We’ve covered the core of the regime matrix, Growth and Inflation. Those two pillars give you the fundamental picture. Growth up, inflation down: Goldilocks. Growth down, inflation up: Stagflation. The combinations drive the playbook.

The fundamental picture alone doesn’t tell you enough. Two identical Growth x Inflation readings can produce completely different outcomes depending on what policy is doing. Growth positive and inflation stable sounds great until you realise policy is restrictive and tightening further. Then your “Goldilocks” is actually Late Cycle, and the clock is ticking.

Policy is the first of three modifier pillars (alongside Liquidity and Risk Appetite) that shape how the fundamental regime expresses across asset classes. If Growth and Inflation tell you what the economy is doing, Policy tells you what the authorities are doing about it. What the authorities do determines how long the current regime persists, how violently it transitions and which assets benefit or suffer along the way.

This is a tough pillar to get right, not because the data is hard to find, it’s actually the most transparent of all five pillars. Central banks publish their decisions, their projections, their minutes. Fiscal data is publicly available. The information is everywhere. The difficulty is knowing what to focus on, what to ignore, and how the different pieces interact.

Let me walk you through how I think about it now.

Why Policy Matters More Than You Think

I used to treat policy as background noise. The Fed cuts, the Fed hikes, markets react for a day and then fundamentals take over.

Policy doesn’t just respond to the economy. It shapes it and it shapes asset prices through channels that are far more powerful than the headline rate decision.

Think about it mechanically. When the Fed sets the overnight rate, it anchors the entire short end of the yield curve. That rate ripples into mortgages, corporate borrowing costs, auto loans, credit cards, every form of debt in the economy. When those costs change, so does behaviour. Businesses invest more or less. Consumers spend or save. Housing booms or busts. The transmission mechanism from a 25 basis point change in Fed Funds to actual economic activity is slow, messy, and uncertain but very real.

Then there’s the balance sheet. When the Fed buys Treasuries and MBS, it pushes investors out the risk curve. Yields compress, asset prices rise, financial conditions loosen. When it sells, or lets holdings roll off, the reverse happens. This is the plumbing that connects monetary policy to markets and it operates independently of the rate decision itself.

And fiscal policy, which most macro frameworks underweight or ignore entirely, puts money directly into the economy. Not through the banking system, not through credit channels, but directly: government spending, transfer payments, tax cuts. When the deficit widens, the government is adding demand. When it narrows, it’s withdrawing demand. This is arguably the most powerful short-term driver of growth and yet I see frameworks that track fourteen different Fed indicators and nothing on fiscal.

The reason I weight Policy at 20% of the regime composite (behind Growth at 30% and Inflation at 25% but ahead of Liquidity and Risk Appetite) is that it’s the bridge between fundamentals and market outcomes. It tells you whether the authorities are reinforcing the current trend or fighting it. And that distinction is the difference between a regime that persists and one that’s about to flip.

The Three-Layer Model

Just as the inflation pillar uses a pipeline (upstream, midstream, downstream), the policy pillar works in layers. Each one captures a different mechanism through which policy affects the economy and markets. Get comfortable with this structure, because it’s the architecture that makes everything else work.

Layer 1: Monetary Policy. The Fed’s rate decisions, forward guidance and balance sheet operations. This is what everyone watches, what CNBC covers on FOMC day, what moves markets in the short term. It matters, but it’s not the whole story.

Layer 2: Fiscal Policy. Government spending, taxation and the resulting deficit or surplus. This is the layer most macro analysts underweight. It’s harder to track in real-time, harder to quantify and harder to forecast. But its impact on growth is often larger and more direct than anything the Fed does.

Layer 3: Financial Conditions. The aggregate result of both monetary and fiscal policy, filtered through markets. Are conditions loose or tight? Is credit flowing or constrained? Are asset prices inflated or depressed? Financial conditions indices capture the net effect of everything the authorities are doing and everything the market is doing in response.

The power of this model is in the combinations. Monetary loose and fiscal tight is a completely different animal from monetary tight and fiscal loose. Each combination has distinct implications for growth, inflation, and asset prices. I’ll walk you through the specifics, but first let me explain each layer in detail.

Layer 1: Monetary Policy - Beyond the Headline Rate

Every six weeks, the Fed announces a rate decision. Markets move. Pundits opine and then the next day, everyone goes back to watching the same indicators they were watching before. The actual analysis of monetary policy gets lost in the noise of the event.

Here’s what I’ve learned to focus on instead.

Real Rates, Not Nominal Rates.

This is probably the single most important concept in the entire policy pillar, and it’s the one most retail investors miss completely.

The nominal Fed Funds rate (the number you see on the news) tells you almost nothing by itself. A rate of 4% is restrictive when inflation is 2% (real rate = +2%). The same 4% is accommodative when inflation is 5% (real rate = -1%). The economy doesn’t care what the nominal rate is. It cares what the rate is relative to inflation.

Real Fed Funds = Nominal Fed Funds minus Core PCE. That’s it. Simple arithmetic, but the implications are profound.

When real rates are deeply negative, say below -1%, the Fed is effectively paying people to borrow. Money is free in real terms. That’s aggressively accommodative. When real rates are significantly positive, above +1%, the Fed is making borrowing genuinely expensive. That’s restrictive. Between those thresholds, policy is roughly neutral.

The scoring in my framework reflects this directly. Real Fed Funds below -1% scores as accommodative (+1). Around zero scores as neutral (0). Above +1% with a shrinking balance sheet scores as restrictive (-1). The level matters, but the direction matters just as much, real rates falling from +2% to +1% is still restrictive, but the trajectory is easing.

Source: FRED FEDFUNDS, FRED PCEPILFE | Real Rate = Fed Funds Effective − Core PCE YoY

I used to just track the nominal rate and the dot plot. That cost me money in 2022 when the Fed was hiking aggressively but real rates were still negative because inflation was running above the funds rate. Everyone was screaming about “tight policy” but real rates told a completely different story. Policy wasn’t actually restrictive until well into the hiking cycle. If you were positioned for “tight money” too early, you got run over.

Source: FRED FEDFUNDS, FRED PCEPILFE

The Yield Curve as a Policy Signal

The yield curve is probably the most over-discussed, under-understood signal in all of macro. Everyone knows an inverted curve “predicts recession.” Few understand why it works or, more importantly, when it stops working.

The basic logic is sound. The curve shape reflects the market’s expectation for future short-term rates. When the curve is steep, long-term rates much higher than short-term, the market expects rates to rise, which usually means growth and inflation are expected to pick up. When the curve inverts, short rates above long rates, the market expects the Fed will need to cut, which usually means a downturn is coming.

The specific spread I watch most closely is the 10-year minus 3-month. It has the cleanest historical track record as a recession signal. The 2s10s gets more attention, but the 10Y-3M is more mechanically tied to Fed policy, the 3-month rate is essentially the Fed Funds rate with a small premium, so the spread captures the gap between where the Fed is and where the market thinks the economy is headed.

But here’s the nuance that most people miss: the inversion itself isn’t the recession signal. The un-inversion is. Historically, recessions don’t begin when the curve inverts, they begin when the curve steepens after a prolonged inversion. That’s when the market starts pricing in emergency cuts. That’s when the data actually breaks.

We saw this play out in late 2024 and early 2025. The 10Y-3M had been inverted for months. Then it steepened aggressively as the Fed started cutting. The curve going from -50 basis points to +50 wasn’t an “all clear” it was the market saying “the Fed is going to have to ease a lot more than you think.” That’s a very different signal, and getting it right matters for positioning.

Source: FRED T10Y3M, FRED T10Y2Y | Red shading = inversion periods

I weight the curve as a confirmation tool, not a leading signal. It tells me whether the bond market agrees with my fundamental assessment. If I’m reading growth as positive but the curve is deeply inverted, one of us is wrong and the bond market has a better track record than I do.

Forward Guidance and the Dot Plot

This is the part of monetary policy that’s changed the most over the past two decades. The Fed now communicates extensively about future intentions, through the dot plot, the Summary of Economic Projections, press conferences, and individual governor speeches. This forward guidance is arguably more important than the rate decision itself, because markets are forward-looking.

The dot plot tells you what each FOMC member expects the funds rate to be at year-end over the next few years. The median dot gets the headline. But the dispersion matters more than the median. When dots are tightly clustered, there’s consensus, the Fed is confident in its path. When dots are widely spread, there’s disagreement and disagreement means uncertainty about the reaction function.

I watch the gap between market pricing and Fed guidance closely. When futures are pricing significantly more cuts than the dots suggest, the market is betting the Fed will be forced to ease by deteriorating data. When futures are pricing fewer cuts, the market thinks the Fed is too dovish. This gap is a tension that resolves one way or another and the resolution drives rates volatility.

The December 2025 FOMC was a perfect example. The dot plot showed only one cut for 2026. The market had been pricing three to four. That gap, between what the Fed said and what the market believed, drove an 87 basis point swing in the 2s10s spread over a matter of weeks. If you were paying attention to the guidance mismatch, you were prepared. If you were just watching the headline rate, you were blindsided.

Source: Federal Reserve SEP, CME FedWatch Tool

Balance Sheet Operations

The rate is the price of money. The balance sheet is the quantity. Both matter.

When the Fed buys bonds (QE), it increases the supply of reserves in the banking system. Banks have more cash, lending conditions ease, asset prices tend to rise. When the Fed shrinks its holdings (QT), either by selling or letting bonds mature without reinvestment, reserves drain, conditions tighten, and the risk premium on financial assets tends to increase.

I track the balance sheet as a year-over-year change. Expanding balance sheet = accommodative. Shrinking = restrictive. The pace matters too. A gradual runoff is very different from active selling. The Fed learned this the hard way in 2018-19 when “autopilot” QT nearly triggered a funding crisis.

Source: FRED WALCL

For the regime framework, the balance sheet is a binary overlay. If the balance sheet is expanding year-over-year, that’s a positive signal regardless of the rate level, the Fed is actively adding liquidity. If it’s contracting, that’s a drag even if rates are low. The combination of rate stance and balance sheet direction gives you the full picture of what the Fed is actually doing, not just what it’s saying.

Layer 2: Fiscal Policy — The Elephant Nobody Watches

This is the layer I was slowest to incorporate into the framework. For years, I basically ignored fiscal policy. The Fed was the show. Fiscal was something economists argued about and politicians used as a talking point. It didn’t seem tradeable.

I was completely wrong. And the 2020-2023 period proved it.

When the US government ran 15%+ of GDP in deficit spending during COVID, it didn’t matter what the Fed was doing with rates. The fiscal fire hose overwhelmed everything. Growth surged. Inflation surged. Asset prices surged. The monetary policy signal was noise compared to the fiscal signal.

That was an extreme case, but the principle holds in normal times too. A budget deficit of 6% of GDP, which is roughly where we’ve been recently, is meaningful demand injection even when rates are at 5%. The government is still adding to spending, still employing people, still transferring money to households. That has a direct and immediate effect on growth that monetary policy can only influence indirectly through credit channels.

The Deficit Level vs The Fiscal Impulse

This distinction is to fiscal policy what the “level vs direction” lesson was to growth indicators. The deficit level tells you the current stance, how much the government is spending relative to what it takes in. The fiscal impulse tells you whether that stance is getting more or less stimulative.

The fiscal impulse is the year-over-year change in the deficit. A deficit that widens from 4% to 6% of GDP is stimulative, the government is adding more demand than last year. A deficit that narrows from 6% to 4% is contractionary, even though the government is still running a deficit, it’s adding less demand than before. The economy feels the change, not the level.

This is counterintuitive and trips up a lot of people. A 5% deficit sounds stimulative and in absolute terms, it is. But if last year the deficit was 7%, the economy is experiencing a 2-point fiscal drag. That drag matters for growth in exactly the same way that a rate hike matters, it’s removing stimulus that was previously supporting activity.

I track both. The deficit level tells me the structural stance. The fiscal impulse tells me the direction. When both are positive and rising, fiscal policy is a strong tailwind. When the deficit is large but narrowing, fiscal is shifting from tailwind to headwind regardless of the absolute number.

Source: FRED FYFSGDA188S, Treasury Monthly Statements, CBO

Why Most Frameworks Get Fiscal Wrong

Three reasons I see people consistently misread fiscal policy:

First, they confuse the political debate with the economic reality. Whether the deficit is “too big” is a political question. Whether the deficit is adding to or subtracting from demand is an economic question. I don’t care about the former for regime classification. I only care about the latter.

Second, they treat fiscal as slow-moving and therefore ignorable. It’s true that fiscal policy changes less frequently than monetary policy, there’s no fiscal FOMC meeting every six weeks. But when it does change, the impact is often larger and more direct. The fiscal cliff, sequestration, COVID stimulus, the IRA, these were all regime-level events that moved markets for months or years.

Third, they underweight the interaction between fiscal and monetary policy. This is where the Policy Mix Matrix comes in, and I’ll cover it in detail shortly. But the short version is: the same fiscal stance means very different things depending on what the Fed is doing, and vice versa.

Layer 3: Financial Conditions - The Net Effect

Financial conditions indices attempt to answer a simple question: taking everything into account, rates, credit spreads, equity prices, the dollar, volatility, how easy or hard is it for the economy to access money?

This is the layer where monetary and fiscal policy meet market reality. The Fed can cut rates, but if credit spreads widen simultaneously, conditions might actually tighten. The government can run a deficit, but if it crowds out private borrowing by pushing yields higher, the net effect could be neutral or negative.

I watch two main indices:

The Chicago Fed National Financial Conditions Index (NFCI) is the workhorse. It’s comprehensive, covering risk, credit, and leverage subcomponents. Zero means average conditions. Negative readings mean conditions are looser than average. Positive means tighter. It’s the one I weight most heavily because it captures the broadest range of inputs and has a good track record relating to economic outcomes.

The Goldman Sachs Financial Conditions Index (GS FCI) gets more attention in markets. It weights equity prices more heavily, so it tends to be more responsive to stock market moves. That’s both a feature and a bug—it captures sentiment shifts quickly, but it can overreact to equity corrections that don’t affect the real economy.

Source: FRED NFCI, S&P 500 | Negative NFCI = conditions looser than average

The key insight with financial conditions is that they’re the final arbiter of whether policy is actually working. The Fed can intend to tighten, but if markets front-run rate cuts and loosen conditions prematurely, the tightening doesn’t reach the real economy. Conversely, the Fed can cut rates, but if a credit event causes spreads to blow out, conditions tighten despite the cuts. Financial conditions capture what’s actually happening, not what policymakers want to happen.

A personal example: in early 2025, the Fed was on hold and QT was winding down. By any textbook measure, policy was restrictive. But financial conditions were loose, equity prices were elevated, credit spreads were tight, the dollar was stable. The economy was behaving as if policy was accommodative because markets had already done the easing. Watching only the Fed would have given you the wrong read. Watching financial conditions told the actual story.

The Policy Mix Matrix

This is the framework within the framework. It’s probably the single most valuable tool in the entire policy pillar, and I wish someone had explained it to me earlier in my career.

The idea is simple: plot the monetary stance against the fiscal stance as a 2×2 matrix. Four quadrants, four very different market environments.

Source: MacroAnalytix Regime Framework

Quadrant 1: Monetary Loose + Fiscal Loose. Maximum stimulus. This is the “money everywhere” environment. Growth gets a boost from both channels. Asset prices inflate. Risk appetite surges. The danger is bubbles and overheating. When both levers are pulling in the same direction, the economy can overshoot in ways that require painful correction later. We saw this in 2021, rates at zero, QE running at full speed, deficit spending at wartime levels. Everything went up. And then everything had to come back down.

Quadrant 2: Monetary Tight + Fiscal Loose. The tug-of-war. The Fed is trying to slow the economy while the government is stimulating it. This is where real rates tend to rise, the curve steepens, and the long end sells off. The growth picture stays muddled because the two forces partially offset each other. This has been roughly the configuration for much of 2024-2025: the Fed restrictive with real rates above +1%, but the deficit running at 6% of GDP providing a fiscal offset. The result? An economy that refuses to roll over despite “tight” monetary policy.

Quadrant 3: Monetary Tight + Fiscal Tight. Maximum restriction. Both levers are pulling demand out of the economy. This is the environment that produces recessions, credit events, and deep bear markets. Growth collapses, earnings fall, defaults rise. Duration works. Risk assets don’t. Historically, these periods are rare because policymakers usually react before both become restrictive simultaneously. But when they happen—think early 1980s under Volcker with concurrent fiscal tightening, the outcomes are severe.

Quadrant 4: Monetary Loose + Fiscal Tight. The healing environment. The Fed is easing while fiscal consolidation reduces demand. The net effect depends on the relative magnitudes, but typically this favours duration and quality growth, rates fall, the curve steepens, and the economy recovers gradually. Credit eases. This is roughly what you get in the early stages of a recovery when the Fed is still accommodative but the post-crisis fiscal impulse is fading.

I check the Policy Mix Matrix every week when I update the dashboard. It takes thirty seconds and tells you immediately whether the two biggest policy forces in the economy are working together or against each other. That single chart has saved me from more bad trades than any other tool in the framework.

Right now, as I write this, both monetary and fiscal policy are loose, the Fed has cut and the deficit remains elevated. The matrix flags this as “Bubble Risk.” That doesn’t mean a crash is imminent. It means the conditions exist for asset prices to overshoot fundamentals, and when both levers eventually need to tighten simultaneously, the correction could be sharp. Worth knowing. Worth sizing for.

Putting It Together: How I Score Policy

The scoring system integrates all three layers into a single pillar reading. Here’s the architecture:

Source: MacroAnalytix Regime Framework | Weights are approximate and subject to judgment

Monetary policy gets roughly half the weight. I look at real Fed Funds (primary signal), the yield curve (confirmation), forward guidance (direction), and the balance sheet (quantity). These combine into a monetary sub-score that ranges from accommodative to restrictive.

Fiscal policy gets roughly a quarter of the weight. I track the deficit level (structural stance) and the fiscal impulse (direction). A widening deficit with rising impulse scores as stimulative. A narrowing deficit with negative impulse scores as contractionary.

Financial conditions get the remaining quarter. NFCI and GS FCI together tell me whether the combined effect of monetary and fiscal policy is actually reaching the economy. Loose conditions confirm the accommodative read. Tight conditions override the headline numbers.

The composite score maps to the same -1/0/+1 framework as the other pillars:

+1 (Accommodative): Real Fed Funds below -1%, OR balance sheet expanding year-over-year, OR financial conditions meaningfully loose. The Fed is actively supporting. Growth tailwind.

0 (Neutral): Real rate around zero, financial conditions near average. Policy is neither helping nor hurting. The economy is on its own.

-1 (Restrictive): Real rate above +1% AND shrinking balance sheet. The Fed is actively restraining. Growth headwind.

As always, the trend matters as much as the level. Policy at 0 and moving toward accommodation is very different from policy at 0 and moving toward restriction. I track both the current score and the direction arrow in every weekly Setup.

Mistakes I’ve Made (So You Don’t Have To)

The policy pillar has its own set of traps. Here are the ones that caught me.

Fighting the Fed.

The oldest cliché in finance for a reason. When the Fed is actively easing, the path of least resistance for risk assets is up. When they’re actively tightening, it’s down. Simple. And yet I’ve repeatedly tried to be clever, shorting into easing cycles because valuations looked stretched, or going long into tightening cycles because growth looked strong.

The Fed moves slower than you think, but it moves more than you think. They don’t stop after one hike or one cut. The cycle continues until conditions force a change. If you fight the direction, you’re fighting the most powerful force in financial markets. I’ve learned to respect it, even when I disagree with the policy.

Watching FOMC Day Instead of the Reaction Function.

FOMC days are exciting. Markets move. Headlines fly. It feels like the most important day of the quarter. And it’s mostly noise.

What matters isn’t what the Fed did today. It’s what the Fed will do in response to the data that hasn’t come out yet. That’s the reaction function, the mapping from economic data to policy response. If you understand the reaction function, you can anticipate the Fed’s moves. If you’re just reacting to FOMC day, you’re trading yesterday’s news.

I spend far more time now thinking about what data would cause the Fed to change course than I do parsing the statement. What level of Core PCE makes them comfortable cutting? What unemployment rate triggers concern? What financial conditions reading makes them nervous? Those are the questions that matter. The statement is just a snapshot of where those answers stood six weeks ago.

Ignoring Fiscal Entirely

I already mentioned this, but it’s worth emphasising as a standalone mistake because I made it for years and it cost me real money.

In 2023, I kept expecting the economy to slow because the Fed was hiking aggressively. Real rates were high. The balance sheet was shrinking. Everything in my monetary toolkit said “restriction.” But the economy kept growing. Consumer spending held up. The labour market refused to crack.

The answer was fiscal. The deficit was running at 6%+ of GDP. The government was pumping money into the economy at a rate that partially offset everything the Fed was doing. If I’d been tracking fiscal impulse, I’d have seen the offset and adjusted my expectations. Instead, I kept waiting for the slowdown that didn’t come.

Now fiscal is a permanent part of the framework. Not as a forecasting tool—I don’t try to predict what Congress will do—but as a measurement of what’s actually happening. The deficit doesn’t lie.

Confusing Nominal and Real

I covered this in the real rates section, but it bears repeating as a mistake. I’ve watched otherwise smart analysts call policy “tight” during periods when real rates were negative, or “loose” when real rates were actually restrictive. The nominal rate is meaningless without the inflation context.

The fix is simple: always think in real terms. Always. When someone says “rates are at 5%,” your first thought should be “what’s inflation?” If inflation is 3%, real rates are +2%—that’s tight. If inflation is 6%, real rates are -1%—that’s loose. The absolute level of rates tells you nothing about the stance of policy.

Overreacting to Fed Language

The Fed communicates in code. “Patient” means something. “Flexible” means something different. “Data dependent” is the ultimate hedge. The cottage industry of Fed statement parsing is a real thing, and I got sucked into it for years.

Here’s what I’ve learned: the words matter less than the numbers. Real rates, the balance sheet, financial conditions, these are the data. The statement is the narrative the Fed wraps around the data. When the narrative and the data agree, fine. When they disagree, trust the data.

Powell can talk about being hawkish all day long. If real rates are falling and financial conditions are loosening, policy is accommodative regardless of what he says. The markets know this, which is why the reaction to statements is often counterintuitive. Don’t parse words. Track numbers.

From Policy Reading to Actual Positions

Here’s roughly how the policy pillar translates into cross-asset positioning:

When policy is accommodative (+1): Duration is less attractive because rates are low and likely to rise eventually. Risk assets benefit from cheap money and easy credit. Equities and credit outperform. The dollar tends to weaken as real rates fall. Commodities benefit from the demand support. Curve steepens as the front end stays anchored and the back end reprices for growth and inflation. This is the environment where you lean into risk, but watch for the Policy Mix Matrix. If fiscal is also loose, you’re in bubble risk territory.

When policy is neutral (0): The most ambiguous environment. Policy isn’t helping or hurting. Asset returns depend more on the Growth and Inflation readings. Position based on the fundamental regime, not the policy modifier. Maintain balanced positioning and don’t overweight any directional bet based on the policy stance alone.

When policy is restrictive (-1): Duration becomes increasingly attractive as the market prices in eventual easing. Risk assets face headwinds from tighter credit and higher borrowing costs. Quality over speculative beta. Cash becomes competitive, you’re getting paid to wait. Credit spreads tend to widen. The curve flattens or inverts as the front end gets pulled up by the Fed while the back end falls on growth fears. This is the environment for patience and defensiveness.

But always, always, overlay the Policy Mix Matrix. Restrictive monetary plus loose fiscal is a very different trade from restrictive monetary plus restrictive fiscal. The matrix tells you which combination you’re actually dealing with, and the cross-asset implications are distinct for each quadrant.

Building Your Own Policy Framework

Same principle as the previous two pillars. The architecture matters more than the exact indicators. Here’s how I’d build it from scratch:

Start with the three layers. Separate monetary, fiscal, and financial conditions. This is the structure that prevents you from overweighting any single policy dimension and ensures you capture the interactions.

For monetary policy, choose indicators that reflect the actual stance, not just the headline. You need something that measures the real rate (nominal rate minus inflation). You need a curve spread that captures the market’s forward view. You need a measure of balance sheet direction. And you should track market expectations versus Fed guidance to spot the tension that drives rates vol.

For fiscal policy, track the deficit level and the impulse. The level is easy, CBO publishes it regularly, and Treasury monthly statements give you near-real-time data. The impulse requires a simple year-over-year calculation that tells you whether fiscal is becoming more or less stimulative.

For financial conditions, pick at least one broad index. NFCI is my primary. Some people prefer the GS FCI or Bloomberg’s index. The specific choice matters less than consistently tracking the same measure over time so you develop intuition for what “tight” and “loose” actually look like in real-time.

Build the Policy Mix Matrix. Plot monetary versus fiscal on a 2×2. Update it weekly. It takes thirty seconds and is genuinely the most underrated tool in the entire framework.

Set your thresholds. What real rate level constitutes “restrictive” versus “accommodative”? What deficit level is stimulative? What NFCI reading counts as loose? Write them down before you need them. Same lesson as Growth and Inflation—explicit thresholds prevent you from moving the goalposts when the data is inconvenient.

Connect to the regime matrix. Policy alone doesn’t tell you the playbook. It’s a modifier that shapes how the Growth × Inflation picture expresses across assets. The same Growth/Inflation regime with accommodative policy is a completely different animal from the same regime with restrictive policy. That interaction is where the framework generates its best signals.

The Bottom Line

Policy is the bridge between the economy and markets. Growth and Inflation tell you what’s happening. Policy tells you what the authorities are doing about it and whether that help or hindrance is reinforcing the current regime or setting up the next transition.

The three-layer model, monetary, fiscal, financial conditions, gives you a complete picture. Real rates matter more than nominal rates. The yield curve confirms but doesn’t lead. Fiscal impulse is the most underrated variable in macro. Financial conditions capture the net effect of everything. And the Policy Mix Matrix tells you whether the two biggest forces in the economy are working together or against each other.

Right now, my Policy reading is 0 (Neutral), trend flat. The Fed is on hold, watching the data. Real rates are modestly positive. Fiscal remains stimulative through deficit spending, but the impulse has stabilised rather than accelerated. Financial conditions are loose. The Policy Mix Matrix flags both levers as accommodative, bubble risk territory in the classification, worth monitoring but not yet actionable.

Next time, we’ll cover Liquidity, the fourth pillar. The plumbing of the financial system. Central bank balance sheets, reserve dynamics, funding market stress and the indicators that tell you whether the system has enough grease to keep running. If Policy is what the authorities intend, Liquidity is what actually flows through the pipes. That distinction matters more than most people realise.

Until then, start building your own policy framework. What would you track for real rates? How would you measure fiscal impulse? Which financial conditions index would you choose? Map out the Policy Mix Matrix for the last five years and see which quadrant produced which market outcomes. The exercise is worth it.

Questions or feedback? macroanalytix.research@gmail.com

Disclaimer: All content published by MacroAnalytix is for research and educational purposes only. Nothing on this site or in our publications should be interpreted as financial advice, investment recommendations, or a solicitation to buy or sell securities. You are solely responsible for your own financial decisions. Please conduct your own due diligence and consult with a licensed financial professional before making any investment decisions.

Last time we covered Growth, the engine of regime classification. This time we’re tackling Inflation, the second axis of the regime matrix. Get these two right and you’ve got the fundamental picture. Growth up + inflation down = Goldilocks. Growth down + inflation up = Stagflation. Obviously liquidity and policy play a part but you get the picture.

But here’s the thing I wish someone had told me years ago: by the time inflation shows up in the CPI print, the trade is already stale. You’re seeing where prices were, not where they’re going. Everyone’s watching the same number, reacting to the same headlines, trading the same information. There’s no edge in that.

The edge (if there is one) comes from understanding inflation as a process, not a number. A pipeline with stages. Pressures that build upstream, flow through expectations and eventually show up in the sticky prices the Fed actually targets. If you can read the pipeline, you can see the regime shift before CPI day confirms it.

This took me longer to figure out than the growth framework and I’m still learning, but let me walk you through how I think about it now.

Why Inflation Is Different From Growth

Growth has a relatively clean relationship with asset returns. Economy expands, earnings grow, risk assets do well. Economy contracts, earnings fall, defensive assets win. Cause and effect.

Inflation is messier. The relationship with returns depends entirely on why inflation is doing what it’s doing and what else is happening at the same time.

Rising inflation with strong growth? That’s Reflation, commodities rip, value beats growth, duration suffers, but equities can still work. Rising inflation with weak growth? That’s Stagflation, pretty much everything hurts except gold and maybe energy. Same inflation direction, completely different playbooks.

Falling inflation with strong growth? Goldilocks, the best environment for risk assets. Falling inflation with weak growth? Could be Deflation risk, duration wins big, but you’re also worried about credit and equity exposure.

This is why the regime framework plots Growth against Inflation as a 2x2 matrix. Neither dimension alone tells you enough. You need both to know which playbook to run.

But it also means inflation analysis has to work differently than growth analysis. You’re not just asking “is inflation going up or down?” You’re asking “where are the pressures, how persistent are they, and what’s the interaction with the growth picture?” That requires understanding the mechanics in a way that pure growth analysis doesn’t.

The Mental Model That Changed How I Think About This

For years I tracked CPI and PCE like everyone else. Month comes out, number goes up or down, I update my view.

The breakthrough and I really do mean breakthrough, this was a genuine level-up in how I think, came when I started conceptualizing inflation as a pipeline with three stages.

Stage 1: Upstream. Input costs, commodity prices, shipping rates. This is where inflationary pressure enters the system. Raw materials, energy, transportation. The stuff that flows into everything else. When upstream is building pressure, it’s going to show up downstream eventually.

Stage 2: Midstream. Expectations. What bond markets, businesses and consumers believe inflation will be. This matters because expectations influence behavior. Wage demands, pricing decisions, spending patterns. If expectations de-anchor, transitory inflation can become permanent.

Stage 3: Downstream. The sticky prices consumers actually pay. Services, wages, rent. This is what the Fed targets and what CPI/PCE measure. By the time pressure shows up here, it’s already flowed through the first two stages. That’s why downstream is coincident or lagging, it confirms where you are, not where you’re going.

The power of this model is pattern recognition. Different combinations tell you different stories:

All three stages falling? Pipeline clearing - disinflation is coming, duration looks attractive.

Upstream rising but downstream stable? Pressure building - prepare to reduce duration before the CPI prints confirm it.

Upstream falling but downstream sticky? Normal lag - be patient, don’t overreact to current levels, the clearing is coming.

All three rising? Full pipeline pressure - Stagflation risk is real, get defensive.

Midstream de-anchoring while upstream is stable? That’s a credibility problem, the Fed has to act regardless of current inflation levels.

The framework gives you a bigger picture than any single number and it gives you lead time, which is the whole point.

Upstream: Where Inflation Begins

Inflationary pressures almost always start upstream. Commodity price spikes, shipping cost surges, supply chain disruptions, these are the first dominoes. By the time they show up in consumer prices, you’ve had months of warning if you were paying attention.

The 2021-22 inflation surge is the perfect example. The Baltic Dry Index, which measures shipping costs for bulk cargo went from around 1,000 in early 2020 to over 5,600 by late 2021. A 460% increase. ISM Prices Paid hit 92.1 in June 2021, the highest reading in decades. Anyone watching upstream saw the wave coming.

Source: https://fred.stlouisfed.org/series/PPIFIS/#

Source: https://fred.stlouisfed.org/series/CPIAUCSL#

Core CPI didn't peak until August 2022, a full 12 months later. The pipeline told the story first. If you waited for CPI confirmation, you were late.

Source: https://fred.stlouisfed.org/series/CPILFESL/#

Types of Upstream Indicators

Commodity indices: Broad baskets tracking energy, grains, metals, and softs. When commodity indices are accelerating, input costs are rising across the board. The level matters less than the trend and acceleration.

Shipping and logistics costs: Indices measuring freight rates for bulk cargo, container shipping and air freight. These can’t be stored or speculated on easily, they reflect real-time supply/demand for global trade. A collapsing shipping index tells you goods inflation is softening at the source. A surging one tells you pressure is building.

Business input cost surveys: What businesses are actually paying for inputs. Manufacturing surveys often include a ‘prices paid’ component that captures the pass-through from commodity markets to business costs. Readings above historical averages signal pressure; readings below signal relief.

I weight upstream at about 20% of the inflation pillar. It’s leading, so it gets less weight than the stuff the Fed actually targets. But it’s the early warning system.

One thing worth noting: upstream isn’t just a US story. Global commodity markets and shipping routes mean that China’s slowdown, European energy dynamics and emerging market demand all feed into the same pipeline. Shipping indices don’t care where the cargo is headed, they measure global trade flows. When China’s property sector struggles and construction demand falls, that shows up in copper and iron ore prices that eventually hit US input costs. When European natural gas spikes, it ripples through global energy markets. The upstream indicators are inherently international, which is part of what makes them useful as early warning signals.

Midstream: The Credibility Meter

This is the probably the toughest one to get right because the expectations are self-fulling. If businesses expect higher inflation, they raise prices preemptively. If workers expect higher inflation, they demand higher wages. If consumers expect higher prices, they pull forward spending, increasing demand. The expectation creates the reality.

This is why the Fed talks about expectations so obsessively. They can tolerate above-target inflation if expectations stay anchored, it’s seen as temporary, a blip that will pass. But once expectations change so does the math. Transitory becomes permanent. The Fed has to break and that usually means inflicting economic pain.

Types of Midstream Indicators

Long-term market expectations: Bond market measures of what inflation will average over the next 5-10 years. These are slow-moving but powerful. When long-term expectations stay near 2%, the Fed has credibility. When they drift materially above or below that anchor, something has shifted.

Breakeven inflation rates: The difference between nominal and inflation-protected bond yields. More volatile than long-term forwards but more responsive to near-term outlook changes. Useful for gauging how the market is pricing inflation risk over the next 5-10 years.

Consumer and business surveys: What households and firms expect inflation to be over the next 1-5 years. The Fed watches these because expectations can become self-fulfilling through wage demands and pricing decisions. Short-term measures (1-year ahead) are volatile and noisy; longer-term measures (3-5 years) are more stable and more useful. Right now there’s often a divergence worth noting: 1-year expectations tend to run higher, reflecting recent price memory, while 5-year expectations sit lower suggesting people believe the Fed will ultimately get inflation under control. That gap is actually healthy. It’s when longer-term expectations start drifting toward the short-term read that the Fed gets nervous.

Source: https://www.newyorkfed.org/microeconomics/sce#/inflexp-3 - 3y Expectations

Source: https://www.newyorkfed.org/microeconomics/sce#/inflexp-4 - 5y Expectations

I weight midstream at about 25% of the inflation pillar. It's the credibility check more than anything, as long as expectations stay anchored, the Fed has room to maneuver.

Source: https://fred.stlouisfed.org/series/EXPINF10YR

Downstream: What Actually Matters to the Fed

This is ultimately the end point. The sticky prices consumers pay. The stuff that doesn’t go away on its own. I weight downstream at 55% of the inflation pillar, more than upstream and midstream combined because this is what drives monetary policy. The Fed doesn’t target commodity prices or breakevens directly. They target PCE inflation.

Types of Downstream Indicators

Services inflation ex-shelter: This is arguably the indicator central bankers watch most closely. Strip out volatile goods and housing, which has its own dynamics. What’s left is the most persistent inflation. The stuff that requires actual demand destruction to bring down. When this is falling, the Fed can be patient but when it’s rising, they’re behind the curve.

Core PCE: The Fed’s official target measure. Strips out food and energy to reduce volatility. The gap between current readings and the 2% target tells you how much work remains. The trend tells you whether you’re getting there.

Wage growth trackers: Wages are the ultimate sticky price. Once they go up they only grow more slowly. Look for measures that track median wage growth for continuously employed workers, removing the composition effects that distort averages. Wage growth decelerating is one of the clearest signs that the inflation cycle is turning.

Source: https://fred.stlouisfed.org/series/CUUR0000SASL2RS#

Source: https://www.atlantafed.org/chcs/wage-growth-tracker

Why PCE, Not CPI?

CPI gets all the headlines. Why does the Fed target PCE instead?

Three reasons:

Substitution effect. PCE updates its weights as spending patterns change. If beef gets expensive and people buy chicken instead, PCE adjusts. CPI uses fixed weights, so it overstates inflation when people substitute away from expensive items.

Broader coverage. PCE includes spending on your behalf, employer-paid healthcare, Medicare, Medicaid. CPI only captures out-of-pocket. Since healthcare is a huge and fast-growing spending category, PCE gives a more complete picture.

Housing treatment. Both use “owners’ equivalent rent” to measure housing costs, but they weight it differently. Housing is about 33% of CPI and 15% of PCE. When shelter inflation diverges from other components (like it has recently) this creates meaningful gaps between the two measures.

The practical implication: PCE historically runs 0.3-0.5% below CPI. So when you see a “hot” CPI print, check whether PCE confirms before freaking out. Recently we’ve seen Core CPI at 2.6% while Core PCE sits at 2.7% and the divergence is housing-driven. The Fed watches PCE.

Source: https://fred.stlouisfed.org/series/PCEPILFE#

Source: https://fred.stlouisfed.org/series/CPILFESL#

Mistakes I’ve Made

The growth pillar cost me money through bad timing. Here are five more lessons I learned the hard way. You might recognise at least one from your own trading.

Reacting to CPI Day

This was my biggest inflation mistake for years. Number comes out, market moves, I’d chase and I was usually wrong by the time I’d executed.

The problem is that CPI day is already priced. The professionals trading inflation have been watching the upstream data, the regional Fed surveys, the components that print earlier in the month. By the time the headline drops, the move is often exhausted or the opposite direction from what the number “should” imply because positioning was wrong-footed.

Now I use CPI for confirmation, not initiation. If my pipeline read says disinflation and CPI confirms, I get more confident in the thesis. If CPI contradicts the pipeline read, I figure out why before changing anything.

Ignoring Services

For a long time I focused too much on goods inflation because that’s what moves around. Used cars, electronics, apparel, these components can spike 20% and then collapse. They get headlines. They feel important. Goods are roughly a quarter to a third of the basket. Services are the rest and services inflation is sticky. Wages drive services inflation, and wages don’t fall. They just grow more slowly.

During 2022-23 goods inflation collapsed as supply chains normalized. Used car prices fell 10%+. I thought inflation was solved. But services kept grinding higher. The Fed kept hiking. My trades suffered.

Now services inflation ex-housing is my single most important downstream indicator type. It’s what takes the longest to move. And it’s what tells you whether the inflation problem is actually solved or just masked by volatile components.

Confusing Level With Trend

“Core PCE at 2.7% is too high. Stay defensive.”

I was so focused on the level being above target that I missed the trend, which was clearly down. The Fed saw the trend too.

The regime framework cares about direction as much as level. Core PCE at 2.7% and falling is completely different from Core PCE at 2.7% and rising. The first is progress toward target. The second is re-acceleration. Same number, opposite implications.

Now I always ask: which way is this moving? That single question has improved my inflation reads dramatically.

Forgetting the Lags

Inflation has momentum. Upstream leads downstream by 6-12 months. Policy affects inflation with “long and variable lags” the Fed’s own words. But I’d see commodity prices fall and expect CPI to follow immediately. It doesn’t work that way.

In 2022 shipping costs collapsed. Commodity prices peaked mid-2022. But core inflation didn’t peak until much later, and services kept rising into 2023. The pipeline takes time to clear.

Patience is required. If upstream is falling and downstream is still elevated, that’s not a contradiction. The clearing is coming. Don’t panic out of positions because the headline hasn’t moved yet.

Over-Reacting to Single Prints

One hot CPI doesn’t make Stagflation. One soft print doesn’t mean Goldilocks. Economic data is noisy. Seasonal adjustments are imperfect. There’s genuine randomness in any single reading.

I used to flip my view every month based on whether the number beat or missed. The framework uses 3-month trends for a reason. If your model changes regime on every data release, it’s just a glorified newsfeed.

The short version: Use CPI for confirmation, not initiation. Focus on services, not goods. Track trends, not just levels. Respect the lags. And don’t flip your view on every print.

Applying the Framework

The power of the pipeline model is pattern recognition. Let me walk through how it works in practice, not to tell you what I’m watching specifically, but to show how the framework generates signals.

When the pipeline is clearing: Upstream stable or falling, midstream anchored, downstream decelerating. This is the pattern that precedes disinflationary environments.

When pressure is building: Upstream rising, midstream starting to drift higher, downstream not yet responding. The lead times tell you the downstream move is coming.

When there’s a credibility problem: Midstream de-anchoring even while upstream is stable. This is the Fed’s nightmare, expectations becoming disengaged from fundamentals.

When you need patience: Upstream falling but downstream still sticky. This is normal lag. The clearing is coming, don’t panic out of positions because the headline hasn’t moved yet.

The specific numbers matter less than the pattern. Once you internalise the pipeline structure, you start seeing these combinations everywhere and you start recognising regime shifts before the consensus catches on.

Building Your Own Inflation Framework

Same principle as the growth framework: I’m not going to hand you the exact recipe. The structure matters more than the specific ingredients. Here’s how I’d approach it if I were starting over:

Build the pipeline structure. Separate upstream (leading), midstream (expectations), and downstream (sticky/target). This is the architecture that makes everything else work.

Choose indicators for each stage. You need at least one commodity measure, one shipping/supply chain measure, one market-based expectation measure, one survey-based expectation measure, and two or three downstream measures including something that captures sticky services inflation and wage dynamics.

Weight toward downstream. That’s what the Fed targets and what drives policy. Upstream and midstream get less weight but provide the lead time. I use roughly 20/25/55 but you can adjust based on what works for your style.

Track levels AND trends. For each indicator, you want to know both where it is and which direction it’s moving. A simple 3-month vs 12-month moving average comparison works. Above and rising = accelerating. Below and falling = decelerating.

Set your signposts. What thresholds would trigger a reassessment? Core inflation above a certain level? Long-term expectations drifting higher? Services inflation breaking below target? Write them down before you need them so you’re not moving goalposts when the data gets inconvenient.

Connect to the regime matrix. Inflation alone doesn’t tell you the playbook. You need to combine it with your growth read. Same inflation number means different things depending on whether growth is positive, neutral or negative.

The Bottom Line

Inflation analysis isn’t about predicting the next CPI print. It’s about understanding where pressures are in the system and whether they’re building or clearing.

The pipeline model gives you that edge. Upstream tells you what’s coming. Midstream tells you whether expectations are anchored. Downstream tells you where you actually are. Together, they give you a complete picture and lead time over the reactive approach of just watching the headline number.

Right now, the pipeline is clearing. That’s the framework doing its job and identifying the regime shift before the headlines confirmed it. The Stagflation call from a few weeks ago was real, but the data improved. The inflation pillar moved from +1 to 0. The regime matrix shifted accordingly.

Next time we’ll cover Policy, the third pillar. How to read the Fed beyond the headlines, what real rates actually tell you, why the curve shape matters more than the level, and the distinction between monetary and fiscal that most people miss. Until then, start building your inflation pipeline.

Questions or feedback? macroanalytix.research@gmail.com

Disclaimer:

All content published by MacroAnalytix is for research and educational purposes only. Nothing on this site or in our publications should be interpreted as financial advice, investment recommendations, or a solicitation to buy or sell securities. You are solely responsible for your own financial decisions. Please conduct your own due diligence and consult with a licensed financial professional before making any investment decisions.

I’ve been thinking about how to do this series for a while.

The Setup gives you my weekly regime call. The Breakdown goes deep on themes. The Pulse catches stuff that can’t wait until Sunday. But none of that explains how I actually think about this stuff. the mental models, the frameworks, the hard-won lessons from getting it wrong more times than I’d like to admit.

That’s what The Mastermind is for. Less “here’s the answer,” more “here’s how I think about the question.”

I’m not going to hand you a recipe. Partly because the specific indicators I use aren’t magic, they’re just the ones I’ve found work for me. Partly because if you just copy someone else’s framework without understanding why it works, you’ll abandon it the first time it gives you a signal you don’t like. And partly because, honestly, I think you’ll build something better if you understand the principles and adapt them to your own style.

We’re starting with Growth. It’s the pillar I weight most heavily, and it took me the longest to get right. Let me tell you why.

Why I Obsess Over Growth

Here’s something that took me years to internalize: most of what matters in macro comes down to two questions. Is growth going up or down? Is inflation going up or down? That’s it. Everything else; policy, liquidity, sentiment, is just context for how those two forces play out.

I used to make this way more complicated than it needed to be. I’d track dozens of indicators, build elaborate models, convince myself that some obscure data series held the key to everything. And you know what? Most of it was noise.

The reason is pretty intuitive once you stop and think about it. Corporate earnings are basically a derivative of economic growth. When the economy expands, companies sell more stuff, margins improve, earnings grow. When the economy contracts, the whole thing reverses. And since asset prices ultimately reflect earnings - present and expected - growth is what drives returns.

Same logic works across asset classes. Credit spreads? They’re pricing default risk, which is a function of whether companies can generate enough cash to service their debt. Rates? Growth expectations drive term premium and curve shape. FX? Capital flows to where growth is. Commodities? Demand growth drives prices. It all comes back to the same thing.

I’m not saying you can ignore everything else. But if you get growth right, you’re playing with the wind at your back. Get it wrong, and no amount of cleverness in the other pillars will save you.

The Third Question Nobody Talks About

Okay, so growth up or down, inflation up or down. That’s the fundamental picture. But here’s the thing that separates okay macro analysis from actually making money: you also need to know who’s positioned wrong.

Because markets don’t just move on fundamentals. They move on the unwind of wrong-footed positions. The violence of a move - how fast, how far, how much it overshoots, depends almost entirely on who’s offside and has to capitulate.

Think about it this way. If growth is about to roll over and everyone already expects it - they’re positioned defensively, hedges are in place, cash is high, then when the data confirms, the move might be a yawn. The bad news was priced. But if growth is about to roll over and everyone’s levered long, chasing the last bit of upside, convinced the expansion has years to run? That’s when you get the face-ripping moves. The forced selling. The “how did it drop this much on that data?” moments.

I’ve learned to always ask: who gets fucked if I’m right?

If growth surprises to the upside, who’s short and will be forced to cover? If it surprises to the downside, who’s levered long and will need to liquidate? The answer tells you not just direction but magnitude. It tells you whether you’re trading with the forced flows or against them.

This is why consensus matters, not because the consensus is usually wrong, but because when it is wrong, the move is amplified by everyone scrambling for the same exit. The pain trade isn’t just about being right. It’s about being right when being wrong would hurt the most people.

Some practical things I look at: fund positioning surveys, futures commitment of traders data, ETF flows, option skew. None of these are perfect, but together they paint a picture of where the crowding is. When my fundamental view aligns with a crowded position on the other side, that’s when I get more aggressive. When my view would require fighting entrenched positioning, I stay smaller and more patient.

S&P500 Commitment of Traders Report Jan 6, 2026. https://www.cftc.gov/dea/options/deacmelof.htm

The regime framework gives you the fundamental read. But the money is made and lost in understanding who’s positioned for a different outcome.

The Distinction That Changed Everything

Okay, here’s the thing nobody told me when I started: not all growth data is created equal. There’s a fundamental difference between indicators that tell you where growth is going and indicators that tell you where growth is now. Mixing them up will cost you money. I know because it cost me money.

Leading indicators are forward-looking. They reflect decisions being made today that will show up in economic activity months from now. Orders placed, permits filed, credit extended. These are commitments - bets on the future that become reality with a lag.

Coincident indicators tell you what’s happening right now. Employment, production, income, spending. These confirm the current state of the economy, not where it’s headed.

The mistake I used to make and I see this constantly, is treating these as interchangeable. You see a strong jobs report and think “economy’s great!” But employment is actually a lagging indicator. It often peaks after the economy has already turned. By the time the labor market cracks, you’re already deep into the downturn.

Or you see a weak survey and panic. But surveys can give false signals for months. They measure what people say they’re experiencing, not what they’re actually doing. Sometimes sentiment gets ahead of reality, or behind it.

The solution is to track both categories separately and watch for agreement or divergence. When leading and coincident indicators tell the same story, you can act with conviction. When they disagree, something’s about to change. The leading indicators are usually right about direction; the coincident data helps you calibrate timing.

What Actually Makes a Good Leading Indicator

I’ve wasted a lot of time on indicators that looked good on paper but didn’t actually work. Here’s what I’ve learned to look for:

It has to reflect real commitments, not just intentions. This is huge. Survey data asking firms about their “expectations” is way less reliable than data showing what they’ve actually ordered or contracted for. Talk is cheap. Money on the table is signal. When someone places an order or files a permit, they’re making a commitment that will turn into economic activity. When they tell a poll they’re “optimistic,” that could mean anything.

You need to be able to explain why it leads. If you can’t articulate the transmission mechanism, be suspicious. New orders lead production because firms produce to fill orders, that’s a clear chain. But some indicators have correlations that don’t make logical sense. Those tend to break down exactly when you need them most.

It should have worked across multiple cycles. Anyone can find an indicator that “predicted” the last recession if they torture the data enough. What you want is something with a track record across different environments. And even then, every indicator has given false signals at some point. The goal isn’t perfection, it’s a useful edge.

Take the Conference Board’s Leading Economic Index as an example. It’s one of the most-watched composites, with a track record going back decades. But look at how it’s actually performed:

Three false positives in the last decade alone. That’s not a knock on the LEI, it’s a reminder that no indicator is infallible. The 2016 and 2019 episodes both featured manufacturing weakness that didn’t spread to the broader economy. The 2022-23 signal was the longest false positive on record, confounded by post-COVID services strength. These patterns repeat which is why diversifying across indicator types matters more than finding the “perfect” single measure.

It needs to be timely enough to act on. An indicator that comes out quarterly with a two-month lag isn’t much use for real-time positioning. Monthly data with a short reporting delay is the sweet spot for most purposes.

The specific indicators you choose matter less than whether they meet these criteria. I’ve seen good frameworks built on completely different data. The principle is what matters.

And Coincident Indicators?

Coincident indicators serve a different purpose. They’re not for prediction, they’re for confirmation. You want to know whether the economy is actually doing what the leading indicators suggested it would do.

A few things I look for:

Measure actual activity, not proxies. GDP is the ultimate coincident indicator, but it comes out quarterly with big lags. You want higher-frequency data that moves closely with GDP - actual production volumes, actual employment, actual spending.

Source: https://fred.stlouisfed.org/series/GDPC1#

Source: https://fred.stlouisfed.org/series/USPHCI#

Cover a meaningful share of the economy. This is where a lot of people go wrong. Manufacturing data gets tons of attention but manufacturing is only about 11% of the US economy. Services are closer to 70%. If you’re over-weighting manufacturing, you’re getting a skewed picture. I’ve seen people get bearish because manufacturing was contracting while the overall economy was fine because services held up.

Source: https://www.bea.gov/data/gdp/gdp-industry

Hard to manipulate or distort. The more an indicator relies on actual counts of real things; units produced, people employed, dollars spent, the more I trust it. Some data series have seasonal adjustment quirks or model-based estimates that can introduce noise. The simpler and more direct, the better.

When coincident data confirms what the leading indicators suggested, I act with more conviction. When it contradicts, I get cautious and start looking for what I might be missing.

The Divergence Framework

This is probably the most useful thing I can share with you.

The real value of tracking leading and coincident separately isn’t either score in isolation, it’s the relationship between them. When they agree, you have a clear signal. When they diverge, you have an early warning that the current state is about to change.

Let me show you how this worked in real time over the past couple months:

December 2025: My leading indicators had turned negative. Forward-looking surveys were weak, market-based signals were depressed, the composite measures were flashing warnings. But coincident indicators were still positive. Services expanding, jobs being added, consumer spending holding up. Classic divergence pattern: “caution, slowdown ahead.” That’s exactly what drove my Stagflation call in Issue #004. The leading data was telling me something the headlines weren’t.

January 2026: The leading indicators stabilized. Coincident data held positive. Now both categories were aligned, telling the same story: deceleration without collapse. That’s why the regime shifted from Stagflation to Neutral/Chop. The divergence had resolved.

This framework won’t tell you exactly what’s going to happen. Nothing will. But it tells you how much confidence to have in your current read and what to watch for next. Honestly, that’s often more valuable than a point estimate.

Mistakes I’ve Made (So You Don’t Have To)

I’ve made every mistake you can make with growth data. Here are the ones that cost me the most:

Getting Obsessed With One Number

Early in my career, I’d anchor on whatever indicator had worked recently. ISM was my thing for a while. Then I got burned when ISM was weak but the economy kept trucking along because services were fine.

No single indicator tells the whole story. Ever. The economy is too complex. You need multiple measures, ideally ones that capture different aspects of activity. When they agree, great. When they disagree, that’s information too.

Confusing Level With Direction

This one still trips me up sometimes. A reading of 52 on a diffusion index means “expansion.” Good, right? Well, it depends. If it was 48 last month, then 52 is great: things are improving. If it was 56 last month, then 52 is concerning, things are deteriorating, even though the level is still positive.

Direction often matters more than level, especially at turning points. I try to track both now, but it took me too long to learn this.

Fighting The Trend

This is the one that’s cost me the most money. Growth indicators have momentum. When things are weakening, they tend to keep weakening. When things are strengthening, they tend to keep strengthening. Mean reversion happens eventually, but timing it is brutally hard.

I used to see a weak reading and think “okay, it’s got to bounce from here.” Spoiler: it usually didn’t. Now I try to respect the trend until there’s genuine evidence of a turn, usually from the divergence framework, from extreme readings that historically don’t persist, or from policy changes that alter the trajectory.

Ignoring Positioning

I talked about this earlier, but it’s worth repeating as a mistake I made repeatedly: being right on the fundamentals but wrong on timing because I ignored who was positioned the other way.

I’d see growth weakening, get short, and then watch the market grind higher for weeks because everyone else saw the same thing and was already positioned for it. The trade was crowded. There was no one left to sell. It took the data actually confirming signals repeatedly before the move happened.

Now I try to think about not just whether I’m right, but whether being right will actually move prices. If everyone already agrees with me, the move might already be done.

Over-weighting What Gets Attention

Financial media loves a headline. ISM day, factory orders, industrial production. But manufacturing is 11% of GDP. Services are 70%.

The “two-speed economy” thing we’ve been tracking recently - manufacturing in recession while services hold up - isn’t unusual. It’s happened multiple times in the past decade. If you weight by economic significance rather than media attention, you’d have gotten much better reads.

Treating Surveys Like Hard Data

Surveys measure what people say. Hard data measures what they actually do. These diverge more often than you’d expect.

I remember a period where business sentiment surveys were extremely weak, firms were telling polls they were worried about the economy, planning to cut back, etc. But actual spending, hiring and investment data stayed resilient. If you’d traded the surveys, you’d have gotten killed. The hard data was right.

Surveys are still useful, they’re timely and sometimes genuinely predictive. But when surveys say one thing and hard data says another, I’ve learned to trust the hard data until proven otherwise.

From Growth Reading to Actual Positions

Okay, so you’ve got a growth read. Now what? Here’s roughly how I think about it:

When growth is clearly positive: Lean into risk. Equities over bonds, cyclicals over defensives, small over large (usually), tight credit, commodity demand. Corporate earnings are growing, default risk is low, the wind is at your back. Duration is less attractive because yields tend to rise with growth.

When growth is neutral: Welcome to the chop zone. This is honestly the hardest environment because nothing has a clear edge. I focus on quality over speculation, carry strategies if vol stays low, and being more selective about what I own.

When growth is clearly negative: Get defensive. Bonds over equities, defensives over cyclicals, large over small, expect credit to widen. Duration tends to work as yields fall on growth fears. Cash becomes competitive. Default risk is rising, so credit quality matters a lot.

But remember, always overlay the positioning question. If growth is negative but everyone’s already hiding in bonds and defensives, the pain trade might actually be up. If growth is positive but everyone’s already levered long, the risk is a pullback on the slightest disappointment. The fundamental read tells you the direction; the positioning read tells you the path.

Building Your Own Framework

I’m not going to tell you exactly which indicators to use. Partly because I don’t want to give away every detail of my process. But mostly because I think you’ll build something better if you figure it out yourself.

Here’s how I’d approach it:

Start with the structure. The architecture matters more than the contents. Separate leading from coincident indicators. Define what “positive,” “neutral,” and “negative” mean for each bucket. Build a divergence check between the two categories

Choose indicators that make sense to you. Diversify across data types. Leading: forward commitments where you can explain the causal mechanism. Coincident: actual activity across meaningful parts of the economy. Avoid: Over-weighting any single sector.

Add a positioning overlay. Know who’s on what side of the trade. Surveys, COT data, flows, option markets, pick your poison. The fundamental read without positioning context is incomplete.

Set thresholds explicitly. Write them down before you need them. What constitutes expansion vs contraction? What counts as a meaningful change? Clear thresholds prevent moving the goalposts when data is inconvenient.

Backtest but don’t overfit. Simpler is usually better. Your framework should relate to historical outcomes. Twenty perfectly-tuned indicators will probably fail on the next cycle.

Iterate as you go. No framework is perfect out of the box. Track what works and what doesn’t. When wrong, figure out why; when right, ask if it was skill or luck.

The Bottom Line

Growth is the engine of regime classification because it has the strongest and most consistent relationship with cross-asset returns. But knowing the fundamental direction is only half the battle. You also need to know who’s positioned wrong because that determines whether your correct call makes you money or just makes you early.

The methodology; separating leading from coincident, watching for divergences, overlaying positioning, avoiding the mistakes I’ve walked through here matters more than any specific indicator choice.

Right now, my Growth reading is 0 (Neutral), with leading and coincident aligned in a soft landing configuration. That’s better than December’s -1, when the divergence was flashing warnings. But it’s not an all-clear. The leading indicators stabilized rather than improved, and the margin of safety is thin.

Next time, we’ll cover Inflation. The second axis of the regime matrix and the pillar that determines whether you’re in Goldilocks or Stagflation, Reflation or Deflation. I’ll walk through how I think about the inflation pipeline, why different inflation measures diverge, and how to spot pressure building before it shows up in the headlines.

Until then, start thinking about how you’d structure your own growth framework. What would you track as leading? What as coincident? How would you gauge positioning? Sometimes the exercise of thinking it through is more valuable than any specific answer I could give you.

Questions or feedback? macroanalytix.research@gmail.com

Disclaimer

All content published by MacroAnalytix is for research and educational purposes only. Nothing on this site or in our publications should be interpreted as financial advice, investment recommendations, or a solicitation to buy or sell securities. You are solely responsible for your own financial decisions. Please conduct your own due diligence and consult with a licensed financial professional before making any investment decisions.