The fedex index white paper

The FedEx Index

Ground-Level Delivery Volume as a Real-Time Consumption Signal

Ryan Binkley | March 2026

Abstract

Most economic indicators tell you what already happened. Employment reports are a month old when they publish. Consumer confidence surveys measure how people feel, not what they actually do. GDP figures arrive a quarter late and get revised twice. This paper proposes a simpler, faster alternative: delivery volume at the ground level of a major parcel carrier, observed in real time by the people running the routes.

The FedEx Index is not a formal economic model. It is a ground-truth signal with zero reporting lag, cross-income-bracket coverage, and a direct read on discretionary consumption behavior. When stop counts stay elevated outside of peak season, especially in a high-anxiety economic environment, the most likely explanation is not wage-driven spending. It is debt-financed consumption. Understanding that distinction matters enormously for interpreting what the economy is actually doing versus what the headlines say it is doing.

1. The Problem With Standard Indicators

Economic forecasting has a lag problem. Every major indicator used to assess the health of the U.S. consumer comes with a built-in delay, a sampling methodology, or a sentiment bias that puts it at least one step behind reality.

The Bureau of Labor Statistics releases monthly jobs data approximately three weeks after the reference period ends. Those figures get revised twice. Consumer confidence surveys ask people how they feel about their financial situation, which is a behavioral and psychological measurement, not an economic one. PMI data relies on purchasing managers estimating forward-looking activity. Even retail sales figures, one of the more direct spending measurements available, have a two-to-four week collection and publication lag.

The result is that analysts, investors, and policymakers are perpetually interpreting the economy through a rearview mirror. By the time a data set confirms a trend, that trend is often already reversing.

There is a secondary problem with aggregation. National employment figures are not useful for understanding consumer spending behavior because employment is a production metric, not a consumption metric. A software company laying off 3,000 employees in Seattle and a distribution center opening in Memphis adding 500 jobs appear in the same national jobs report as a net negative of 2,500. Neither tells you whether consumers are buying furniture on credit in Colorado Springs today.

Employment tells you how many people got paid. Delivery volume tells you what they did with it.

1. The FedEx Index: What It Measures and Why It Works

A FedEx Ground route driver in a mid-sized U.S. metro runs between 160 and 180 stops per day during normal operating conditions. Peak season, defined as late November through late December, drives that number higher. January and February historically see a meaningful drop in stop count as post-holiday spending normalizes. March picks up modestly around tax refund activity. April through October represents the baseline, with summer showing a modest uptick.

That seasonal pattern is well-established. The FedEx Index is not the pattern itself. It is the deviation from the pattern.

When stop counts remain above historical baseline during months that should reflect seasonal contraction, and when the composition of those deliveries is weighted toward discretionary goods, that is a signal worth examining. It means consumers are spending at a rate that their income cycle alone does not fully explain.

What Makes This Signal Different

Three characteristics separate ground-level delivery volume from standard economic indicators.

First, it is real-time. A route driver knows today's stop count today. There is no collection delay, no publication schedule, no revision cycle. The signal is same-day.

Second, it is cross-bracket. A single route in a mid-sized city will pass through low-income, middle-income, and higher-income neighborhoods. The aggregate stop count reflects purchasing behavior across the income spectrum, not a sample of a particular demographic.

Third, it measures revealed preference, not stated preference. The box on the doorstep is proof of a completed transaction. The card was not declined. The purchase was made. No survey captures that with the same fidelity.

1. The Debt Thesis

The 2025 delivery environment presented an anomaly. January and February showed modest softening, consistent with post-peak seasonal normalization. But the volume floor was higher than historical baselines suggested it should be, and the goods being delivered skewed heavily discretionary. Subscription services, pet supplies, meal kit fulfillments, large furniture items including sofas and mattresses, and home goods were well-represented in daily stop composition.

This is not the composition of a consumer under financial stress making essential purchases. Essential purchases, pharmaceuticals, staple groceries, household consumables, are largely handled by USPS, same-day local delivery, or retail pickup. What moves through FedEx Ground at elevated volume outside of peak season is what people want, not just what they need.

The reconciling variable for sustained discretionary volume during a period of elevated economic anxiety is revolving credit.

The Data

The Federal Reserve Bank of New York reported total U.S. credit card balances of $1.277 trillion at the end of Q4 2025, the highest figure since tracking began in 1999 and a 66 percent increase from the Q1 2021 pandemic-era trough of $770 billion. Average APRs on cards accruing interest were running at 22.30 percent in Q4 2025.

Buy Now Pay Later adoption accelerated through 2025. Emarketer reported BNPL services facilitated $10.1 billion in online spending in November 2025 alone, a 9 percent year-over-year increase. BNPL is particularly prevalent in large discretionary purchases including furniture and appliances, exactly the category of goods appearing at elevated rates in ground delivery volume.

The St. Louis Federal Reserve documented that delinquency rates in the lowest-income ZIP codes grew by 63 percent in relative terms between Q2 2021 and Q1 2025. Even in the highest-income ZIP codes, delinquency rates grew by 44 percent over the same period. This is not a story confined to subprime borrowers. Stress is cross-bracket, which mirrors what ground-level delivery volume also shows: elevated spending across income levels.

Sustained discretionary delivery volume during a high-anxiety economy is not a sign of financial health. It is the last spending spree running on borrowed time.

The K-Shape Nuance

Equifax data through 2025 identified a growing divergence in consumer behavior sometimes described as K-shaped. Higher-income households are largely cycling credit card balances monthly without carrying interest-accruing debt. Lower-income and subprime households are carrying balances at APRs above 22 percent with rising delinquency rates.

For the purposes of the FedEx Index, this nuance does not change the signal. It clarifies the mechanism. Total delivery volume is being sustained by a combination of genuinely healthy spending from the top of the income distribution and debt-financed spending from the middle and bottom. The aggregate stop count stays elevated either way. But the composition of risk underneath that number is not uniform.

1. Recession Suppression by Revolving Credit

The conventional recession debate asks whether we are in a contraction or not. The FedEx Index suggests that framing misses the more important question: how long can revolving credit and BNPL financing maintain the appearance of consumption health before the structural limits assert themselves?

Three mechanisms could break the current equilibrium, roughly in order of probability.

Credit Supply Contraction

Banks have significant discretion over credit limit adjustments, new account approvals, and balance transfer availability. Tightening does not announce itself. Consumers do not receive press releases. It shows up quietly in declining available credit, rejected applications, and reduced promotional offers. By the time tightened credit supply is visible in aggregate data, the spending behavior change will already be underway. In the FedEx Index, it would appear as a gradual compression in stop counts beginning in the non-peak months.

Minimum Payment Threshold

At $1.28 trillion in outstanding balances and a weighted average APR above 22 percent, the aggregate monthly interest burden on U.S. credit card debt exceeds $23 billion per month. At some point, the minimum payment on existing balances consumes enough discretionary income that new purchases cannot be initiated. The card does not get declined at checkout. The consumer simply stops putting new charges on it. This is a quieter mechanism than an outright credit crisis but ultimately more corrosive because it is invisible to most indicators until spending falls off materially.

Employment Shock

A meaningful spike in unemployment would stress-test all of the above simultaneously. Revolving credit can sustain consumption through periods of reduced income confidence, but it cannot replace a paycheck indefinitely. A layoff wave would accelerate delinquency rates, trigger credit supply contraction, and eliminate the minimum payment capacity described above in rapid sequence. This is the traditional recession trigger, and it remains the most consequential scenario because it compounds the two mechanisms above rather than operating independently.

1. The Threshold Framework

The FedEx Index becomes actionable when a specific threshold is defined for what constitutes a signal versus normal variation.

A single week of compressed stop counts is not a signal. Delivery volume is subject to weather disruptions, local events, carrier network adjustments, and calendar effects that have nothing