An EV Launch Looks Like One Date. The Federal Record Shows Five.

A vehicle reveal is a scheduled media event. It is not a regulatory event, and it is not a revenue event. Between the moment a new model is shown and the moment the first unit is handed to a retail buyer, a sequence of federal filings has to be completed, and most of those filings leave a dated public record. Anyone treating the reveal as the tradable moment is trading the one point in the sequence that carries no verification requirement at all.

This matters more for electric models than for conventional ones, because the electric segment attracts announcement coverage far out of proportion to the volume it moves. The reveal generates the headline. The filings generate the evidence. They rarely happen on the same day, and the gap between them is where the observable information lives.

What follows is a description of the U.S. federal sequence as the regulations currently read, with citations, and a description of what each stage does and does not tell an observer. It is a framework for reading a launch, not a recommendation about any company or security.

The reveal is the least regulated moment in the sequence

Nothing in federal motor vehicle law requires a manufacturer to have completed anything before showing a vehicle to the press. Specifications shown at a reveal are not certified figures. Range claims made at a reveal are not label figures. A delivery date announced at a reveal is a commercial statement, not a regulatory milestone.

The regulatory milestones come later, and they are specific. Three of them sit with the National Highway Traffic Safety Administration, one with the Environmental Protection Agency, and one is a publication step that follows from the EPA filing. Each has its own statutory timing, and the timings are not aligned with each other.

Five records stand between the reveal and the first delivery

The sequence below is drawn from the regulations themselves. It is worth reading the timing language closely, because the deadlines are anchored to different events: one to the first offer for sale, one to the start of manufacture, one to introduction into commerce.

Three details in that timeline carry more weight than the rest.

The VIN filing is the earliest fixed deadline. Under 49 CFR 565.16(d), VIN deciphering information must be submitted to NHTSA "at least 60 days prior to offering for sale the first vehicle identified by a VIN containing that information." The regulation provides an alternative: if the manufacturer does not have sufficient vehicle characteristic information by then, submission is due within one week after that information first becomes available. That alternative is the escape hatch, and its existence is a reminder that a 60-day lead is a floor with a documented exception, not a guarantee.

The manufacturer identification filing is backward-looking. Under 49 CFR 566.6, a manufacturer beginning production of a new vehicle type submits identifying information "not later than 30 days after he begins manufacture." That deadline runs after production has already started. It confirms that manufacture began; it does not signal that it is about to.

The EPA certificate is the only true gate. Under 40 CFR 86.1848-10, a manufacturer "must obtain a certificate of conformity covering such vehicles from the Administrator prior to selling, offering for sale, introducing into commerce, delivering for introduction into commerce, or importing into the United States the new vehicle." A certificate is issued "for a period not to exceed one model year." Without it, lawful sale does not happen.

Only one of these gates is opened by a government agency

The most common misreading of U.S. vehicle regulation is the assumption that a federal agency inspects and approves new models before they go on sale. For emissions, something close to that is true. For safety, it is not true at all.

NHTSA's own guidance for new manufacturers states the position plainly: the agency "does not issue type approval certifications and does not certify any motor vehicles or motor vehicle equipment as complying with applicable FMVSS." Compliance with the Federal Motor Vehicle Safety Standards is self-certified. The manufacturer runs the testing, the manufacturer signs the statement, and the manufacturer affixes it to the vehicle.

That statement has fixed wording. Under 49 CFR 567.4, a passenger car label must read: "This vehicle conforms to all applicable Federal motor vehicle safety, bumper, and theft prevention standards in effect on the date of manufacture shown above." The label must also carry the date of manufacture, the gross vehicle weight rating, the gross axle weight ratings, the VIN, and the vehicle type classification.

The practical consequence for an observer is a difference in observability, not a difference in rigor. The EPA certificate exists as a dated entry in a federal database before a vehicle can be sold. The safety certification exists as a manufacturer's own statement that becomes visible on the vehicle and in NHTSA filings clustered around the start of manufacture. One is a leading record. The other is roughly coincident.

The range number on the window sticker is derived, not measured

The single figure most likely to be quoted from a launch is the range number. It is worth understanding how that number is produced, because its construction explains why it can differ from both the reveal claim and from owner experience.

The label itself is mandatory and timed. Under 49 CFR 575.401, a manufacturer "must affix or cause to be affixed and each dealer must maintain or cause to be maintained on each passenger car or light truck a label," and it must be there "prior to being offered for sale." For an electric model the label carries driving range, an energy consumption figure, and a combined value marked MPGe. The regulation also requires the label to display the web address fueleconomy.gov alongside the line "Calculate personalized estimates and compare vehicles."

The numbers on that label come from testing that the manufacturer runs. EPA states that "it is the manufacturers that are responsible for providing the fuel economy data used for labeling," and that the agency confirms roughly 15 percent of test results itself at its National Vehicle and Fuel Emissions Laboratory. Vehicles are run over five defined cycles: city, highway, a higher-speed aggressive cycle, a 95°F test with air conditioning, and a 20°F test with heating and defrost.

For an electric vehicle, range is established by running a fully charged vehicle over the city cycle until the battery is depleted, then repeating for the highway cycle on a laboratory dynamometer. Laboratory results are then scaled by an adjustment factor of 0.7 to approximate real-world conditions — a vehicle achieving 200 miles on the highway laboratory test yields 200 × 0.7 = 140 miles of adjusted highway range. Adjusted city and highway results are combined at a 55 percent city / 45 percent highway weighting to produce the label figure.

Three implications follow. A label range is a weighted, adjusted composite, so it is not comparable to a manufacturer's unadjusted engineering claim. Because 85 percent of results are accepted as submitted, the label is a compliance filing rather than an independent measurement. And because the label must be affixed before the vehicle is offered for sale, the appearance of a certified label figure is a firmer marker of market readiness than any reveal-stage specification.

The tax-credit gate that no longer exists

Between 2023 and 2025, a further gate governed the economics of an electric launch in the United States: eligibility for the clean vehicle tax credits, which turned on conditions including final assembly location and battery sourcing, and which produced a published list of qualifying vehicles. Reading a launch through that list was a defensible exercise during that window.

It is no longer the current framework. Public Law 119-21, enacted July 4, 2025, accelerated the termination of these provisions. The IRS states that the New Clean Vehicle Credit (section 30D), the Previously-Owned Clean Vehicle Credit (section 25E), and the Qualified Commercial Clean Vehicle Credit (section 45W) are unavailable for vehicles acquired after September 30, 2025. The agency also defines the term: a vehicle is "acquired as of the date a written binding contract is entered into and a payment has been made," where a payment "includes a nominal downpayment or a vehicle trade-in." A vehicle acquired on or before that date but placed in service later could still qualify; a vehicle acquired after it cannot.

This is the most important thing to check before reusing any launch framework written during that window. A model-eligibility list that was a live constraint through much of 2025 is not a live constraint for vehicles acquired afterward, and analysis that still leans on it is analysing a rule that has lapsed. It is also a general caution: federal agency web pages describing a credit do not all update on the same schedule, and the administering agency's guidance is the one that governs.

Model counts move slower than headlines

One way to see how loosely announcement volume tracks market structure is to look at how many distinct models are offered per model year. The Department of Energy publishes this count by technology.

Source: U.S. Department of Energy, Alternative Fuels Data Center, dataset 10303, "Light-Duty AFV, HEV, and Diesel Model Offerings, by Technology/Fuel." Counts derived by NLR using NHTSA Vehicle Product Information Catalog (vPIC) and registration counts from Experian Solutions.

The series is not smooth. All-electric offerings stood at 33 for model year 2016 and 34 for 2017, fell to 17 for 2018, and did not exceed the 2017 level again until 2022, at 32. The count then rose to 41 for 2023 and 61 for 2024, and stands at 44 for 2025 in the published table. Plug-in hybrid offerings follow a different shape entirely: 51 for 2020, 36 for 2022, 54 for 2024, and 32 for 2025.

Two cautions about reading it. First, this is a count of models offered, not units sold; a model with negligible volume counts the same as a high-volume one. Second, the source note states the counts are "derived by NLR using NHTSA Vehicle Product Information Catalog (vPIC) and registration counts from Experian Solutions" — the figures are derived from registration data rather than being a direct tally of certificates issued, so the most recent model year is the one most exposed to later revision. A single year's move in this series is weak evidence of anything. The multi-year shape is the usable part.

What Would Invalidate This

This framework rests on assumptions that can fail, and several of them are worth stating explicitly.

• If the regulations change. Every deadline cited here is current regulatory text, and regulatory text is revised. The clean vehicle credits demonstrate exactly this: a framework built on a published eligibility list became obsolete on a specific date by statute. Anyone reusing this sequence should re-read the cited sections rather than trusting this summary.
• If the filings are not observable in time. A certificate of conformity must exist before sale, but the timing between issuance and its appearance in a public dataset is a data-publication question, not a regulatory one. EPA's interactive certificate report updates daily and the downloadable spreadsheets update quarterly, so the lag depends entirely on which source is being watched.
• If the information is already reflected. These are public filings. Any observation available from a daily-updated federal database is available to everyone reading that database. Treating a public filing as private information is the error this framework is most likely to invite.
• If the launch is not the driver. A single model launch is one line item inside a diversified manufacturer. For a large automaker, regulatory progress on one model may be immaterial next to input costs, tariffs, labour agreements, or credit conditions. The framework is more informative the narrower the issuer's product base, and correspondingly less informative the broader it is.
• If deliveries are constrained downstream. Clearing every federal gate establishes that a vehicle may lawfully be sold. It establishes nothing about production rate, component supply, logistics, or dealer allocation. The gates are necessary conditions, not sufficient ones.

Concrete Framework

A repeatable way to read a launch as a sequence rather than a date:

1. Write down the reveal date and mark it as unverified. Record the claimed range, the claimed delivery window, and the claimed price separately from anything certified. These are the figures that will later be compared against label values.
2. Identify the model year the vehicle is being certified under. A certificate of conformity covers a period not to exceed one model year. A model-year boundary is a re-certification boundary, and it constrains how long a given certificate can carry a product.
3. Check whether an EPA certificate exists. This is the only gate a federal agency actively opens. Use the interactive certificate report for currency and note that the downloadable spreadsheets lag it by up to a quarter.
4. Check whether label values have been published. The label must be affixed prior to the vehicle being offered for sale, and it carries range, MPGe, and consumption. Compare the label range against the reveal claim and record the difference rather than explaining it away.
5. Apply the derivation before comparing numbers. Label range is laboratory results scaled by 0.7 and weighted 55 percent city to 45 percent highway. An engineering claim quoted without that treatment is not the same quantity as a label figure.
6. Verify the current status of any incentive assumption. Do not carry a tax-credit assumption forward from an older note. Confirm against the administering agency's current guidance, not a secondary page.
7. Size the position to the gap, not to the headline. If the thesis is that the interval between certification and delivery is mispriced, then the interval is the exposure. Define in advance which specific filing would confirm the thesis and which would end it, and set the exit at that filing rather than at a price level chosen afterward.
8. Log what each stage resolved in practice. After delivery begins, record which of the five records moved the market and which passed unremarked. Over several launches this is what separates a sequence that carries information from a sequence that only looks like it should.

The underlying point is narrow. A launch is not one event with one date. It is a chain of dated, citable, public filings, each with a different deadline anchor and a different degree of visibility, and the headline sits at the front of that chain where the verification requirement is weakest.