The Core Objective

Our forecasting engine is built to answer one question: "What should I realistically expect from this property over the next few years?"

To answer that with confidence, we combine property-level data with broader market intelligence — producing forecasts that are grounded in reality and usable for real investment and asset management decisions.

The Problem with Traditional Forecasts

Many rent forecasts depend on outdated or sampled data. Quarterly surveys are the industry standard, but because data is collected only every few months, these reports inevitably create large gaps between what the data shows and what's actually happening in the market.

When rental demand changes quickly — or supply enters the market faster than expected — lagging data means forecasts miss turning points entirely. By the time the signal appears, operators are scrambling to react rather than planning ahead.

ApartmentIQ is built to solve that.

How Our Forecasting Model Works

Data Foundation

Accuracy starts with better data. ApartmentIQ monitors daily unit-level pricing across over 40 million units, backed by five years of historical data. A daily cadence matters — instead of sampling the market every few months, we track how the industry behaves in real time.

Three Core Inputs

Our model combines three types of inputs to explain both demand and supply pressures — the primary drivers of rent and occupancy outcomes:

1. Historical Performance

• Long-term rent growth history
• Occupancy trends

2. Macroeconomic Conditions

• Employment and wage trends
• Consumer Price Index (CPI) and inflation
• Unemployment rates
• Income levels
• Retail spending patterns
• Population shifts and demographic data

3. Supply-Side Dynamics

• Development pipelines with delivery timelines
• Building permit activity

How Often Forecasts Update

The model updates on a monthly basis, factoring in labor markets, construction supply, inflation, and demographics — rather than waiting for the end of a quarter. This means if market conditions shift, forecasts adjust accordingly without waiting months to reflect the change.

Two Forecasting Approaches

Depending on how much data is available for a given property, ApartmentIQ uses one of two complementary methods:

1. Property-Specific Forecasts

When a property has sufficient high-quality historical data and a critical mass of similar comparable assets, we generate a standalone forecast specific to that property.

• Inputs: The asset's historical performance, daily unit-level rents, local demographics, and new supply pipelines
• Output: A forward-looking view reflecting actual asset performance, including a specific year-over-year (YoY) rent growth metric displayed on the Property Detail Page

This approach delivers a transparent way to compare future performance, offering higher confidence for underwriting, renewal, and pricing decisions.

2. Submarket-Anchored Forecasts, Calibrated for the Asset

For properties that lack the historical depth to support a standalone model, we anchor the forecast to the submarket and then specifically tailor it to the subject property.

• Inputs: Expected market performance, calibrated using the property's daily unit rents, surrounding demographics, and local supply dynamics
• Output: A submarket-informed forecast tuned to the specific property — never a generic submarket line

This provides a credible, market-grounded forecast that ensures consistency across asset types and gives a clear view for evaluating new or recently stabilized properties.

Note: As ApartmentIQ integrates new data over time, properties continuously "graduate" from submarket-anchored forecasting to property-specific forecasting.

Understanding Forecast Scenarios

Baseline, Upside, and Downside

Forecasts include three scenario ranges based on the model's probabilistic outputs, reflecting the range of outcomes the model considers plausible:

In practical terms:

• The upside scenario represents a reasonable expansionary case — for example, stronger employment growth, steady demand, or less supply pressure than anticipated.
• The downside scenario reflects a moderate softening environment — such as slower demand growth, slightly elevated vacancy, or higher-than-expected supply.

These scenarios are intended to capture a realistic range of outcomes useful for planning, underwriting, and risk assessment under typical market uncertainty.

Why Forecasts May Look Optimistic or Pessimistic

The forecast reflects the underlying economic and supply signals in the data — not a fixed assumption. For example:

• More optimistic: Strong employment growth or limited new supply
• More pessimistic: Rising supply or weakening economic indicators

Because the model is forward-looking, it may sometimes diverge from recent trends if leading indicators suggest a change in direction.

Forecast Accuracy

Performance Benchmarks

On a one-month horizon, forecasts average within 0.3 percentage points of actual year-over-year rent growth. Over a twelve-month horizon, they are typically within 0.8 percentage points.

How We Compare to Industry Benchmarks

When compared with other publicly available 2025 rent growth forecasts — which average roughly 2.2% mean absolute error — ApartmentIQ's Explore Pro forecasts deliver materially lower error by approximately 0.7%. Given that total market movement is typically only 3–4%, that improvement represents a 33% gain in forecast accuracy — a significant competitive advantage for portfolio decision-making.

How the Model Handles Major Market Events

We intentionally include periods like COVID-19 and past economic downturns in our training data. The model learns how rent and occupancy responded under those conditions and links those outcomes to macroeconomic drivers — allowing it to better anticipate how similar shocks may impact the market in the future, rather than treating those periods as noise.

Forecast Geography and Granularity

How Localized Are Forecasts?

The model produces forecasts at multiple geographic levels:

• Metro
• Submarket
• ZIP code
• Property

It combines local historical data with broader economic signals, allowing it to reflect both micro-level performance and macro-level trends.

Why Forecasts Sometimes Disappear When Filtering

This comes down to the level at which each forecast is generated:

• Rent growth and occupancy are forecast at the individual property level, so those projections remain visible regardless of how the view is filtered.
• Absorption and deliveries are forecast at the submarket- and metro-level. When a filter narrows the view below a submarket, the app removes those forecast lines because property-level forecasts for these metrics don't currently exist.

Absorption, Deliveries & Occupancy: What to Watch For

When reading the absorption, deliveries, and occupancy chart, understanding the relationship between these three metrics tells the story of market health.

The Ideal Scenario

Positive net absorption that meets or exceeds deliveries, paired with stable or rising occupancy. This means demand is keeping pace with new supply, units are being leased up, and the market is healthy for owners and investors.

Key Signals

The key narrative is the relationship between the three lines: absorption tracking at or above deliveries with stable occupancy signals a resilient market, while a widening spread between deliveries and absorption with falling occupancy signals supply pressure or demand strain.

For more information: 

1. A deeper dive into the Forecasting Methodology from our Head of Forecasting 

2. ApartmentIQ Forecasting Blog Post

3. FAQ: Explore & Research  

4. Contact our Support Team