A residual value curve plots how a vehicle's market value changes as it ages, typically expressed as a percentage of a reference value at each age milestone. For lenders, the most practically useful version is a recovery-at-auction curve: at each time-on-lot stage, what percentage of the unit's stated value would it realistically recover if sold at regional auction today?
Not all vehicles have the same residual value curve shape. A high-demand model with a deep secondary market might retain 88% of its stated value through 45 days on lot and decline to 79% by day 90. A niche import with a shallow buyer base might decline from 78% at day 45 to 62% by day 90. The shape of the curve, specifically its slope and the depth of its baseline, has direct implications for how much credit exposure a lender is taking on, and how quickly that exposure changes as inventory ages.
Why Curve Shape Matters More Than the Starting Point
Floor-plan credit is typically sized based on a percentage of stated value at the time the credit is established. The advance rate is set against current market value, and the assumption embedded in that advance rate is that the collateral will retain enough value through the expected holding period to cover the credit if recovery is needed.
That assumption holds well when the residual value curve is relatively flat, meaning the unit retains most of its value even as it ages. It holds less well when the curve drops sharply, because a unit that was well-collateralized at day 20 may have burned through the lender's cushion by day 60 if the curve descends steeply in that interval.
The starting point and the advance rate together define the buffer the lender has at origination. The slope of the curve defines how quickly that buffer is consumed as time passes without the unit selling. A lender who knows the curve shape for each model category can calculate precisely when a given unit is expected to reach the point where the outstanding advance exceeds the realistic recovery value. This is not a vague risk assessment; it is a calculable timeline based on observable data.
Building Residual Value Curves from Auction Data
Residual value curves for the Philippine used-car market are not published by any central authority. They need to be derived from actual auction transaction data: what units of a given model actually sold for at regional auction, at what age they sold, and in what condition. Accumulating enough transaction data to produce reliable curves requires consistent data collection over time, segmented by model, mileage band, and geographic market.
The practical data pipeline for this is: collect auction transaction records for each unit sold, record the time-on-lot at sale and the sale price, compute the sale price as a percentage of the acquisition price or a reference market value, and aggregate across comparable units to produce a curve for each model category. The curve is the average trajectory, with confidence intervals that widen as the sample size decreases.
Some model categories have abundant transaction data because units of that type sell frequently at auction. Toyota Hilux pickups and Mitsubishi Montero Sports in the three-to-five-year-old age range, for example, are high-volume at Philippine used-car auctions, and a curve built from their transaction history will be stable and well-supported. Other models, particularly European imports and lower-volume Japanese brands, sell less frequently and have less reliable curves due to smaller sample sizes. Lenders should apply wider confidence intervals to recovery estimates for low-volume models.
Applying Curves to Portfolio Monitoring
Once residual value curves are available by model category, they can be applied to a dealer's portfolio in real time to produce a collateral quality assessment that updates with the passage of time, not just at audit events.
The calculation is straightforward. For each unit in the dealer's portfolio, identify the model category, the time already on lot, and the outstanding advance against the unit. Apply the residual value curve for that model category at the current time-on-lot to estimate the realistic recovery value today. Compare the recovery estimate to the outstanding advance to compute the current coverage ratio. Aggregate across all units in the portfolio to produce a portfolio-level coverage ratio.
A portfolio with a weighted average coverage ratio of 1.22 (meaning realistic recovery value is 22% above outstanding advances) is in a different position from one with a ratio of 0.94. A ratio below 1.0 means the portfolio's realistic liquidation value has fallen below the outstanding credit balance, which is an effective negative equity position for the lender. Identifying which specific units and model categories are dragging the coverage ratio below safe thresholds allows the lender to address the concentrated risk rather than the portfolio average.
Portfolio Composition as a Risk Management Tool
Residual value curve analysis is not only useful at the individual dealer level. Applied across a lender's full dealer portfolio, it surfaces patterns in how model mix composition affects aggregate exposure.
A lender whose floor-plan portfolio is heavily weighted toward models with steep residual value curves, meaning models that lose value quickly, is in a different risk position than one whose portfolio is concentrated in flat-curve, high-recovery models. This concentration is not visible in aggregate statistics about outstanding balances and stated values. It requires knowing the curve shapes of the models in the portfolio and calculating the aggregate exposure to steep-curve depreciation.
This analysis also helps lenders understand how correlated their collateral risk is. Two dealers who both hold large positions in the same low-recovery model category are correlated risks: a market development that affects that model category affects both simultaneously. A portfolio spread across model categories with different curve shapes and different buyer bases is less correlated, meaning market developments that hurt one segment are less likely to hit the whole portfolio at the same time.
Limitations of Residual Value Models
Residual value curves are built from historical transaction data and reflect how the market has priced aged vehicles in the past. They are useful proxies for the future, but they carry limitations that lenders should hold in mind.
First, they are averages over conditions that will not always hold. A model that has consistently shown strong residual value retention may see its curve shift if a new model launch, a regulatory change affecting that vehicle type, or a broader fuel economy shift reduces demand. Market-cycle sensitivity varies by model, and a curve that was accurate during stable conditions may understate the depreciation risk during a demand downturn for that category.
Second, curves are built from completed transactions, meaning they capture what sold, not what did not sell. A unit that was taken off the lot without auction sale because the owner determined the recovery would be too low does not appear in the data. This creates some upward bias in recovery estimates, because the transactions that generate data are selectively the ones where recovery was deemed worthwhile.
Neither limitation means residual value curves are not useful. They are more useful than applying a flat haircut to all inventory regardless of model category, which is the implicit assumption in most standard advance rate policies. A lender who accounts for curve shape differences across model categories is building a more accurate picture of their collateral position than one who treats all inventory equivalently at a given stated value. The curves improve the estimate; they do not make it perfect.