01The ten worst over three years
Most of this series measures value at five years, because that is where a used-car market forms. Three years is a different and more personal question: it is the point at which many owners first consider changing, and the point at which what they owe and what the car is worth are furthest apart.
65.9%
Worst three-year retention
MINI Countryman, on 740 listings across 40 cohorts
RM104,018
Largest three-year loss in ringgit
BMW X5, which retains a larger share than four cars above it
81.4%
Median three-year retention
Across all thirty corroborated models
| # | Model | 3-year retention | Lost in 3 years | Near-new ask | 5-year retention | Listings |
|---|---|---|---|---|---|---|
| 1 | MINI Countryman | 65.9% | RM64,669 | RM189,644 | 59.9% | 740 |
| 2 | BMW X3 | 67.5% | RM81,503 | RM250,780 | 48.6% | 877 |
| 3 | BMW 530i | 68.1% | RM91,264 | RM286,095 | 54.0% | 867 |
| 4 | Nissan Serena | 69.1% | RM33,935 | RM109,823 | 59.2% | 1,888 |
| 5 | Nissan Almera | 69.5% | RM20,881 | RM68,461 | 59.0% | 1,374 |
| 6 | BMW X5 | 71.1% | RM104,018 | RM359,925 | 48.1% | 1,152 |
| 7 | Volvo XC90 | 74.2% | RM76,788 | RM297,627 | 49.2% | 416 |
| 8 | Perodua Aruz | 74.3% | RM14,617 | RM56,877 | 71.4% | 1,295 |
| 9 | Mazda 3 | 74.5% | RM32,134 | RM126,017 | 67.1% | 1,284 |
| 10 | Proton Saga | 74.8% | RM7,789 | RM30,909 | 70.5% | 3,530 |
02The three-year ranking is a different ranking
If the three-year and five-year tables ordered cars the same way, this article would not need to exist. They do not. Rank agreement across the thirty corroborated models is Spearman 0.672 — related, but far from interchangeable.
| Model | 3-year retention | Rank at 3 years | 5-year retention | Rank at 5 years | Places moved |
|---|---|---|---|---|---|
| Toyota Alphard | 93.0% | 1st | 62.2% | 20th | −19 |
| Perodua Aruz | 74.3% | 23rd | 71.4% | 10th | +13 |
| Lexus RX350 | 82.0% | 13th | 57.8% | 24th | −11 |
| Volkswagen Golf | 90.0% | 3rd | 67.7% | 14th | −11 |
| Proton Saga | 74.8% | 21st | 70.5% | 11th | +10 |
| Honda City | 82.6% | 11th | 77.2% | 2nd | +9 |
| MINI Countryman | 65.9% | 30th | 59.9% | 21st | +9 |
The pattern is that a strong three-year figure is not evidence of a strong five-year one. Cars supplied through import channels and large Continental models both hold up unusually well to three years and then fall, while several national models do the opposite — an early, gentle loss followed by a long flat stretch.
For a buyer this cuts both ways. If you intend to keep a car for a decade, the five-year and ten-year behaviour is what matters and a flattering three-year number is noise. If you expect to change at three, this is your table and the headline one is not.
03Why three years is the dangerous point on a Malaysian loan
The three-year mark matters here for a reason specific to how Malaysians buy cars. Our hire-purchase terms are long, and a long term repays principal slowly at the start.
Malaysian hire purchase is governed by the Hire-Purchase Act 1967, with tenures commonly written over five, seven or nine years. The Hire-Purchase (Amendment) Act 2026 took effect on 1 June 2026 and banks have introduced early-settlement changes alongside it, but the underlying arithmetic has not changed: on a long term, three years in, you have paid a modest share of the principal.
Set that against a median three-year retention of 81.4%. For most owners the car has lost about a fifth of its value while the loan has not fallen by a similar proportion — and for the models at the top of this table, which have lost a third, the gap is far wider. That gap is negative equity, and it is the reason a three-year trade-in so often feels worse than expected.
The ringgit column tells a different story again
The BMW X5 sits sixth on percentage but first on money: RM104,018 gone in three years, against RM64,669 for the Countryman that tops the percentage table. At the other end the Proton Saga appears in the bottom ten on percentage while losing RM7,789 — less over three years than the X5 loses in three months.
04How we measured it, and what we will not claim
Three-year retention is read from the same fitted curve as the rest of this series: a decay rate estimated across model-year cohorts inside one market snapshot. This is editorial analysis of asking-price patterns, not how a Carvaly valuation is calculated. Understanding your Carvaly valuation explains how to read a result.
Two limits worth stating plainly:
- Only the five-year figure is corroborated by a second method. Our independent price-ladder check compares cars aged 0–1 against cars aged 4–6, which cannot validate a three-year number. Every model here cleared that five-year check, so we are decomposing curves that have already survived scrutiny — but the three-year column itself rests on one method.
- We do not publish a first-year figure at all. Our data is organised by model year, which means it cannot resolve inside the first twelve months: only one model-year cohort can ever sit in that window. Any 'first-year depreciation' number derived from cohort data of this shape is borrowed from the following period rather than measured, so we leave it out.
The second point is the reason this article starts at three years rather than one. It is a real limit of measuring depreciation from listing cohorts, and we would rather name it than publish a number that looks precise and is not.
05What to actually do with this
The useful action here is matching the horizon to your plan before you fall in love with a number.
- Decide your horizon first. If you change cars every three years, use this table. If you keep them, use the five-year ranking. They disagree often enough to matter.
- Check your equity position, not the percentage. Value at three years minus outstanding balance is the number that decides whether you can change cars at all.
- Be suspicious of a strong three-year figure on an expensive car. The Alphard, the Golf and the RX350 all look excellent at three years and materially worse at five.
- If you are buying used, three years old is rarely the bargain. Most of the models here still hold three-quarters of their value at that age. The discount arrives later.
- In every case: the model-level figure is a starting point. Year, variant, mileage, service history and condition decide the number that actually matters.
Frequently asked questions
Which car has the worst 3-year depreciation in Malaysia?
In our 14 August 2026 analysis the MINI Countryman retained the least at three years, at an estimated 65.9%, backed by 740 listings. The BMW X3 (67.5%) and BMW 530i (68.1%) followed. In ringgit the order differs: the BMW X5 loses the most over three years at about RM104,018, despite retaining a larger share than four models above it.
How much value does a car lose in 3 years in Malaysia?
The median across our thirty corroborated models is 81.4% retention at three years — so roughly a fifth of the value is gone. The range is wide: the Toyota Alphard retains 93.0% and the MINI Countryman 65.9%. Cheaper national models cluster in the mid-80s to low-90s.
Is the 3-year ranking the same as the 5-year ranking?
No, and that is the main finding here. Rank agreement between the two horizons is Spearman 0.672 across our thirty corroborated models. The Toyota Alphard is first at three years and twentieth at five; the Perodua Aruz is twenty-third at three years and tenth at five. Match the horizon to how long you actually intend to keep the car.
Why is three years a risky point to sell a car in Malaysia?
Because Malaysian hire-purchase terms commonly run five, seven or nine years, and a long term repays principal slowly at the start. Set that against a median three-year retention of 81.4% and many owners find the car is worth less than the outstanding balance. The gap is widest on the models at the top of this table, which have lost about a third of their value by then.
Why does this article not cover first-year depreciation?
Because our data cannot measure it. Listing cohorts are organised by model year, so at most one cohort can ever fall inside the first twelve months, and a rate fitted from a single point is not a measurement. Any first-year figure produced from data of this shape is borrowed from the following period, so we do not publish one.
Sources and references
- Carvaly: Understanding your Carvaly valuation
- Carvaly Market Intelligence: Malaysia Used-Car Market Report · July 2026
- Consumer Credit Commission Malaysia (SKP): Hire-Purchase (Amendment) Act 2026 Takes Effect on 1 June 2026
- The Association of Banks in Malaysia (ABM): Banks Introduce Changes to Hire-Purchase Financing
- Ministry of Domestic Trade and Cost of Living (KPDN): Hire-Purchase Act 1967 (Act 212)
- data.gov.my: Car Registration Transactions