01Start with the data class, not the headline
A credible market read begins by asking what each number actually counts. New vehicles sold, vehicles registered, active used-car advertisements, ownership transfers, and final sale prices are related—but they are not interchangeable.
This brief uses exact public figures where the source supports them and states what remains unknown. Carvaly's editorial standards define the same boundaries for every article: context from one data class is never presented as proof from another.
| Data class | What it can show | What it cannot establish alone |
|---|---|---|
| New-vehicle sales | New units reported sold in a stated period | Used-car volume, used prices, or resale strength |
| Registrations | One record per JPJ car-registration transaction in the cited data | A purchase or used sale, its date, or its price |
| Active used-car listings | Current seller asking positions and available supply | Final negotiated price or completed transaction |
| Completed transactions | A transfer or sale when the dataset defines it that way | Condition-adjusted value without vehicle detail |
| Final sale prices | Amounts actually paid when directly observed | A national index without representative coverage |
02May 2026: the new-vehicle backdrop
The Malaysian Automotive Association's May 2026 production and sales release reported 61,250 total new-vehicle sales for May, compared with 69,607 in May 2025. It reported 315,568 units year to date through May, compared with 319,846 over the same 2025 period.
| MAA measure | 2026 | Comparison | Reported change |
|---|---|---|---|
| Total sales · May | 61,250 | 69,607 · May 2025 | −12% |
| Total sales · Jan–May | 315,568 | 319,846 · Jan–May 2025 | −1% |
| Passenger vehicles · May | 57,360 | Part of May total | — |
| Commercial vehicles · May | 3,890 | Part of May total | — |
The same MAA release said May total industry volume was 15% below April after two months of growth and attributed the monthly decline mainly to long festive holidays. That is useful context for new-vehicle activity only; it does not show whether used-car prices rose, fell, or stayed flat.
03Registrations show registration mix, not resale strength
The official data.gov.my Car Popularity Explorer, last updated 10 July 2026 with data through 30 June 2026, showed 2026 registrations of 158,295 for Perodua, 98,010 for Proton, 50,591 for Toyota, and 27,549 for Honda. The underlying Car Registration Transactions catalogue documents the registration dataset.
| Marque | 2026 registrations through 30 June |
|---|---|
| Perodua | 158,295 |
| Proton | 98,010 |
| Toyota | 50,591 |
| Honda | 27,549 |
These counts describe registration transactions in the official data, not the number of used cars sold or the stock of vehicles currently in use. A higher count shows more registrations recorded for that marque during the selected period; it does not prove resale liquidity, depreciation, buyer preference, service-network reach, or a final sale price.
04What the public figures do not answer
Neither cited source is a national used-car transaction-price series. They do not provide the joined vehicle-level fields needed to compare like with like across condition, mileage, variant, repair history, ownership history, seller type, region, time on market, and final negotiated price.
Questions this brief cannot answer from the cited figures:
- How many used-car sales completed nationally in a month and what buyers finally paid.
- Whether a particular model retained more value after controlling for age, mileage, variant, and condition.
- Whether an advertised car sold, how long it took, or how far its final price moved from the asking price.
- Whether EV, hybrid, petrol, or diesel residual values strengthened or weakened in 2026.
- Whether one Malaysian region consistently clears a higher condition-adjusted sale price than another.
A third-party projection can answer a forecasting question only within its own model and assumptions. It should be labelled as a projection, not blended with observed official counts or presented as an actual 2026 transaction result.
05Electrification: registration evidence is not residual evidence
The registration catalogue can help analysts inspect registration activity by the available vehicle attributes and period. It does not include the complete condition, battery-health, warranty, asking-price, and completed-sale evidence needed to calculate EV or hybrid residual values.
For an individual electrified vehicle, battery-health documentation, remaining warranty, charging compatibility, repair history, exact variant, age, mileage, and current comparable listings may all be relevant. Their influence must be assessed from evidence about that vehicle and comparable market—not from a national registration count alone.
06How to use current asking-price evidence
Active listings can answer a narrower and useful question: what are sellers currently asking for comparable vehicles in the observable market? They cannot reveal an unobserved final price, and a single ambitious advertisement should never set the benchmark.
A defensible vehicle-level check should:
- Match make, model, year, variant, transmission, and region as closely as the available evidence allows.
- Compare mileage, condition, service evidence, accident or flood history, ownership, and relevant equipment.
- Use a group of recent comparables, remove clearly unusable records, and show when evidence is thin or widely spread.
- Keep the observation date and source class visible, then verify the physical car and documents independently.
Carvaly's public methodology explains how asking-price signals become a decision-support range, including freshness, comparable depth, confidence, and limitations. It does not claim that an asking-price range is an official appraisal or guaranteed sale price.
Frequently asked questions
Does this brief measure the size of Malaysia's used-car market?
No. The MAA figures are new-vehicle sales and the data.gov.my figures are registrations. Neither is a complete national count of used-car transactions or their final prices.
Does one registration equal one used-car sale?
No. Registration data records vehicles under the dataset's registration definition. It should not be relabelled as a used-car transaction without a source that explicitly establishes that event.
Do the registration counts prove which marque holds value best?
No. They describe registration mix. Retained value requires price evidence adjusted for age, mileage, variant, condition, history, location, and the point in time being measured.
Can asking listings be used to value a car?
They can provide current asking-price context when several close, recent comparables are cleaned and interpreted with their limits. They do not reveal the final negotiated sale price and should be combined with a physical and document check.