H1 2026 COE Review: Results from January Through June
Compare the full first half of 2026 using exercise ranges and averages, with clear distinctions between premiums, quota and registration policy.
A first-half review must cover January through June. The earlier version relied heavily on first-quarter figures while presenting them as H1. The table below uses the complete six-month interval and reports how many exercises are represented.
| Category | Exercises | Minimum | Mean premium | Maximum |
|---|---|---|---|---|
| A | 12 | S$102,009 | S$115,361 | S$126,009 |
| B | 12 | S$105,001 | S$119,535 | S$129,501 |
| C | 12 | S$74,801 | S$82,059 | S$94,000 |
| D | 12 | S$7,989 | S$9,203 | S$10,000 |
| E | 12 | S$112,890 | S$122,208 | S$130,000 |
Arithmetic mean across available exercises; no weighting by quota. Missing exercises are not invented. This table updates when a source record is corrected.
Read the mean and range together
The mean is the arithmetic average of available exercise premiums, without weighting by quota. The minimum and maximum show the observed range. Neither is the amount every buyer paid for a complete car, and the mean should not be labelled a volume-weighted market price.
An opening-to-closing change answers a different question. If quoting one, identify both exercises instead of subtracting a quarterly average from a single June result.
Supply announcements use different windows
February–April and May–July are LTA allocation periods. Each overlaps a different part of H1. The May–July total rose to 19,052 from 18,824, but A’s allocation fell while B’s rose. That announcement alone does not show that all first-half buyers experienced more supply.
Policy affected different groups differently
The revised PARF schedule applies by original COE cohort. Existing cars retained the relevant older schedule; there was no universal retrospective reduction. EV registration incentives also use their announced effective dates and should not be described as benefits available to every car.
What the half-year does not predict
Six months of outcomes do not establish a permanent premium floor. Earlier claims assigning a probability to lower H2 prices lacked a documented model and validation sample. They have been removed.
A scenario can still be useful: test a higher or lower complete purchase price against your deposit, monthly budget and waiting costs. Label that as sensitivity analysis rather than the most likely market outcome.
Use the correct tool for the decision
New buyers need an actual package and its COE terms. Renewing owners need the relevant monthly PQP. Sellers need a vehicle-specific offer and rebate calculation. H1 statistics provide context for those decisions; they do not replace the current inputs.
Sources and review date
Reviewed on 13 September 2026. Historical results retain their exercise dates; worked budgets are assumptions, not quotations.