Figure 1: Uncertainty indicators over the Business cycle.
Brief description: Four-panel time-series chart of common uncertainty proxies: the BAA-10-year corporate bond spread, the VIX implied volatility index, the Economic Policy Uncertainty Index, and the JLN 3-month uncertainty index. Gray recession bands mark downturns; all four measures generally rise or spike around recessions, with especially visible jumps around the 2008-09 financial crisis and the 2020 downturn.
Figure 2: Uncertainty and Economic Activity. Consumption
corresponds to the year-over-year changes in Personal Consumption
Expenditures (PCE) as recorded by the BEA, while Employment
tracks the year over year changes to the level of total non-farm,
quarterly employment. Capacity Utilization refers to the
percentage of industrial capacity currently being used by firms
domestically to produce the demanded finished products as compiled by
the Board of Governors of the Federal Reserve System. Retail
Sales correspond to the yearly change in the level of retail and
food services sales as measured by the U.S. Census Bureau, and Credit
Conditions refer to the Federal Reserve Bank of Chicago’s National
Financial Conditions Index (NFCI), where positive values of the index
indicate that financial conditions are tighter than average. Finally,
Firms’ Net Worth tracks the evolution of the non-financial
corporate business sector’s net worth as a percentage of
GDP.
Brief description: Six time-series panels compare credit conditions, capacity utilization, retail sales, private consumption, firms' net worth, and employment over the business cycle. Recession bands coincide with tighter credit conditions and weaker real activity, including lower capacity utilization, retail sales, consumption, employment, and corporate net worth around major downturns.
Figure 3: Uncertainty over the business cycle
Brief description: Two time-series panels plot cross-sectional dispersion in sales growth and employment growth, with recessions shaded. Sales-growth dispersion is volatile and tends to rise or remain elevated around several downturns, while employment-growth dispersion shows a more mixed and weaker cyclical pattern.
Figure 4: Model’s Timing
Brief description: Timeline from period t to t+1 divided into three subperiods. The aggregate credit shock theta_t is realized at the start; entrepreneurs enter with assets b_t^i, choose labor h_t^i, and wages are set before idiosyncratic matching n_t^i is realized; output y_t^i is then realized, consumption c_t is chosen, wages are paid, balances are settled, and next-period assets b_{t+1}^i are selected.
Figure 5: Distribution of sales per worker
Brief description: Schematic density chart for sales per worker y_t^i. A red distribution is shifted to the right relative to a blue distribution with similar spread, illustrating that better aggregate credit conditions raise expected sales per worker as a first-moment effect without changing dispersion.
Figure 6: Distribution of customers per worker
Brief description: Schematic density chart for customers per worker n_t^i with a vertical cutoff at n_bar(theta_t). The shaded area shows the part of the customer distribution that can be served before the capacity-related censoring point, emphasizing that per-worker demand is bounded by the credit-conditioned customer threshold.
Figure 7: \(\theta_t^2 >
\theta_t^1\)
Brief description: Schematic density chart for customers per worker showing two censoring thresholds, $$n_bar(theta_t^2)$$ to the left of $$n_bar(theta_t^1)$$. Because $$theta_t^2$$ is greater than $$theta_t^1$$, the cutoff moves left and the shaded probability mass between the two thresholds is reallocated to the new censoring point, reducing idiosyncratic uncertainty per worker.
Figure 8: Endogenous Uncertainty
Brief description: Schematic density chart for sales per worker with two distributions and a capacity ceiling y_bar. The higher-credit distribution shifts sales toward the capacity limit while the shaded area near the cutoff represents compression of dispersion, combining a higher expected sales level with lower demand uncertainty.
Figure 9: Distribution of customers per firm
Brief description: Schematic density chart comparing customer distributions for firms of different employment sizes. Hiring more workers shifts the firm-level customer distribution toward a higher mean and wider spread, increasing expected sales but also increasing the entrepreneur's exposure to demand risk and a larger wage bill.
Figure 10: Profit distribution simulations
Brief description: Density plot of simulated profit distributions for firms with h = 2, h = 4, and h = 6 workers. As employment rises, the distribution becomes wider and flatter, with more mass in low-profit or loss outcomes as well as more potential high-profit outcomes, illustrating the risk-return trade-off from scaling up employment.
Figure 11: Impulse responses for a 1% shock to \(\theta_t\)
Brief description: Four impulse-response panels show aggregate output, employment, sales per worker, and the real wage after a positive one percent aggregate credit shock. All four variables increase on impact and then gradually decay toward zero over about 40 periods; output and employment show the largest responses, while sales per worker and the real wage rise more modestly.
Figure 12: Impulse responses for a 1% shock to \(\theta_t\)
Brief description: Four impulse-response panels show demand uncertainty, average capacity utilization, aggregate assets, and the interest rate after the same positive credit shock. Demand uncertainty and aggregate assets fall on impact and return gradually toward zero, while capacity utilization and the interest rate rise on impact before fading.
Figure 13: Effect decomposition for a 1% shock to aggregate
credit
Brief description: Four panels compare baseline impulse responses with a level-only counterfactual for aggregate output, employment, sales per worker, and the real wage. Baseline output, employment, and wages respond more strongly than in the level-only economy, while sales per worker is nearly the same in both cases, showing that endogenous uncertainty amplifies activity mainly through hiring and wage effects.
Figure 14: Effect decomposition for a 1% shock to aggregate
credit
Brief description: Four panels compare baseline and level-only responses for aggregate assets, the interest rate, entrepreneur consumption, and household consumption. In the baseline economy, assets and entrepreneur consumption fall while the interest rate and household consumption rise; in the level-only economy, assets and entrepreneur consumption rise slightly and the interest-rate response is negative, highlighting the financial-market role of endogenous uncertainty.
Figure 15: Effects of a 1% shock to \(\theta_t\) under alternative capacity
utilization levels.
Brief description: Six impulse-response panels compare the baseline economy with a counterfactual economy that begins with lower capacity use. The counterfactual shows larger positive responses of output, employment, sales per worker, real wages, and capacity utilization, while the decline in demand uncertainty is smaller than in the baseline; this indicates that unused capacity strengthens the level response but weakens the uncertainty-compression channel.
Figure 16: Disagreement amongst professional forecasters. The figure
above plots the cross-sectional dispersion in private sector forecasts
over the business cycle. The data comes from the Federal Reserve Bank of
Philadelphia’s survey of professional forecasters from 1968Q4 - 2014Q3
for the first four variables and 1981Q3 - 2014Q3 for the remaining two.
Beginning from top left we have the forecasts for Real GDP, the
Price Deflator, Industrial Production, the
Unemployment rate, Real Consumption and
Non-residential fixed investment. In times of higher
uncertainty forecasts become less precise and dispersion amongst
predictions increases. Not surprisingly, recessions tend to be periods
of greatest disagreement amongst forecasters.
Brief description: Six time-series panels plot cross-sectional dispersion in private-sector forecasts for real GDP, inflation, industrial production, unemployment, real consumption, and non-residential fixed investment, with recessions shaded. Forecast disagreement is generally elevated around downturns and periods of macroeconomic stress, with prominent spikes in several series during early-1980s recessions and around the 2008-09 recession.
Figure 17: Net Financial Assets in the nonfinancial business sector as
a percentage of total nonfinancial assets (1980-2025).
Source: Federal Reserve Z.1 Financial
Accounts.
Brief description: Line chart comparing corporate and noncorporate net financial assets as a share of total nonfinancial assets. The corporate sector moves from negative values in the 1980s and early 1990s to positive net lending around the 2000s, falls sharply around 2008-09, and rises again by 2025; the noncorporate sector remains negative throughout, with a deep trough around the financial crisis.
Figure 18: Consumer Traffic and Business Cycle
Source: Own calculations based on ICSC
data
Brief description: Monthly ICSC customer-traffic diffusion index from 2000 to 2014, with recession periods shaded and the index centered around 50. The series drops below 50 during the 2001 and 2007-09 recessions, reaches its lowest point during the 2008-09 downturn, and then recovers above 50 as economic conditions improve. Table Descriptions Table 1: U.S. Compustat Moments 1980-2024 Brief description: Summary statistics for log sales and log employment in the Compustat sample. The table compares dispersion, skewness, kurtosis, correlations of higher moments with the business cycle, and observation counts; it shows sizable cross-sectional dispersion in both variables, mild negative skewness, and a negative business-cycle correlation for sales-growth dispersion. Table 2: Calibration Values Brief description: List of model parameters, their economic descriptions, calibrated values, and empirical targets or sources. The table reports externally chosen and jointly calibrated parameters, including the discount factor, labor disutility, borrowing limit, inverse Frisch elasticity, matching-function parameters, credit-shock process, and maximum output per worker. Table 3: ShopperTrak Data Moments Brief description: Summary statistics for the proprietary customer-traffic data. The table reports cross-sectional dispersion, skewness, kurtosis, the positive correlation of traffic growth with the business cycle, average visits per store and per year, total visits, ZIP-level units, and annual observations. Table 4: Targeted Moments Brief description: Comparison of data targets and model-implied values for the steady-state interest rate, hours worked, and unsecured debt-to-income ratio. The model closely matches all three calibration targets. Table 5: Non-targeted Moments Brief description: Comparison of data and model values for non-targeted cross-sectional moments of employment and sales. The model reproduces the negative cyclical correlation of sales-growth dispersion and the relative ordering of sales and employment dispersion, but overstates some dispersion levels and differs in skewness. Table 6: Model's sensitivity to tau Brief description: Sensitivity table showing how calibrated parameters and impact responses change as the labor-supply parameter tau varies. Lower tau values produce larger employment responses to a positive credit shock, while the wage response remains relatively similar across calibrations. Table 7: Validation of Steady-State and Ergodic Moments Brief description: Appendix validation table comparing steady-state moments with ergodic means from the baseline stochastic economy. Output, employment, wages, assets, debt-income ratios, sales per worker, capacity utilization, and cross-sectional moments are nearly identical across the two columns, supporting the use of steady-state targets in the calibration.