MonteFI

Core Datasets

MonteFI utilizes derivations of the Census American Community Survey (ACS) to provide a view of economic conditions.

1. Average Housing & Income

This dataset aggregates total values reported by a state and divides them by the total count of households, renters, or owners. This is useful for understanding the broad economic volume but can be skewed by extreme outliers (e.g., billionaires or luxury mega-mansions).

• Source Keys: B19025_001E (Agg Income), B25065_001E (Agg Rent), B25079_001E (Agg Home Value), B19001_001E (Total Count: Households), B25063_002E (Total Count: Renters), B25075_001E (Total Count: Home Owners)
• Calculation: Total $ / Total Count
• API Link: Average Housing & Income (2023 Example)

2. Median Housing & Income

This dataset represents the "middle" value of the population. 50% of the population earns/pays more, and 50% earns/pays less. This is generally considered a more accurate reflection of the "typical" experience as it ignores outliers.

• Source Keys: B19013_001E (Income), B25064_001E (Rent), B25077_001E (Home Value)
• Calculation: Direct median estimate from Census Bureau.
• API Link: Median Housing & Income (2023 Example)

3. Monte Carlo Simulations

This dataset projects future economic conditions (2024-2123) based on historical trends (2009-2023). It uses a Geometric Brownian Motion (GBM) model to simulate 10,000 possible future paths for each state and metric, helping to quantify uncertainty and potential risk.

• Source Keys: Generated via Python `pandas` and `numpy` simulations.
• Calculation: Geometric Broiwnian Motion (GBM) (see methodology below).

Derived Metrics & Methodology

To measure affordability and project future trends, we perform custom calculations. These metrics help us understand the relationship between earnings, housing costs, and potential future volatility.

Income to Rent Ratio

This metric answers: "How much income is available to cover rent?"

Interpretation: A higher number is better. A ratio of 40+ implies rent is a small fraction of income. A ratio below 36 might indicate rent burden (i.e. 36 represents 1/3rd of income).

Income to Home Value Ratio

This metric answers: "How much income is available to purchase a home?"

Interpretation: A higher number suggests homes are more affordable relative to income. FYI, this dataset does not take into account property taxes.

Average/Median Ratio (Skewness)

This metric answers: How do average and median values align? It acts as a proxy for the skewness of the economic distribution.

Interpretation:

• Ratio > 1 (Right-Skewed): The Average is pulled "up" by high outliers (e.g., billionaires). The bulk of data is on the lower end.
• Ratio < 1 (Left-Skewed): The Average is dragged "down" by low outliers. The bulk of data is on the higher end.
• Ratio ≈ 1 (Symmetric): The distribution is balanced. Average and Median are roughly equal.

Monte Carlo Projection Model

To forecast future values, we employ a Geometric Brownian Motion (GBM) model. This stochastic process assumes that the logarithm of the underlying metric follows a Brownian motion with drift and volatility.

We calculate two key parameters from historical data (2009-2023):

• Drift (μ): The average annual growth rate (log returns).
• Volatility (σ): The standard deviation of the annual growth rate.

The simulation for year t is calculated as:

• S_t: Projected value at year t.
• S_t-1: Value at the previous year.
• μ: Drift parameter.
• σ: Volatility parameter.
• Z_t: A random shock from a standard normal distribution, Z_t ~ N(0, 1).

We run 10,000 simulations for each state and extract three representative paths: High (Maximum Return), Average (Mean of all paths), and Low (Minimum Return).