Vehicle Demand Forecasting & Inventory Risk Report

Misaligning production schedules with market realities introduces massive financial exposure. Under-stocking leaves millions in unrealized revenue on the table, while over-stocking traps working capital in depreciating assets and compounding holding costs. By anchoring internal supply targets to external econometric demand trends, this model directly serves the Telemetry pillar—ensuring corporate planning coordinates smoothly with macroeconomic shifts rather than fighting them.

Time-Series Architecture & Stationarity Testing

To construct a reliable forecasting framework

the historical baseline (2010–2024) was tested for statistical equilibrium.

• The Stationarity Hurdle: Running the Augmented Dickey-Fuller (ADF) test yielded a $p$-value of 0.457

confirming the presence of a unit root (the data is non-stationary). Unaltered

this would cause standard regression metrics to break down.

• The Mathematical Remedy: Applying one order of differencing ($d=1$) successfully stabilized the drifting mean.

• Model Selection Optimization: While an initial complex ARIMA(1

1

1. architecture was evaluated

it exhibited parameter instability. The system was ultimately streamlined to an **ARIMA(0

1

0)** model—mathematically equivalent to a Random Walk with Drift. This elegant reduction achieved a superior

stable Akaike Information Criterion (AIC = 419.516).

Inventory Risk Analysis (2025–2027)

A. Baseline Demand vs. Corporate Inventory Gap

The predictive model calculates a consistent demand baseline using historical drift dynamics applied to the final 2024 data ceiling.

| Horizon | Forecasted Sales | Planned Inventory Target | Projected Gap | Risk Profile |

| --- | --- | --- | --- | --- |

| 2025–2027 | 17

527

854 units | 17

200

000 units | **+327

854 units** | Persistent Shortage Risk |

Strategic Directive: The baseline forecast shows that management’s current inventory ceiling is systematically under-budgeted by an average of **327

854 units**. Relying solely on this fixed target guarantees unfulfilled customer demand.

B. Volatility Bounding & Extreme Risk Exposure

By generating a 95% Confidence Interval (CI) around the timeline

we mapped the operational outer limits driven by underlying market variance.

| Market Scenario (2027 Target Horizon) | Forecast Boundary | Realized Gap (vs. 17.2M Target) | Operational Risk Exposure |

| --- | --- | --- | --- |

| Upper Confidence Bound | $\approx$ 19.89 Million Units | +2.70 Million Units | Extreme Supply Shortage: Drastic stockouts

brand erosion

competitor poaching. |

| Lower Confidence Bound | $\approx$ 15.15 Million Units | -2.04 Million Units | Major Asset Surplus: Critical capital lockup

soaring lot holding costs

aggressive discounting required. |

Core Visualization Summary

The generated predictive output condenses these statistical boundaries into a clear operational map:

• Historical Baseline (Blue Line): Illustrates the organic market behavior and macro-driven volatility across the 2010–2024 timeline.

• Planned Corporate Target (Red Line): Sits fixed at 17

200

000 units

operating completely independent of market feedback loops.

• The Forecast (Orange Dashed Line): Hovers continuously above management's planned threshold

visually highlighting an intrinsic

ongoing supply deficit.

• The Shaded Uncertainty Band (95% CI): Flares out rapidly as time progresses. This widening cone mathematically visualizes the compounding risk of external

unmodeled macroeconomic factors (e.g.

economic cycles

consumer sentiment shifts)

proving that static planning exposes the supply chain to major systemic shocks.