more likely geopolitical or financial pricing, and persistence should be discounted. The comparison between natural gas and crude separates a single crude shock from broad energy inflation — oil and gas rising together transmits more readily into inflation expectations, which in turn constrains the rate-cut trade.

Agriculture

This is the group where macro has the least explanatory power, supplying only two background constraints: dollar strength and the direction of global demand. Real trade confirmation comes from contract-level evidence — Managed Money positioning, producer / merchant hedging, open interest and crowding, plus weather and supply information. The correct reading is not "macro is bullish agriculture" but "the macro backdrop does not argue against it; direction is decided by contract-level evidence".

Softs

More extreme than agriculture: macro generates almost no direction and supplies the dollar as its only background constraint. The "Unconfirmed" grade is an explicit acknowledgment that the evidence is thin. OI has its highest diagnostic value in this group: expansion alongside a price trend shows new money entering and better trend quality; contraction while price still moves shows existing positions are driving it, and the trend is fragile.

US Dollar

The dollar transmits more widely than any other variable, entering the drivers or invalidation conditions of seven cards — gold, copper, agriculture, softs, euro, yen and EM. A rising broad dollar tightens global dollar liquidity and suppresses non-US assets and financial demand for commodities. REER enters the card as its 6-month change, a confirming leg read against the series' own historical norm: stronger-than-usual real appreciation (or milder-than-usual depreciation) supports the card's positive side — the dollar strengthening — and the opposite supports the negative side. The REER level measures the dollar's real purchasing power and external competitiveness, but it is shown only as valuation context and does not vote — a high real dollar is an expensive dollar, which is mean-reversion risk rather than trend evidence. Once the dollar bias changes, every card that uses the dollar as an invalidation condition must be re-checked.

Euro

Euro pricing is essentially the mirror of the US advantage over Europe in growth, rate differentials and policy, and rarely generates an independent direction. It stays on watch mainly because independent confirming evidence is thin. The euro area REER enters as its 6-month change, a confirming leg: stronger-than-usual real appreciation (or milder-than-usual depreciation) relative to the series' own historical norm supports the card's positive side — the euro strengthening — and the opposite supports the negative side. The REER level is valuation context and does not vote; a persistently high level erodes external-demand competitiveness — the self-limiting mechanism of euro appreciation and a check on whether the trend is sustainable — but that is a slow valuation drag, not a trend signal.

Japanese Yen

The core drivers are rate differentials and the carry trade, so US real rates matter more than domestic Japanese variables. The yen REER enters as its 6-month change, a confirming leg read against its own historical norm: stronger-than-usual real appreciation (or milder-than-usual depreciation) supports the card's positive side — the yen strengthening — and the opposite supports the negative side. The yen's typical 6-month change is negative, so a milder-than-usual decline already reads as mild support for the positive side. The REER level is valuation context and does not vote: when it is excessively weak, imported inflation and carry-unwind risk warrant extra attention, but that is a watch item, not a directional vote. Yen positioning offering no reverse confirmation is a fragility flag, not an invalidation condition: in a carry unwind, positioning usually moves before price, so its absence raises reversal risk without overturning the drivers above.

EM / China Assets

The bias is the intersection of a dollar condition and a growth-diffusion condition, and both must hold. A weaker dollar without diffusion typically delivers no more than a brief valuation repair. Improving activity alongside a strengthening dollar leaves tighter liquidity to offset most of the activity dividend. On the China side, judge from the credit proxy together with leading indicators: money supply improving while activity indicators stay flat shows credit transmission is still weak.

HY Credit

This card requires simultaneous confirmation from multiple sources, not a single-indicator trigger. The four indicators cover different links in the transmission chain: HY-IG strips out the rate effect and shows whether the market is starting to discriminate on asset quality; the financial conditions index captures marginal changes in funding, leverage and risk-bearing capacity, where direction matters more than level; bank lending standards are low-frequency but verify whether credit stress is reaching real-economy lending; initial jobless claims supply cross-validation from the employment side. Deterioration on this card should be treated as an advance warning for the S&P and for growth.

Credit Quality Spread (BB / CCC)

This is a late-cycle fragility reading, not a direction on credit. The HY Credit card above prices outright exposure: long high yield, bullish when spreads compress. This card prices a relative-value leg inside the same index and is bullish when the low-quality tier moves apart from the high-quality one. Positive here means "low-quality credit is expected to underperform high-quality credit" and says nothing about whether credit as a whole is going up or down. The two cards are therefore complementary rather than duplicative, and they can both read positive without contradiction.

Evidence overlap audit, stated in the open. The CCC-BB spread was already wired into the HY Credit card before this one existed, so the two cards do share evidence. The intersection is exactly two indicators — the CCC-BB spread and NFCI — and both are read at different roles for different questions. On HY Credit the CCC-BB spread is a risk leg: a fragility check on an outright long, able to cap conviction. Here it is the primary leg, because it is this card's payoff variable; the same widening that argues against the outright long is the return on this trade. NFCI is primary on HY Credit and deliberately demoted to context here, since broad financial conditions inform a quality spread without pricing it. The other HY Credit primaries — HY-IG, bank lending standards, the HYG/JNK trend — are not reused at all, and this card's own confirming and context legs do not appear on HY Credit. Two further mitigations are structural rather than promised: both cards sit in the credit group, whose rule is that same-group risk budgets are aggregated, not added, so a reader cannot spend the shared evidence twice; and the coverage cap below means this card can never publish a conviction high enough to outweigh the card it overlaps.

No crowding leg, and the compensating discipline. Credit indices carry no CFTC commitment-of-traders series — COT covers futures, and there is no BB or CCC futures contract — so this card can never display the signed fragility label that the rates, equity, commodity and FX cards display. Its fragility check is weaker than theirs by construction, not by oversight — the absence of a label here means "not measurable", not "not crowded". As with TIPS real duration, no strong-evidence rule is registered for the card, so its coverage grade tops out at "Adequate" and its published conviction can never reach "High". Matrix conviction is capped at low/medium in all eight regimes.

Two cells are worth stating explicitly because they are where the naive version of this trade goes wrong. Under overheating the card is neutral, not positive: late cycle is when the fragility argument is most tempting and least supported by the tape, because growth is still carrying earnings and quality spreads historically stay tight through the hawkish repricing. Decompression lags that regime rather than coinciding with it, and the trade belongs in the growth scare. Under stagflation risk the card is positive but held at low conviction, and the reason is sample rather than logic: the US has had few clean stagflation episodes with a modern high yield index to measure, so the historical support for that specific cell is thin.

6. A Unified Reading of Invalidation Conditions

Placing the invalidation conditions of all 18 cards side by side yields four recurring patterns:

Shared invalidation variables also run across cards: the dollar and real rates are global constraints, credit spreads sit upstream of the equity group, and crowding is the fragility check on the commodity and equity groups. When a shared variable triggers, review the whole group rather than working through cards one at a time.

7. Usage Guidance

Confirm the prevailing regime first, then read the asset bias — read a bias detached from its regime and every piece of regime-conditional information is lost. Treat the evidence basis and the pending indicator count as confidence weights: "Market priced" and "Unconfirmed" should not be allocated the same risk budget. Aggregate exposure within a group: when several cards in the same group point the same way, what is being carried is a magnified exposure to one macro risk. Invalidation conditions must be written into execution discipline in advance. Crowding only ever subtracts — it can cut size, tighten stops or raise the bar for entry, but it should not be used to generate direction.