How CarCast Works: The Technology Behind Our Predictions
Used car price forecasting is a hard problem. Here's exactly how we solve it.
The Data
Every week, CarCast ingests pricing data from over 200 vehicle segments covering 84,000+ dealer listings nationwide (via MarketCheck). We track weekly median prices, listing volumes, and days-on-market trends for each Year/Make/Model combination.
We also incorporate macroeconomic signals: BLS CPI data for used vehicles, gasoline prices, and overall inflation -- factors that historically correlate with used car demand shifts.
The Model: Our Proprietary Forecasting Engine
At the core of CarCast is a proprietary AI forecasting stack built around a 200-million-parameter foundation model for time-series forecasting.
Unlike traditional statistical models (ARIMA, Prophet) which require per-segment fitting, AI forecasting is a zero-shot forecaster -- it can generate accurate forecasts on vehicle segments it has never explicitly trained on, by learning universal temporal patterns from billions of time-series data points across domains.
How We Generate Forecasts
- Data preparation: We take the full weekly price history for a segment (typically 100-200+ weeks) and apply a log transform to stabilize variance.
- Inference: The transformed series is fed to AI forecasting running on dedicated GPU infrastructure via cloud GPU infrastructure. The model generates point forecasts and quantile predictions.
- Post-processing: We extract P10, P50 (median), and P90 quantiles to create confidence bands, then reverse the log transform to get dollar values.
- Trend classification: If the predicted 30-day price change exceeds +1.5%, we classify the segment as Rising. Below -1.5%, Softening. Everything in between is Stable. These classifications are informational analytics, not financial advice or a recommendation to purchase any vehicle.
- Confidence scoring: We combine the prediction band width, historical volatility, and context length to produce a 0-100% confidence score.
Accuracy Tracking
We score every forecast against the real market and publish the result on our /accuracy page the moment it is old enough to check — no cherry-picking. We are early in building that track record: as of September 2026 only a small, exotic-skewed set of forecasts has matured on the 30-day horizon, so we do not yet publish a blended error or directional figure across all segments. Mainstream and longer-horizon results begin maturing in late October 2026. Every forecast also ships with a P10/P50/P90 confidence band so you can read the uncertainty directly, and the accuracy page always states the sample size and scope behind any number it shows.
Limitations
No model is perfect. CarCast forecasts are probabilistic estimates, not guarantees. They work best for:
- Segments with 50+ weeks of history
- Normal market conditions (not black swan events)
- 30-day horizons (accuracy decreases with longer horizons)
See the forecasts in action. Try CarCast free