The Danish used-car market is not falling as fast as it looks
How asking prices and residual values of used cars in Denmark moved from January 2025 to September 2026, by powertrain, and how much of the movement comes from which cars were on offer. Asking prices, not transactions. Residual value and asking prices by powertrain · January 2025 – September 2026
Every month, Danish dealers create used-car listings. We take the price they ask and divide it by what that specific car cost new when it was first registered, including registration tax. That ratio, RV%, tells you how much of the original price a car is still being offered for.
This report covers roughly 420,000 listings across twenty-one months.
One thing to know before the numbers: Danish registration tax sits in both the asking price and the reference price. That makes RV% here a different measure than in a market without that tax. It is interpretable within Denmark, and not comparable to Germany.
The question this report answers
The question this report answersUsed electric-car prices in Denmark have risen since the turn of the year. Residual values across the rest of the market have fallen in most of the past twenty-one months.
Both are averages. And an average can move for two quite different reasons:
Are the same cars being priced differently, or has the supply shifted toward different cars?
A price average at two points in time cannot tell those apart. Both produce the same movement. But they mean very different things to a dealer holding stock.
Because we measure every car against its own new price, and because we can hold age constant, we can separate them. That is what this report does, applied in both directions: to the rise since winter, and to the long decline that came before it.
What the rise is made of
What the rise is made ofThe charts: two line charts, one for asking prices and one for RV%, indexed so that January–March 2025 = 100. Each powertrain is indexed against its own starting point.
Reading the index. 100 is where each powertrain stood in January–March 2025. An index of 103 to 104 means residual values are 103 to 104% of where they were then. It does not mean cars are asked at that share of their new price, and it does not mean any car is asked at more as it ages. Every car in this data is still asked at a lower share of its new price the older it gets. The index tracks whether that share falls faster or slower than before.
View data
| Jan 2025 | Feb 2025 | Mar 2025 | Apr 2025 | May 2025 | Jun 2025 | Jul 2025 | Aug 2025 | Sep 2025 | Oct 2025 | Nov 2025 | Dec 2025 | Jan 2026 | Feb 2026 | Mar 2026 | Apr 2026 | May 2026 | Jun 2026 | Jul 2026 | Aug 2026 | Sep 2026 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Electric | 104 | 100 | 99.8 | 100 | 100 | 100 | 100 | 104 | 99.9 | 100 | 98 | 96.7 | 96 | 95.2 | 94 | 98 | 99.8 | 104 | 104 | 104 | 104 |
| Plug-in hybrid | 100.2 | 97.9 | 99.7 | 97.9 | 95.6 | 95.6 | 95.6 | 95.6 | 91.1 | 91.1 | 91.1 | 88.8 | 88.8 | 86.5 | 86.5 | 86.5 | 84.1 | 84.2 | 84.2 | 84.2 | 82 |
| Petrol | 105.4 | 105.4 | 100 | 97.4 | 97.6 | 100 | 97.6 | 94.1 | 100 | 94.2 | 94.1 | 91.5 | 92.8 | 93.4 | 92.7 | 91.5 | 88.3 | 93.9 | 92.9 | 88.2 | 82.3 |
| Diesel | 110.9 | 100 | 99.6 | 99.6 | 94.4 | 94.4 | 100 | 94.4 | 100 | 99.8 | 94.4 | 94.4 | 88.9 | 94.2 | 94.2 | 88.4 | 88.9 | 88.3 | 86.7 | 83.3 | 81.2 |
Brand-weighted median asking price, indexed per powertrain · Baseline Jan–Mar 2025 = 100. Values at the central cut-off for implausibly high residual values (1.20); where the text gives a range, it spans the cut-offs from 1.10 to 1.30.
Source: National used-car listing data · n = 416,000 · Lens: flow (listings created per month) · January 2025 – September 2026 · Asking prices divided by Danish list price when new, including registration tax. Neither end is a transaction.
View data
| Jan 2025 | Feb 2025 | Mar 2025 | Apr 2025 | May 2025 | Jun 2025 | Jul 2025 | Aug 2025 | Sep 2025 | Oct 2025 | Nov 2025 | Dec 2025 | Jan 2026 | Feb 2026 | Mar 2026 | Apr 2026 | May 2026 | Jun 2026 | Jul 2026 | Aug 2026 | Sep 2026 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Electric | 101.2 | 99.7 | 99.6 | 99.6 | 99.3 | 99.6 | 99.9 | 97.3 | 97.1 | 97.6 | 95.7 | 94.8 | 93.5 | 93.9 | 92 | 95.8 | 96.9 | 100.4 | 101.6 | 101.1 | 103.5 |
| Plug-in hybrid | 101 | 99.2 | 99.6 | 98.4 | 96.3 | 96.4 | 96.8 | 95.1 | 94.2 | 93.2 | 91.6 | 90.4 | 89.5 | 87.8 | 87 | 87.2 | 85.2 | 84.6 | 83 | 82.3 | 82.3 |
| Petrol | 101.4 | 99.7 | 99.2 | 97.7 | 97.1 | 97.8 | 96.8 | 95 | 99.2 | 94.5 | 93.3 | 91.4 | 88.4 | 91.5 | 91 | 92.4 | 90.5 | 89.7 | 89 | 87.3 | 85 |
| Diesel | 105.8 | 97.3 | 97.6 | 97.2 | 95.6 | 93.8 | 95 | 94.3 | 97 | 94.7 | 94.3 | 92.2 | 90.8 | 87.8 | 91.4 | 88.6 | 86.9 | 90.6 | 85.1 | 82 | 78.9 |
Brand-weighted median RV%, indexed per powertrain · Baseline Jan–Mar 2025 = 100. Values at the central cut-off for implausibly high residual values (1.20); where the text gives a range, it spans the cut-offs from 1.10 to 1.30.
Source: National used-car listing data · n = 323,000 · Lens: flow (listings created per month) · January 2025 – September 2026 · Asking prices divided by Danish list price when new, including registration tax. Neither end is a transaction.
Where things stand in September 2026:
| Asking price | RV% | Difference |
|---|---|---|---|
Electric | 104 | 103 to 104 | 0 to 1 |
Petrol | 82 | 85 | -3 |
Diesel | 81 | 79 | 2 |
Plug-in hybrid | 82 | 82 | 0 |
Electric at 103 to 104 means electric residual values stand 3 to 4% above where they stood in January–March 2025. Across the window, a two- to three-year-old electric car is asked at a median of 59% of what it cost new.
On “brand-weighted”. Not every listing comes with a new price, and the gaps are not spread evenly across brands. So instead of a plain average, each month’s figure gives every brand the same share it holds in the market as a whole. A brand we happen to see more of does not get to count for more.
Read the two columns against each other. That comparison is the point.
If dealers were listing more expensive electric cars than before, meaning newer models or higher trims, the asking price would rise while RV% stayed flat. RV% measures each car against its own new price, so a more expensive car does not lift it. The two lines would separate.
They do not. For electric they sit 0 to 1 points apart, and across all four powertrains the gap runs from -3 for petrol to 2 for diesel.
The gap is the measure, not just the evidence. Where price runs ahead of residual value, the cars on offer carry higher new prices than before; where it runs behind, lower ones. That covers model and trim, and it covers age as well: an older car carries the new price of its own, earlier registration year. Diesel’s 2 points say the new prices of the diesel cars being listed are higher than before; petrol’s -3 say lower. Electric’s gap is 0 to 1 points. Each line is a median of its own, so the distance between them is an indication, not an exact split.
What this comparison cannot see. It catches a shift toward cars with higher or lower new prices, because a more expensive car does not lift RV%. It does not catch a shift toward cars that are asked at a higher share of their new price: younger cars, or models that hold their price better. That part lifts both lines together; the gap only catches the difference in new price that comes with it. Age is the correction section 3 turns out to be about, and it is handled separately below. Brands are held fixed by the weighting described in section 2; models within a brand are not.
So the rise is not a shift toward more expensive cars.
The same holds when we fix the age
A price average can also rise because the cars in the listing pool are simply younger than they were. RV% would not catch that on its own, since a younger car is asked at a higher share of its new price, which is exactly what it should do.
Tracking each age band separately, in index points per month:
| Trend before the new year | Trend after |
|---|---|---|
Electric, 2–3 years | -0.4 | 1.5 |
Electric, 4–5 years | 0.0 | 1.8 |
Two- to three-year-old and four- to five-year-old electric cars turn upward at the same time, with age drift inside those bands of 0.0 and 0.4 years. The cars do not get younger. They are still being asked more for.
Two things could have produced the rise without any car being asked at a higher share of its new price. Neither did. Not a shift toward cars with higher new prices, because price and RV% move together. Not a younger pool, because the rise holds within fixed age bands. A shift toward models that hold their price better, within a brand, is not something either check can see.
What this means
A two- to three-year-old electric car listed in Denmark in September 2026 is being asked at a higher share of its original price than one of the same age was asked at a year ago. It is still asked at less than it cost new. That has not changed. What has changed is the rate.
Why it is happening sits outside what listing data can see. We observe what sellers ask, not who is buying or from where. Within what the checks in this section and the next can see, no change in what is on offer accounts for it.
It is not the brands that moved, except in three months
It is not the brands that moved, except in three monthsBefore turning to the decline that came before, one more check on the same question, because brand mix is the correction most people would reach for.
The chart: one per powertrain, each with two lines: the raw median, and the median after weighting for which brands are in the market each month. The distance between the lines is what brand mix accounts for.
Weighting by brand moves the series by less than two points, except in three months
View data
| Jan 2025 | Feb 2025 | Mar 2025 | Apr 2025 | May 2025 | Jun 2025 | Jul 2025 | Aug 2025 | Sep 2025 | Oct 2025 | Nov 2025 | Dec 2025 | Jan 2026 | Feb 2026 | Mar 2026 | Apr 2026 | May 2026 | Jun 2026 | Jul 2026 | Aug 2026 | Sep 2026 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Electric, raw median | 100.7 | 99.1 | 100.1 | 99.7 | 99.5 | 100.5 | 101.2 | 93.7 | 95.3 | 97.8 | 95.7 | 94.9 | 94 | 94.6 | 92.8 | 97.1 | 97.9 | 100.5 | 102 | 101.5 | 104 |
| Electric, brand-weighted | 101.2 | 99.7 | 99.6 | 99.6 | 99.3 | 99.6 | 99.9 | 97.3 | 97.1 | 97.6 | 95.7 | 94.8 | 93.5 | 93.9 | 92 | 95.8 | 96.9 | 100.4 | 101.6 | 101.1 | 103.5 |
View data
| Jan 2025 | Feb 2025 | Mar 2025 | Apr 2025 | May 2025 | Jun 2025 | Jul 2025 | Aug 2025 | Sep 2025 | Oct 2025 | Nov 2025 | Dec 2025 | Jan 2026 | Feb 2026 | Mar 2026 | Apr 2026 | May 2026 | Jun 2026 | Jul 2026 | Aug 2026 | Sep 2026 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Plug-in hybrid, raw median | 100.9 | 99.3 | 99.5 | 98.7 | 96.5 | 96 | 97 | 95 | 94.1 | 93.3 | 91.5 | 90.2 | 90.1 | 88.2 | 87.1 | 87.3 | 84.6 | 85.2 | 83.5 | 83.3 | 82.8 |
| Plug-in hybrid, brand-weighted | 101 | 99.2 | 99.6 | 98.4 | 96.3 | 96.4 | 96.8 | 95.1 | 94.2 | 93.2 | 91.6 | 90.4 | 89.5 | 87.8 | 87 | 87.2 | 85.2 | 84.6 | 83 | 82.3 | 82.3 |
View data
| Jan 2025 | Feb 2025 | Mar 2025 | Apr 2025 | May 2025 | Jun 2025 | Jul 2025 | Aug 2025 | Sep 2025 | Oct 2025 | Nov 2025 | Dec 2025 | Jan 2026 | Feb 2026 | Mar 2026 | Apr 2026 | May 2026 | Jun 2026 | Jul 2026 | Aug 2026 | Sep 2026 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Petrol, raw median | 101.2 | 99.8 | 99.3 | 97.3 | 97.2 | 98.2 | 97.1 | 95.7 | 103.9 | 94.6 | 93.2 | 92 | 88.6 | 91.8 | 90.9 | 92.5 | 90.9 | 90.5 | 89.2 | 87.9 | 85.1 |
| Petrol, brand-weighted | 101.4 | 99.7 | 99.2 | 97.7 | 97.1 | 97.8 | 96.8 | 95 | 99.2 | 94.5 | 93.3 | 91.4 | 88.4 | 91.5 | 91 | 92.4 | 90.5 | 89.7 | 89 | 87.3 | 85 |
View data
| Jan 2025 | Feb 2025 | Mar 2025 | Apr 2025 | May 2025 | Jun 2025 | Jul 2025 | Aug 2025 | Sep 2025 | Oct 2025 | Nov 2025 | Dec 2025 | Jan 2026 | Feb 2026 | Mar 2026 | Apr 2026 | May 2026 | Jun 2026 | Jul 2026 | Aug 2026 | Sep 2026 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Diesel, raw median | 105.4 | 98 | 98.5 | 98.2 | 96.7 | 95.1 | 96 | 95.8 | 98.5 | 96.4 | 94.9 | 93.5 | 91.8 | 88.9 | 93.2 | 89.8 | 88.1 | 91.6 | 87.2 | 82.9 | 79.8 |
| Diesel, brand-weighted | 105.8 | 97.3 | 97.6 | 97.2 | 95.6 | 93.8 | 95 | 94.3 | 97 | 94.7 | 94.3 | 92.2 | 90.8 | 87.8 | 91.4 | 88.6 | 86.9 | 90.6 | 85.1 | 82 | 78.9 |
RV%-index, raw against brand-weighted · Baseline Jan–Mar 2025 = 100. In a normal month the two lines overlap. Values at the central cut-off for implausibly high residual values (1.20); where the text gives a range, it spans the cut-offs from 1.10 to 1.30.
Source: National used-car listing data · n = 323,000 · Lens: flow (listings created per month) · January 2025 – September 2026 · Asking prices divided by Danish list price when new, including registration tax. Neither end is a transaction.
What to read: in a normal month the lines sit on top of each other. The typical gap, in the table below, is under two index points for all four powertrains. Then there are three months where at least one series parts by two points or more: August 2025 for electric, September 2025 for petrol and July 2026 for diesel.
| Typical month | Largest gap | When |
|---|---|---|---|
Petrol | 0 | 5 | September 2025 |
Electric | 0 to 1 | 3 to 4 | August 2025 |
Diesel | 1 | 2 | July 2026 |
Plug-in hybrid | 0 | 1 | August 2026 |
Those three months stand out from the rest.
Which is what the weighting is for. It is not a formality that leaves the answer unchanged. For eighteen of twenty-one months it moves every series by less than two points, and in the three months that stand out, it keeps the gap out of the series.
One further check on the same question. Brands too small to weight individually are pooled into a single group, so drift inside that group is not corrected. In electric that group made up 5.1% of listings in January 2025 and 6.0% in September 2026. Removing the group entirely leaves nearly all of the rise in place. Without it, electric residual values were 60% of new price in March 2026 and 66 to 67% in September 2026; with it, 60% and 67%. The brands in that group are not what lifted it.
So brand mix is not what moved the average. Something else was, and it changes what the headline decline means.
The fall before it was not what it looked like either
The fall before it was not what it looked like eitherIn the headline series, residual values fell 15% for petrol, 18% for plug-in hybrids and 21% for diesel across twenty-one months. That is a real measurement of a real market average. Electric, which rose, is the subject of section 1.
But less of it happened to the individual cars.
Held at a fixed age, so that three-year-old cars are compared against three-year-old cars, every age band we can measure stands above the headline series. In the oldest petrol and diesel bands, residual values did not fall.
To be clear about what that does and does not say. It is not the finding that older cars are asked at a lower share of their new price than younger ones. That is true, and unremarkable, and every dealer already knows it. The finding is that for petrol, diesel and plug-in hybrids the market average fell faster than the cars in any of their fixed age bands did, because the pool of cars on offer aged underneath the average while it was being taken.
Why the average moved anyway
The cars being listed in Denmark are getting older. Not slightly, but measurably:
| Median age, January 2025 | Median age, September 2026 | Change |
|---|---|---|---|
Diesel | 8.54 yr | 10.15 yr | +1.6 |
Petrol | 7.99 yr | 9.47 yr | +1.5 |
Plug-in hybrid | 3.57 yr | 5.01 yr | +1.4 |
Electric | 2.12 yr | 2.75 yr | +0.6 |
Petrol, diesel and plug-in hybrids aged by between 1.4 years for plug-in hybrids and 1.6 for diesel across twenty-one months. A market average taken over an ageing pool falls even when nothing happens to the cars in it.
Holding the age fixed
Each age band on its own, in September 2026:
| Headline series | Within a fixed age band | Difference |
|---|---|---|---|
Diesel, 6–8 years | 79 | 102 | +23 |
Plug-in hybrid, 6–8 years ◆ | 82 | 104 | +22 |
Diesel, 9+ years | 79 | 100 | +21 |
Petrol, 9+ years | 85 | 100 | +15 |
Electric, 0–1 years | 103 to 104 | 117 | +13 to +14 |
Diesel, 4–5 years ◆ | 79 | 92 | +13 |
Petrol, 6–8 years | 85 | 98 | +13 |
Plug-in hybrid, 2–3 years | 82 | 90 | +8 |
Petrol, 2–3 years | 85 | 91 | +6 |
Plug-in hybrid, 4–5 years ◆ | 82 | 88 | +6 |
Electric, 4–5 years | 103 to 104 | 109 | +5 to +6 |
Petrol, 4–5 years | 85 | 89 to 90 | +4 to +5 |
Electric, 2–3 years | 103 to 104 | 105 | +1 to +2 |
◆ rests on thinner ground. See the method note on the limits of the age bands.
Every one of the thirteen bands sits above the headline series. There are no exceptions.
Held at a fixed age, nine-year-old petrol cars are asked at the same share of their new price as they were in January–March 2025. The headline series for all petrol cars fell 15% over the same months. A nine-year-old petrol car did not. There are simply more of them, and they are older.
Which changes how the current rise should be read. If the underlying cars never fell as far as the average suggested, then what is happening now is less of a reversal and more of a continuation. The average was moving for two reasons. Only one of them was the cars.
What this means in practice: if you are valuing a specific car, the market-wide average is not the number you want, in either direction.
We expected brand composition to be the correction that mattered. Outside those three months it moves the series by less than two points. Age moves it by up to 23. We were looking in the wrong place.
How fast residual value falls, by powertrain
How fast residual value falls, by powertrainThe chart: one curve per powertrain, shown separately, with median RV% against age.
How fast residual value falls, by powertrain
View data
| 0.0 yr | 0.2 yr | 0.5 yr | 0.7 yr | 1.0 yr | 1.2 yr | 1.5 yr | 1.7 yr | 2.0 yr | 2.2 yr | 2.5 yr | 2.7 yr | 3.0 yr | 3.2 yr | 3.4 yr | 3.7 yr | 3.9 yr | 4.2 yr | 4.4 yr | 4.7 yr | 4.9 yr | 5.2 yr | 5.4 yr | 5.7 yr | 5.9 yr | 6.2 yr | 6.4 yr | 6.7 yr | 6.9 yr | 7.1 yr | 7.4 yr | 7.6 yr | 7.9 yr | 8.1 yr | 8.4 yr | 8.6 yr | 8.9 yr | 9.1 yr | 9.4 yr | 9.6 yr | 9.9 yr | 10.1 yr | 10.3 yr | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Electric, median RV% | 98 | 95 | 91 | 87 | 83 | 79 | 75 | 70 | 66 | 63 | 60 | 59 | 58 | 57 | 55 | 54 | 52 | 54 | 49 | 44 | 44 | 42 | 42 | 41 | 38 | 36 | 36 | 32 | 32 | 31 | 32 | 29 | 30 | 23 | 23 | 20 |
View data
| 0.0 yr | 0.2 yr | 0.5 yr | 0.7 yr | 1.0 yr | 1.2 yr | 1.5 yr | 1.7 yr | 2.0 yr | 2.2 yr | 2.5 yr | 2.7 yr | 3.0 yr | 3.2 yr | 3.4 yr | 3.7 yr | 3.9 yr | 4.2 yr | 4.4 yr | 4.7 yr | 4.9 yr | 5.2 yr | 5.4 yr | 5.7 yr | 5.9 yr | 6.2 yr | 6.4 yr | 6.7 yr | 6.9 yr | 7.1 yr | 7.4 yr | 7.6 yr | 7.9 yr | 8.1 yr | 8.4 yr | 8.6 yr | 8.9 yr | 9.1 yr | 9.4 yr | 9.6 yr | 9.9 yr | 10.1 yr | 10.3 yr | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Plug-in hybrid, median RV% | 96 | 95 | 92 | 89 | 85 | 81 | 79 | 77 | 73 | 71 | 70 | 68 | 67 | 65 | 65 | 63 | 62 | 61 | 59 | 57 | 57 | 55 | 53 | 52 | 50 | 51 | 50 | 46 | 46 | 45 | 44 | 43 | 43 | 44 | 43 | 41 | 41 | 41 | 37 | 38 | 38 | 38 | 34 |
View data
| 0.0 yr | 0.2 yr | 0.5 yr | 0.7 yr | 1.0 yr | 1.2 yr | 1.5 yr | 1.7 yr | 2.0 yr | 2.2 yr | 2.5 yr | 2.7 yr | 3.0 yr | 3.2 yr | 3.4 yr | 3.7 yr | 3.9 yr | 4.2 yr | 4.4 yr | 4.7 yr | 4.9 yr | 5.2 yr | 5.4 yr | 5.7 yr | 5.9 yr | 6.2 yr | 6.4 yr | 6.7 yr | 6.9 yr | 7.1 yr | 7.4 yr | 7.6 yr | 7.9 yr | 8.1 yr | 8.4 yr | 8.6 yr | 8.9 yr | 9.1 yr | 9.4 yr | 9.6 yr | 9.9 yr | 10.1 yr | 10.3 yr | 10.6 yr | 10.8 yr | 11.1 yr | 11.3 yr | 11.6 yr | 11.8 yr | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Petrol, median RV% | 100 | 97 | 95 | 94 | 91 | 87 | 84 | 82 | 80 | 77 | 77 | 75 | 74 | 74 | 72 | 72 | 71 | 69 | 68 | 68 | 66 | 60 | 63 | 62 | 60 | 59 | 57 | 54 | 53 | 52 | 50 | 49 | 47 | 46 | 45 | 43 | 42 | 40 | 40 | 38 | 38 | 36 | 35 | 34 | 33 | 32 | 30 | 30 | 28 |
View data
| 3.0 yr | 3.2 yr | 3.4 yr | 3.7 yr | 3.9 yr | 4.2 yr | 4.4 yr | 4.7 yr | 4.9 yr | 5.2 yr | 5.4 yr | 5.7 yr | 5.9 yr | 6.2 yr | 6.4 yr | 6.7 yr | 6.9 yr | 7.1 yr | 7.4 yr | 7.6 yr | 7.9 yr | 8.1 yr | 8.4 yr | 8.6 yr | 8.9 yr | 9.1 yr | 9.4 yr | 9.6 yr | 9.9 yr | 10.1 yr | 10.3 yr | 10.6 yr | 10.8 yr | 11.1 yr | 11.3 yr | 11.6 yr | 11.8 yr | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Diesel, median RV% | 59 | 61 | 59 | 59 | 59 | 58 | 58 | 56 | 55 | 53 | 51 | 49 | 48 | 47 | 45 | 43 | 42 | 40 | 38 | 37 | 36 | 34 | 32 | 30 | 29 | 28 | 27 | 25 | 24 | 23 | 22 | 21 | 20 | 20 | 19 | 18 | 18 |
Median RV% by age, in 90-day steps. One chart per powertrain, on the same scale: registration tax sits in the reference price and differs sharply by powertrain, so compare the shape of each curve, not the heights. Values at the central cut-off for implausibly high residual values (1.20); where the text gives a range, it spans the cut-offs from 1.10 to 1.30.
Source: National used-car listing data · n = 323,000 · Lens: flow (listings created per month) · January 2025 – September 2026 · Asking prices divided by Danish list price when new, including registration tax. Neither end is a transaction.
Compare the shape of each curve, not the heights. Danish registration tax sits in the reference price and differs sharply by powertrain, so the levels are not comparable across powertrains. The tax scales the points lost as well, so where this section compares the drop between powertrains, the comparison holds both in points and as a share of where each curve starts.
Band | Diesel | Petrol | Electric | Plug-in hybrid |
|---|---|---|---|---|
0–1 yr | 80 to 81% | 89 to 90% | 84% | 84% |
2–3 yr | 61% | 74% | 59% | 66% |
4–5 yr | 53% | 66% | 48% | 58% |
6–8 yr | 38% | 51% | 33% | 47% |
9+ yr | 17% | 26% | 22% | 36% |
Each figure is the middle car of its group, not an average price divided by an average new price.
One column reads differently from the others. The plug-in hybrid figures describe a category that changes as the cars get older. Measured by registration year, each later year carries a larger share of hybrids without a plug, so the youngest cars hold more of them than the middle bands. Among the oldest the same is expected, because Denmark had almost no plug-in cars before 2021, but it has not been measured. Read that column for its level at a given age. Reading it downwards traces the changing mixture as much as the cars.
Read down a column, not across a row. Electric cars are asked at 84% of their new price in the first year and 59% by the third, the steepest early drop of the four. Petrol is at 89 to 90% and 74% over the same span.
Age is the only thing held constant here. Mileage, condition and service history are not, so two cars of the same age can sit far apart inside a band. That is why these are market-level figures and not a valuation for a specific car.
Two cells rest on very little, and both are facts about the market rather than gaps in the data. Diesel at 0–1 year is 69 cars across twenty-one months. Almost nobody registers a new diesel in Denmark any more, so almost none reach the used market at that age. Electric at 9+ years is 567, because Denmark’s electric fleet is barely old enough to have any.
This is also why the diesel curve begins at three years. The curve is drawn in 90-day steps, and the youngest diesel steps hold too few cars to plot. The table keeps the band because the emptiness is itself the finding.
When the turn happened
When the turn happenedElectric residual values fall like everything else, bottom out in March 2026 at 92, and rise from there, reaching 103 to 104 by September 2026. It is the only powertrain back above its starting point. The charts in section 1 show it. The trend changes at the turn of the year, which is what sections 1 and 6 measure; the level keeps falling until March 2026, and that low point is what this section dates.
The underlying figures are worth seeing without the index. From the low point in March, the median asking price for a used electric car ran from 235,000 kr to 260,000 kr by September, close to 11% over those months.
Held at a fixed age it is more, not less. Cars in the two- to three-year band rose 14% over the same months, and cars in the four- to five-year band 20%. The two- to three-year band is the larger of the two and the stronger reading. The older band carries its rise while its own cars age slightly, which works against it. The raw figure understates what happened to a given car, in the same way section 3 found for the decline.
What the turn is not
We can rule several explanations out, and we think that is worth stating plainly even though it leaves the cause open.
Not one large seller. Across March to September, the largest single seller accounted for between 1.9% and 2.8% of electric listings in any month, with never fewer than 780 sellers active in any of those months.
Not brand composition. Section 2 shows the weighted and unweighted series sitting on top of each other through the whole rise, and removing the pooled group of small brands entirely leaves nearly all of the rise in place.
Not the age of the cars. Fixed age bands rise more steeply than the raw series, not less.
Not the January change in test method. Section 6 covers that. It affects plug-in hybrids and nothing else.
What remains is outside what listing data can answer. We see what sellers ask, not what buyers pay, and not why. Naming a cause here would be a guess dressed as a finding.
Electric residual values in Germany also stand above where that report’s window starts, on a completely separate source and its own window. The German report runs the same separation there and states its own result.
Two markets and two unrelated data sources showing the same thing is worth noticing. Whether the two markets share a cause is not something listing data can settle.
Petrol and diesel do not turn with it. In the headline series petrol keeps its pace after the turn of the year and diesel does not turn upward. Section 3 shows what the headline series carries: over the window petrol fell 15% there and diesel 21%, while the oldest petrol and diesel cars, held at a fixed age, did not fall. The electric turn is so far the only one.
A change in test method that did not show up
A change in test method that did not show upFrom 1 January 2026, new plug-in hybrids are registered under the EU’s Euro 6e-bis test method. It lifts the official CO₂ figure of most plug-in hybrids above 50 g/km, and above that line they lose the lower Danish registration tax. Existing models had to be recalculated by 31 December 2025. The Danish tax rules themselves did not change. By FDM’s calculation in April 2025, the difference is about DKK 98,000 on a car priced at DKK 250,000, and DKK 200,000 or more on more expensive cars. Cars already registered are exempt.
That is an unusually clean setup: a dated change, decided outside the market and affecting one powertrain. Of the others, petrol turns out to be the clean comparison, as the test below shows.
The expectation was that used plug-in hybrids would become relatively more attractive and their residual values would rise through 2026.
The chart: the residual-value index from section 1, read around the date of the change. Nothing else is different. The line to look at is the plug-in hybrids.
View data
| Jan 2025 | Feb 2025 | Mar 2025 | Apr 2025 | May 2025 | Jun 2025 | Jul 2025 | Aug 2025 | Sep 2025 | Oct 2025 | Nov 2025 | Dec 2025 | Jan 2026 | Feb 2026 | Mar 2026 | Apr 2026 | May 2026 | Jun 2026 | Jul 2026 | Aug 2026 | Sep 2026 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Electric | 101.2 | 99.7 | 99.6 | 99.6 | 99.3 | 99.6 | 99.9 | 97.3 | 97.1 | 97.6 | 95.7 | 94.8 | 93.5 | 93.9 | 92 | 95.8 | 96.9 | 100.4 | 101.6 | 101.1 | 103.5 |
| Plug-in hybrid | 101 | 99.2 | 99.6 | 98.4 | 96.3 | 96.4 | 96.8 | 95.1 | 94.2 | 93.2 | 91.6 | 90.4 | 89.5 | 87.8 | 87 | 87.2 | 85.2 | 84.6 | 83 | 82.3 | 82.3 |
| Petrol | 101.4 | 99.7 | 99.2 | 97.7 | 97.1 | 97.8 | 96.8 | 95 | 99.2 | 94.5 | 93.3 | 91.4 | 88.4 | 91.5 | 91 | 92.4 | 90.5 | 89.7 | 89 | 87.3 | 85 |
| Diesel | 105.8 | 97.3 | 97.6 | 97.2 | 95.6 | 93.8 | 95 | 94.3 | 97 | 94.7 | 94.3 | 92.2 | 90.8 | 87.8 | 91.4 | 88.6 | 86.9 | 90.6 | 85.1 | 82 | 78.9 |
Brand-weighted RV%-index · Baseline Jan–Mar 2025 = 100. From 1 January 2026 new plug-in hybrids are measured under Euro 6e-bis, which lifts most of them out of the lower registration tax. On the chart, the change falls between Dec 2025 and Jan 2026. Values at the central cut-off for implausibly high residual values (1.20); where the text gives a range, it spans the cut-offs from 1.10 to 1.30.
Source: National used-car listing data · n = 323,000 · Lens: flow (listings created per month) · January 2025 – September 2026 · Asking prices divided by Danish list price when new, including registration tax. Neither end is a transaction.
The test is whether the trend changed at the turn of the year, not whether the level did. On a series that was already falling, comparing averages before and after would just measure the existing decline.
| Trend before | Trend after | Change |
|---|---|---|---|
Electric | -0.5 | 1.4 | 1.9 |
Petrol | -0.7 | -0.6 | 0.1 |
Plug-in hybrid | -0.9 | -0.9 | 0.0 |
Diesel | -0.7 | -1.3 | -0.6 |
Plug-in hybrids do not turn upward. Their trend after the turn of the year is the same as before. Electric is the only series that turns upward; petrol and diesel do not. Had the change lifted used plug-in residual values, plug-ins should have separated upward from petrol. They did not: the change in trend is 0.0 for plug-in hybrids against 0.1 for petrol. Diesel is not the comparison here, because its own trend steepened (-0.6), which lifts plug-ins against it whatever the change did.
It holds when age is held fixed too. Within the 2–3 year band the change in trend is -0.1, and within the 4–5 year band -0.1. Near zero in both.
And it holds when the category is not pure. This series counts hybrids with and without a plug together, and the change reached only the first. The share without a plug is not fixed. Measured on 25 September 2026, on listings from January 2025 to July 2026 in the 2–3 year band, it runs from 21% in the first month to 29% in the last. That movement is mechanical rather than behavioural — a fixed age band walks forward through registration years as time passes, and the later years carry more hybrids without a plug.
Measured separately, the two sit about 9 points apart. More to the point, they fall at the same rate: 3.2 points against 3.4 between the two calendar years. Two lines falling at the same rate. A shift in the mix moves where the combined series sits; it does not turn it.
This test reads direction, so the drift costs little. Allowing for it leaves plug-in hybrids at 0.0 — still level, and still far from electric, the one series that turns.
The level difference between the two types is not read here. Danish registration tax sits in the reference price and falls differently on them, the same reason section 1 compares each powertrain only against itself.
So the hypothesis is not supported. Somebody will otherwise assume the change moved the used market. As far as we can see, it did not.
And it does not explain the electric rise either. The change affects plug-in hybrids and nothing else. Electric cars have their own registration tax, and the phase-in of its higher rate was postponed by a year from 1 January 2026. Why they turn upward is still outside what this data can answer.
Reading these numbers
Reading these numbersDirection, not just movement. Twenty-one months is long enough to read a series as a direction rather than as movement inside a window. The turn in section 5 is the one place we hold back, because the change in trend rests on a single January and seasonality cannot yet be ruled out there.
What the figures count. Every figure describes used-car listings created by dealers, counted in the month the listing first appeared. Not cars sold, and not the stock of cars currently for sale.
What the ratio is made of. Prices are asking prices, and the reference price is a list price. Neither end of the ratio is a transaction: a residual value of 65% is 65% of a price nobody paid, not a share of value retained.
Asking prices are an early signal of where sellers think the market is. Transaction data shows something different, and both are worth having.
Age, not condition. RV% here is set against age. It does not adjust for mileage, service history or condition, so two cars of the same age can sit far apart. These are market-level figures, not a valuation for a specific car.
Alongside this: Germany
Alongside this: GermanyWe publish a matching report for the German market. It has six months of data against twenty-one here, so it reads movement within its window rather than direction over time, and it has room to break residual value down by body type as well as by age.
It runs the same separation as section 1 here, on its own market and its own window. The two windows overlap but do not match, so the figures are not the same measurement of the same thing.
Both reports run the same fixed-age check, and each reports what its own market shows. Here it makes the rise larger, not smaller: the electric cars being listed in Denmark are not getting younger. What the German data shows on that check is stated in the German report.
What the German report cannot yet say is what this one says about ageing, because that needs the time span. We can measure it in Denmark.
Method
MethodSource | National used-car listing data |
Population | Dealer listings. The private market is effectively absent from this source |
Window | January 2025 – September 2026 · roughly 420,000 listings across four powertrains |
Lens | Listings created per month, not total inventory |
Price | Asking price. Neither end of the ratio is a transaction |
RV% | Asking price divided by list price when new, including registration tax |
Reference | The registered new price of that specific car at its own first registration. Historical, not today’s model price |
Age | Days from first registration, day precision |
Hybrid | One fuel code: mostly plug-in hybrids, with some conventional hybrids in it |
This report covers | Electric 53%, petrol 29%, diesel 10%, plug-in hybrid 8% in September 2026 |
VAT | Both ends of the ratio include Danish VAT at 25% |
Each figure | The middle car of its group, not an average price divided by an average new price |
Weighting | Median weighted by brand within each powertrain, against the full population |
Baseline | January–March 2025 = 100 |
Currency | Danish kroner |
Data as of | 19 October 2026 |
Extract | |
Method version | 6 |
On the reference price
It is the price that specific car cost when it was first registered, not what the model costs today. That is what makes the separation in section 1 possible: a car with a higher new price does not lift RV% simply by being expensive. It also means a change in today’s new-car prices or taxes cannot move these numbers backwards.
On the two series
The figures under the price charts and under the residual-value charts are not the same count, because the two series do not rest on the same listings. Every listing carries an asking price; only those with a reference price carry a residual value. The two are reported side by side and counted separately.
On VAT
Both the asking price and the reference price include VAT at the Danish rate of 25%, and both include registration tax. Because the rate is the same at both ends and constant through the window, it cancels out of the index entirely. It does not cancel out of the levels, which is one of the reasons the figures here cannot be laid alongside another country’s.
On the baseline
The index is set to 100 at January–March 2025, the window’s first three months. Three months rather than one, because a single powertrain in a single month can run as low as about 1,200 cars, and one month as the zero point would let one quiet month decide where every later figure sits. The whole baseline also falls before the January 2026 change in test method for plug-in hybrids, so the starting point is untouched by it.
On what plug-in hybrid covers
This series is built on a fuel code that is mostly, but not only, cars with a plug. The balance is not steady: in the 2–3 year band the share without a plug runs from 21% in the first month measured to 29% in the last, and later registration years carry more of them. Section 6 shows that this moves where the series sits without turning it. The balance also shifts with age, because Denmark had almost no plug-in cars before 2021. That is why section 4’s plug-in hybrid column is read for its level at a given age rather than down the column.
The German report’s hybrid line counts plug-in and conventional hybrids together as one category, so the two lines are not the same set of cars.
On weighting
New-price coverage varies by brand. Each month’s figure is therefore weighted so that every brand carries the share it holds in the full population, within each powertrain, rather than the share we happen to observe of it. Section 2 shows what that adjustment is worth in practice. No coverage threshold is applied: the skew between brands is handled by weighting by brand, not by leaving out brands or months with low coverage.
On comparing powertrains
Registration tax is in the reference price and differs sharply by powertrain. The gap between electric and petrol in the 2–3 year band is 15 points, but that comparison would measure the tax system, not the market. Development over time within each powertrain is where the tax structure cancels out, and that is what this report shows.
On the limits of the age bands
The fixed age bands in section 3 hold age constant by showing one band at a time rather than by modelling, and the drift inside each band is measured rather than assumed. Ten of the thirteen bands hold their age steady across the whole window, and the report’s argument rests on those ten. The three marked ◆ either thin out in their weakest month or shift enough inside their own band that “fixed” is approximate. They are shown anyway, because leaving them out would be a choice the reader could not see.
Residual values above 1.20 are set aside as implausible. Whether they are data errors or real cars has not been measured. No lower cut-off is applied, because a floor would remove cars that are genuinely listed at very low prices. Where a range is shown, the value depends on where the cut-off for implausibly high values is set. The range covers the cut-offs we consider defensible, from 1.10 to 1.30. It is not a confidence interval.
Differences are computed within each cut-off, from the figures as printed. Where either figure is a range, a difference can therefore be narrower than the distance between the two ranges.