How to track used car prices: from price indices to portfolio risk
How to track used car prices: asking vs achieved prices, price indices, adjusting for supply mix, leading indicators and portfolio risk.

In this article
Key findings
- A listing shows what the seller is asking, not what the buyer paid. Track asking prices, achieved prices and estimated market values as three separate series.
- The average price can change even when no car has become more expensive. In an illustrative example, it rises by €2,000 simply because the share of younger cars on offer increases from 40% to 60%. A fixed-weight index removes this effect.
- Growing supply and longer listing times signal price pressure earlier than a price index. The real warning sign is listing times lengthening in your own group of cars.
- Compare countries using the same group of cars in euros. If the indices of two countries both stand at 100, they only share the same starting point, not the same price level.
- Portfolio risk arises when falling prices, large exposure and an imminent sale date coincide. Scenarios should therefore be calculated separately for groups of cars and months of lease return, not with a single percentage for the whole portfolio.
Used car prices drive dealers' buying limits, proceeds from returned lease cars and the collateral value of car loans. Most companies track them through average prices from classifieds portals or through published price indices. Both figures, however, are easily distorted by a change in the supply mix, i.e. a different ratio of young to older cars, diesel to electric, or a different country mix. A single figure for the whole market can also hide a fall in precisely the part of supply a company holds. A usable overview therefore links prices with the mix of cars, the volume of supply and the speed of sale.
Which price are you actually tracking?
A listing captures the seller's asking price at one point in time. The final transaction price may reflect a discount, a part-exchange or an extended warranty that you cannot see in the listing. Statistics based on listings therefore describe supply. They say something about the amounts actually paid only indirectly.
| Measure | What it represents | What it is useful for |
|---|---|---|
| Asking price | The seller's current asking price | Monitoring competitors, price changes and supply pressure |
| Achieved retail price | The amount the car was sold to the customer for | Assessing your own sales and the gap against the advertised price |
| Achieved wholesale price | The result of a trade deal, for example at a B2B auction | Monitoring the buying market and remarketing |
| Estimated market value | The result of a valuation based on comparable cars | Valuing an individual car or a portfolio |
Keep wholesale and retail series separate. Between buying and selling to a customer lie reconditioning, the dealer's costs and their margin. The difference between the two prices therefore does not show the market rising or falling.
Sales estimates based on listings must be read with equal caution. A listing disappearing does not mean the car was sold. The seller may have withdrawn it, moved it to another channel or re-listed it under a new ID. The number of removed listings is therefore only an estimate of sales. The last asking price is only an estimate of the price at which the car sold.
For your own deals, it pays to record the first asking price, the last asking price and the achieved price. External data then shows the situation on the market, and internal results show how well the company is selling in it. For the external market, Market Data from Carsdata brings supply, prices, time on market and estimated sold vehicles together in one overview. It estimates sales from how listings develop, so you see prices alongside supply movements. The estimate does not pretend to be a record of actual sales.
How should you read a used car price index?
An index converts price developments to a common scale. The base period is given a value of 100. A value of 95 then means a price 5% lower than in that period. An index says nothing about the amount in euros or whether a car is more expensive in one market than in another.
Different calculations are hidden behind the label "used car price index":
| Calculation method | What it captures | Weak point |
|---|---|---|
| Change in the average or median | Prices of cars currently on offer | Any change in the supply mix shifts the result |
| Fixed-weight index | Price change with the same group composition | Age or mileage can still change within groups |
| Quality-adjusted index | Price change after accounting for vehicle characteristics | The result depends on the characteristics chosen and data quality |
A quality-adjusted index is calculated with a hedonic model. It breaks down a car's price into the effect of age, mileage, powertrain, equipment and other characteristics. What remains is the change in the price level itself.
The period you compare the index with also affects the result. A month-on-month change shows a turning point sooner but fluctuates more. A year-on-year change smooths out seasonal swings, but it can look favourable even when prices have already been falling for several months. For operational decisions, read both values side by side.
A market-wide index will not tell you:
- the price of a specific car or the development in the group you hold in stock or in your portfolio;
- achieved prices, if the index is based on listings;
- whether a fall will continue, because the index describes the past.
Derive buying limits or portfolio values from the relevant group of cars. Do not recalculate all items by the same percentage from a market-wide index.
How do you adjust used car price trends for changes in supply?
Imagine that cheaper, older cars leave the market and more younger cars arrive. The average price rises, even though no comparable group has become more expensive. The following calculation is illustrative and uses round figures:
| Group | Price in both periods | Share in first period | Share in second period |
|---|---|---|---|
| Younger cars | €30,000 | 40% | 60% |
| Older cars | €20,000 | 60% | 40% |
| Weighted average price | — | €24,000 | €26,000 |
The average rose by €2,000 purely because the group composition changed. A fixed-weight index would remain at the same value in this example. The median does not solve the problem; it makes it worse. In the first period it lies in the cheaper group at €20,000, in the second it moves into the more expensive group and jumps to €30,000.
What should a comparable group contain?
For everyday work, it pays to split cars by country, make and model range, generation, body type, powertrain, transmission, age and mileage. Equipment explains differences within a group. If you split groups too finely, the samples become small and prices within them jump around.
Working on two levels has proved effective:
- Detailed groups are used to value cars and investigate the causes of changes.
- Broader groups with fixed weights are used to track the trend and for regular reporting.
| Group tracked | Change that can distort the price | Solution |
|---|---|---|
| Compact petrol cars | More cars with automatic transmissions | Separate transmissions or account for their effect statistically |
| Mid-size diesel estates | More cars with higher mileage | Track prices separately in mileage bands |
| Compact electric cars | More versions with a larger battery | Distinguish versions by battery capacity |
Alongside the adjusted index, keep the actual supply mix in the report too. Its effect is a distraction when explaining a price movement, but you need to know it for planning purchases and sales. What sells on the European market and how the share of powertrains and transmissions is changing is described in a separate Carsdata article on European used car market trends.
How do you compare used car prices across Europe?
A country comparison answers two different questions. The first is where a comparable car is more expensive, and answering it requires amounts in euros. The second is where prices are falling faster, and answering it requires price series built in the same way. The indices of two countries can both start at 100, even if their starting prices differ considerably.
Compare both countries in the same sales channel, on the same price basis, over the same period and using cars with the same characteristics. If, for example, German supply has more younger cars with automatic transmissions than French supply, the difference in overall averages says nothing about the difference between the countries. For markets outside the euro area, such as Poland or the Czech Republic, exchange rate movements also feed into prices in euros. Part of the change therefore does not originate in the car market at all.
A single figure per country is only a rough guide. The average asking price tracked by one German classifieds platform was around €26,728 in October 2026. It describes the mix of that supply and does not show the price of a comparable car in Europe. For cross-border analysis, it is more accurate to compare the same group of cars in Germany, France and Italy and only then compile a summary.
A European overview therefore rests on two layers. The first is a common index with fixed country weights, the second is separate national series. Without fixed weights, the aggregate price can change simply because more cars from a more expensive market have entered the data.
What do listing times and supply volume tell you?
Asking prices react with a delay. Sellers first wait longer, then give discounts during negotiations and only finally cut the price in the listing. Indicators of stock levels and speed of sale therefore pick up price pressure earlier than a price index.
| Combination of signals | Likely explanation | Impact on decisions |
|---|---|---|
| Supply is growing and listing times are lengthening | Cars are arriving faster than the market can sell them | More cautious buying and a check on the price position of stock |
| Supply is growing and prices are falling | Sellers are already reacting to high stock levels | Update valuations and sales proceeds scenarios |
| Supply is falling and listing times are shortening | Demand is strengthening, or fewer new cars are coming onto the market | Separate the two effects using new and removed listings |
| The average price is rising and more younger cars are arriving | The rise may be caused by the supply mix | Rely on the adjusted index |
Typical listing times in Germany and the Czech Republic by age and powertrain are summarised in the Carsdata report on used car time to sell. For business decisions, however, how this time is changing in your group of cars carries more weight than comparing a single car with the market-wide average.
Supply volume does not explain itself either. The number of electric cars on one large German classifieds platform fell by 7.2% in October 2026 compared with September. The cause could be faster sales or fewer new cars coming onto the market. Only the number of new and removed listings and their age can tell the two apart.
In the groups that matter to you, track these leading indicators:
- the number of new listings against removed listings;
- the share of listings that have been on offer much longer than is usual in the group;
- the number and size of asking price reductions;
- listing time for listings that have already ended.
Listing time differs from time in stock. A car may be in reconditioning before it is published and may still stand at the dealer after the listing is removed. For active listings, the final listing duration is not yet known. The overview should therefore separate the age of current supply from the duration of completed listings.
How do you translate a price trend into portfolio risk?
For leasing companies and banks, used car prices are a direct component of risk. The European Commission attributes around 30% of annual new passenger car sales to leasing. These cars return to the used car market when their contracts end. Renting contracts in Spain lasted around 49 months on average in 2025, so the cars come back at roughly four years old. For the portfolio, what counts is the prices of this age group at the time of return, not today's market-wide average.
Risk is not necessarily highest where prices are falling fastest. It grows where a price fall coincides with large exposure and an imminent sale date.
| Situation | What to link with price trends | What risk this captures |
|---|---|---|
| A dealer holds cars in stock | Acquisition value, stock age and market valuation | Shrinking margin and tied-up capital |
| A leasing company expects cars to be returned | Agreed residual values and return dates | Gap between expected value and sales proceeds |
| A bank finances cars | Outstanding principal and collateral value | Weakening loan cover by the car's value |
For leasing, it pays to split the portfolio by month or quarter of return. When many similar cars return at the same time, the portfolio is more sensitive than one of the same size spread over more dates and segments. For loans, a fall in the car's value is not the same as a credit loss. The loss also depends on repayments and on the proceeds from selling the collateral. Used car prices are one input into risk management here; they do not replace the whole credit model.
Warning signs for the portfolio
- Prices in a group that makes up a large share of returns in the next 6–12 months fall several months in a row, even though the market-wide index is flat.
- Supply in this group grows faster than the market as a whole, for example when more cars of the same model year come back off lease.
- Listing times lengthen and asking price reductions increase.
- The estimated market value falls below the agreed residual value or below the outstanding principal.
- Manufacturers cut new car prices in the same segment or launch a new model generation. Young used cars then lose their price advantage, even if their supply is not yet growing.
The current value of individual cars in an at-risk group can be estimated by Price Report based on comparable cars on the market. Alongside the price, it adds a range, the confidence level of the estimate and the expected time to sell.
How to build a price scenario
A scenario shows how sensitive the portfolio is to falling prices. It does not forecast market developments. Illustrative example: a portfolio of 100 cars worth €20,000 each has a total value of €2,000,000. A 5% fall in prices reduces its value by €100,000, a 10% fall by €200,000.
In practice, different scenarios for individual groups of cars give a better picture than a single percentage for the whole portfolio. A longer time to sell also belongs in the scenario. Receiving the same amount two months later is a different burden for a company than selling on schedule.
How do you feed the data into your own systems?
Monitoring portals manually means repeatedly searching for comparable cars, removing duplicates and storing the history. When a seller re-lists the same car, without a persistent listing ID it looks like a new car on the market in the data. Supply then appears to grow.
A usable data flow has four parts:
- Consistent vehicle descriptions. The same names for powertrains, transmissions, body types and other characteristics allow comparison across sources.
- Observation history. The capture date, price changes and the first and last appearance of a listing show the development, not just the current state.
- Link to your own portfolio. Internal records add acquisition value, stocking date or lease return date.
- Versioned outputs. For each index and valuation, the calculation date and methodology used remain on record so that decisions can be explained retrospectively.
For operational meetings, a weekly overview of changes in stock and supply is enough; for tracking the price trend, a monthly series. For thinly represented groups, a quarterly series is more stable.
Carsdata data and results can be imported into your own systems via API (REST, JSON) and as exports. Custom dashboards, reports and data feeds are also available for banks, leasing companies and importers. Dealers using Omnetic DMS have Carsdata integrated directly in the system.
When combining with public statistics, expect differences in data structure. Eurostat's statistical API returns data in the JSON-stat 2.0 format, which has its own structure. Before combining it with market data, both datasets therefore need to be converted to a common data model.
What to watch out for when reading the results
- A new data source can create a false trend. When a portal or auction channel is added to the data, the mix of cars changes. Historical comparison therefore requires the same scope of sources or a recalculation of earlier periods.
- A re-listed car looks new. Without linking records, a long-unsold car appears as fresh supply and listing time comes out shorter.
- A weak listing does not necessarily mean a weak market. Missing equipment in the description or poor photos reduce buyer interest even for a correctly priced car.
- A small sample causes sharp swings. For a narrow group, a report should always show the number of observations too. An average without the sample size looks more precise than it really is.
Carsdata: price trends from data covering the whole European market
Tracking prices reliably means cleaning listings, linking re-listed offers and standardising vehicle descriptions from different portals. Carsdata collects used car listings and auction offers from European portals and standardises vehicle data. Decisions on pricing, buying and stock are therefore based on data from the whole market, not on a look at a single portal.
- Market Data shows supply, prices, time on market and estimated sold vehicles through dashboards, regular reports and data feeds. Analysts can see price trends together with leading indicators for their market and segment.
- Price Report estimates the realistic selling price of a specific car from comparable cars selected by make, model, age, mileage, equipment and region. It gives a price range, a confidence level and an expected time to sell. It is suitable both for cars returned from leasing and for cars serving as loan collateral.
- Stock Report compares a dealer's stock with the market, recommends a price for each car and flags cars that need action. It also shows stock value and days in stock, including across a branch network.
Book a Carsdata demo for your market, stock or portfolio.
FAQ
Does it make sense to track inflation-adjusted prices?
Yes, if you are comparing long periods and want to see whether cars are becoming more or less expensive in real terms. For current stock valuation and sales proceeds, however, nominal prices are used, i.e. the amounts at which cars are actually traded.
When is seasonal adjustment appropriate?
Only once you have several years of monthly history and the scope of sources has not changed in the meantime. On a short series, the model cannot tell a seasonal swing from a change in trend and may create a turning point that has not happened in the market.
Why can a historical index value change retrospectively?
Data processed later may be added to the calculation, or the methodology may change. When reviewing decisions retrospectively, keep both the originally published value and the later recalculation.
Can two different price indices be combined into one series?
Only if you know how each is defined and their series overlap for at least some time. Otherwise, simply chaining the values creates an apparent jump or, conversely, hides a change in methodology. In that case, it is clearer to show both series side by side.
Sources
- https://ec.europa.eu/eurostat/web/user-guides/data-browser/api-data-access/api-introduction
- https://data.consilium.europa.eu/doc/document/ST-7476-2025-INIT/en/pdf
- https://ae-renting.es/prensa-noticias/notas-de-prensa/el-renting-de-vehiculos-cierra-el-ano-con-un-incremento-del-695-en-parque-del-550-en-clientes-y-del-735-en-inversiones-en-compra/
- https://www.autohaus.de/nachrichten/gw-trends/ebrauchtwagenmarkt-unter-druck-bestaende-steigen-preise-sinken-3925325
- https://www.autohaus.de/nachrichten/gw-trends/autoscout24-gebrauchte-e-autos-werden-gegen-den-markttrend-teurer-3929761
- https://carsdata.com/about
- https://carsdata.com/products/market-data
- https://carsdata.com/products/price-report
- https://carsdata.com/products/stock-report


