01What a predictable price actually looks like
Every buyer wants to know whether the price in front of them is fair. That question has a very different answer depending on the car, and not for the reason most people assume — it is not about the brand's honesty, it is about how much the cars within a single model year actually differ from each other.
17.9%
Market-wide price spread
Middle half of listings within one model and model year
9.6%
Spread on cars under three years old
Tight enough that the asking price is a real guide
27.5%
Spread on cars over fifteen
Wide enough that the asking price tells you little
We took every combination of make, model and model year with at least twenty listings — 2,493 cohorts covering 311,642 listings — and measured how far apart the twenty-fifth and seventy-fifth percentile asking prices sit, as a share of the median. That interquartile spread is the honest answer to 'how much do identical cars vary'.
| Example | Median asking price | Middle half of listings | Spread | Listings |
|---|---|---|---|---|
| Toyota Vios, 2024 | RM72,000 | RM69,888 – RM75,800 | 8% | 277 |
| Perodua Myvi, 2011 | RM14,500 | RM12,000 – RM16,800 | 33% | 561 |
| Proton Saga, 2016 | RM12,800 | RM9,886 – RM16,000 | 48% | 308 |
02Predictability collapses with age
The single strongest pattern in this data is age. It is not price level, not brand and not country of origin — it is simply how long the cars have been on the road accumulating differences from one another.
| Age of the cars | Cohorts measured | Listings | Typical price spread |
|---|---|---|---|
| Under 3 years | 259 | 27,929 | 9.6% |
| 3 to 6 years | 470 | 81,590 | 13.6% |
| 6 to 10 years | 611 | 98,901 | 17.0% |
| 10 to 15 years | 649 | 70,193 | 19.7% |
| Over 15 years | 504 | 33,029 | 27.5% |
The mechanism is straightforward. A two-year-old car has had two years to diverge from its identical siblings: similar mileage, similar condition, mostly still under warranty. A fifteen-year-old car has had fifteen years, and one may have 90,000 km and a full service record while another has 300,000 km and a repaired front end. The badge is the same; the cars are not.
The same nameplate, tracked across its own life
This shows up cleanly within a single model. The Toyota Vios is priced within 8% at two years old, 13% at five, 14% at ten, 20% at fifteen and 24% at twenty. The Perodua Myvi runs 10%, 15%, 23% and 33% across the same milestones. The car does not become less popular; it becomes less uniform.
03The models at each end
Among the 127 models with at least five model-year cohorts and 500 listings, the extremes are informative — though most of what separates them is how recently the model went on sale.
| Most predictable models | Typical spread | Median price | Listings |
|---|---|---|---|
| Honda Civic Type R | 8.0% | RM216,900 | 1,735 |
| Toyota Corolla Cross | 9.4% | RM103,269 | 1,421 |
| Mazda CX-30 | 10.2% | RM96,000 | 1,142 |
| Mazda CX-3 | 10.4% | RM61,800 | 960 |
| Mitsubishi Xpander | 10.6% | RM66,400 | 900 |
| Perodua Ativa | 10.9% | RM54,800 | 2,535 |
| Toyota Yaris | 11.2% | RM61,750 | 2,524 |
| Perodua Aruz | 11.9% | RM53,900 | 2,197 |
| Least predictable models | Typical spread | Median price | Listings |
|---|---|---|---|
| Land Rover Range Rover | 50.8% | RM283,800 | 931 |
| Volkswagen Golf | 43.1% | RM86,688 | 1,601 |
| Perodua Viva | 40.4% | RM7,950 | 1,664 |
| Proton Wira | 39.5% | RM4,550 | 1,106 |
| Proton Waja | 39.3% | RM4,919 | 646 |
| Proton Preve | 39.0% | RM14,811 | 1,432 |
| Perodua Kelisa | 35.7% | RM9,550 | 799 |
| Proton Saga | 34.0% | RM13,844 | 6,393 |
Two different causes produce a wide spread, and it is worth telling them apart. The Perodua Viva, Proton Wira and Proton Waja are cheap, old and long out of production: their spread is condition. The Land Rover Range Rover and Volkswagen Golf are wide because one nameplate covers many very different variants and engine sizes, so cars that look identical in a listing title are not comparable at all.
The Proton Saga is a special case worth naming. It has the widest spread of any high-volume Malaysian model at 34.0%, and that is because it is the one nameplate still selling in numbers today that also has twenty model years of surviving cars on the road. It spans both causes at once.
04How to use this when you are buying
The practical translation is a rule about how much weight to put on the asking price versus how much to put on the inspection.
- On a car under three years old, the asking price is close to the answer. With a typical spread under 10%, a listing meaningfully above the cohort median is genuinely expensive rather than merely different.
- On a car over ten years old, treat the asking price as an opening claim. A spread near 20% to 30% means the range of legitimate prices is very wide, and where a specific car sits inside it is decided by evidence you have to go and gather.
- Beware of nameplates that span many variants. A Golf or a Range Rover listing tells you much less than a Myvi listing, because the badge covers vehicles with very different specifications and values.
- Compare within the model year, not the model. Almost all of the spread in this analysis is within a single year. Comparing a 2016 car against 2019 prices will mislead you far more than any negotiating error.
- Use the spread as a sanity check. If the car you are looking at sits outside the middle half of comparable listings, there is a reason. It may be a bargain or it may be a problem, but it is not nothing.
Frequently asked questions
How much do used car prices vary for the same model in Malaysia?
Across 2,493 model-year cohorts covering 311,642 listings, the middle half of asking prices within a single model and model year spans a median 17.9% of the median price. That figure hides a strong age effect: it is 9.6% on cars under three years old and 27.5% on cars over fifteen.
Which used cars have the most predictable prices in Malaysia?
Recently introduced, high-volume models. The Honda Civic Type R came out tightest at 8.0%, followed by the Toyota Corolla Cross at 9.4%, the Mazda CX-30 at 10.2% and the Mazda CX-3 at 10.4%. Almost every model at this end of the table went on sale within the past decade, so its cars have not yet had time to diverge from one another.
Why do older cars have such different prices for the same model?
Because time makes identical cars different. A fifteen-year-old fleet contains cars with 90,000 km and full service records alongside cars with 300,000 km and accident repairs, and the listing title cannot tell them apart. In our data a 2016 Proton Saga's middle half of asking prices runs from RM9,886 to RM16,000 — the top of that range is 62% above the bottom for the same model year.
Does a wide price range mean sellers are overcharging?
Not necessarily. A wide spread mostly means the cars genuinely differ — in condition, mileage, history and specification — rather than that some sellers are dishonest. It does mean the asking price carries less information, so the burden shifts onto inspection and on comparing against genuinely similar cars rather than against the model average.
Should I worry more about price on a new or an old used car?
Counter-intuitively, the old one. On a nearly-new car the market has already narrowed the price to within about 10%, so there is limited room to get it badly wrong. On a car over ten years old the legitimate range is two to three times wider, and placing a specific car inside that range requires evidence rather than a rule of thumb.