1. Summary

Q2 2026 was the quarter in which Tesla finally addressed the frozen new Model 3 market that closed our Q1 report. On May 1, Tesla began selling a Shanghai-built Model 3 Premium RWD in Canada at an effective on-road cost of roughly $47,000 in BC, replacing a US-built Long Range that tariffs had priced into effectively zero sales volume, after Ottawa cut tariffs on Chinese-built EVs from 100% to the standard 6.1% rate.

The obvious expectation was that used Model 3 values would come under pressure across the board. The actual outcome was somewhat unexpected and considerably more selective. Values on the newest, facelifted Model 3s were capped — those cars stopped appreciating entirely. But older pre-facelift Model 3s showed no sign of pressure at all, and in fact we observed their prices rising through the quarter.

Across the quarter, four of our five tracked segments closed higher than Q1, and the gains were not simply a matter of better cars coming to dealer secondary markets. Controlling for model year and mileage, Q2 traded 9–18% above Q1 depending on segment. What the new-car repricing did do is flatten the shape of the Model 3 curve: the newest used cars — the ones a shopper would genuinely weigh against a new one — stopped appreciating, while older cars serving an entirely different budget kept climbing.

Underneath the pricing, the shape of local supply changed. We observed a substantial drop in overall volume, close to half against Q1. Against that same backdrop, however, we saw a pronounced increase in used Teslas flowing into the market through Clutch — a rise in one channel set against a decline in the whole.

2. Key Market Metrics

The charts below show approximate median dealer market transaction prices versus median odometer reading for each drivetrain configuration, broken down by model year. Bubble size indicates relative transaction volume. Model years with too few observations to be meaningful are omitted, as are near-new 2025–2026 Model Y units, which are effectively new cars being remarketed rather than ordinary used inventory.

2.1 Model 3 — RWD

Includes SR+, Standard Range, and RWD variants | Q2's highest-volume segment

2.2 Model 3 — AWD / Long Range / Dual Motor

Includes LR AWD, Long Range, and Dual Motor variants | Thinner observation set

2.3 Model Y — RWD

Includes Standard Range RWD and RWD variants | Note: not every model year had a RWD option

2.4 Model Y — AWD / Long Range / Dual Motor

Includes LR AWD, Long Range, and Dual Motor variants | Q2's highest-volume Model Y segment

2.5 Model Y — Performance

Includes Performance AWD variants | Thinner observation set, shown for continuity with our Q1 format

3. Used Market Price Trends

Q2's pattern was calmer than Q1's whipsaw: a soft open in April, a sharp recovery through May, and most of that gain held into June. The table below shows where each tracked segment finished for the quarter as a whole against Q1.

3.1 Price Movement Summary

* Shown for continuity with our Q1 format, but drawn from thinner observation sets and not intended as trend indicators. Model 3 RWD and Model Y AWD are the two segments deep enough to read with confidence. All figures are aggregated observations of general market activity.

3.2 Looking Past Composition: A Like-for-Like Index

Median prices can rise simply because newer, lower-mileage cars came to market. Within Q2 that happened on mileage but not on age: the average pre-refresh Model Y AWD we observed held steady at a 2022 model year from April to June while its average odometer fell by roughly 7,500 km. To separate genuine price movement from a changing mix, we fit each segment's price against model year and mileage, then track how actual transactions sit against that baseline over time.

3.2.1 Like-for-Like Used Tesla Dealer Market Price Index (Observations)

Adjusted for model year and mileage | January 2026 = 100

The adjusted picture is consistent. Both segments bottomed in February, recovered through March and April, and kept climbing after the May 1 Model 3 repricing. Model 3 RWD finished June 21.5% above its January level and Model Y AWD 14.8% above. A second approach — fitting the price relationship on Q1 observations and applying it to Q2 observations — points the same direction with somewhat smaller magnitudes: Q2 traded 17.7% above the Q1 baseline for Model 3 RWD and 12.7% above for Model Y AWD. The two are not independent tests, since both regress price on model year and mileage over the same observations, but together they bracket the move at roughly 13–22%.

4. The New-Car Ceiling: Where It Actually Bites

4.1 What Changed on May 1

Canada's tariff on Chinese-built electric vehicles fell from 100% to the standard most-favoured-nation rate of 6.1%. Tesla responded on May 1 by offering a Shanghai-built Model 3 Premium RWD in Canada at an effective on-road cost of roughly $47,000 in BC. It replaced a US-built Long Range whose tariff-inflated price had left the new Model 3 selling in effectively zero volume — the situation we flagged at the end of our Q1 report. Model 3 Performance also came down substantially.

Supply matters as much as price here. Shanghai's ample capacity means new Model 3 delivery waits have been short — on the order of a few weeks. That stands in sharp contrast to the Model Y, which is built in Berlin for the Canadian market and where capacity constraints have stretched waits to several months. A buyer who wants a Tesla now, rather than in the autumn, has been steered toward the Model 3 by availability as well as by price.

One further detail matters more than the headline number. Because the Model 3 is built in China, it does not qualify for the federal $5,000 Electric Vehicle Affordability Program incentive, which requires manufacture in a country with a Canadian free-trade agreement. The Berlin-built Model Y RWD, at $49,990, does qualify. The two vehicles therefore land closer together on the road than their stickers suggest — but not as close as the stickers imply either.

Freight and PDI are excluded from the EVAP program's $50,000 transaction-value test, which is why the Model Y RWD qualifies. Following the treatment used in our Q1 report, the incentive is applied as a post-tax credit; if it were applied pre-tax the Model Y figure would be approximately $53,200.

4.2 The Ceiling Compresses the Curve Rather Than Lowering It

If a cheaper new car pulls used values down, the effect should be strongest on the cars closest to new. That is exactly the pattern we observed — and it is the only place the effect appears. The table below shows how much each model-year cohort gained from Q1 to Q2 after adjusting for mileage, expressed in percentage points relative to its own Q1 position.

The Model 3 column falls steadily from +28.9 points on 2019 cars to +4.1 points on 2024 cars. Model Y, whose new-car price did not change in Q2, shows no comparable slope. The newest cohorts rest on the thinnest observation sets in the table, so the exact figures carry wide error bars; the shape across six Model 3 cohorts is the more robust observation.

Absolute prices tell the same story. The 2024 Model 3 RWD median went from $39,050 in Q1 to $40,100 in Q2, up 2.7% unadjusted, while the Model 3 RWD segment as a whole rose 10.1%. The newest cars stood still while everything beneath them moved.

The reason is straightforward. A 2024 Model 3 RWD changed hands in the dealer market at a median of $40,100 against a new equivalent costing roughly $47,000 on the road — and that spread still has to absorb reconditioning, warranty risk and retail margin before a used buyer sees any saving at all. There is very little room left in it. By contrast, a 2024 Model Y RWD traded at a similar $40,500 against a new equivalent at roughly $53,800, leaving materially more space in the middle. That difference in remaining headroom is the clearest practical distinction between the two nameplates going into Q3, and it exists because only one of them had its new-car price cut.

4.3 Where the Pressure May Be Going Instead

One possibility we cannot yet test is that the ceiling is being absorbed by sellers rather than by prices. If wholesalers holding high-cost inventory decline low bids and let cars go unsold, the transactions that do complete are the ones sellers were willing to accept, and observed prices stay firm while liquidity thins. Our observations cover completed transactions only; we have no visibility into no-sale rates. We raise this as a hypothesis for Q3, not a finding.

What we can say is that the mix shifted. Model Y's share of the vehicles we observed declined noticeably over the course of Q2, from a clear plurality of activity early in the quarter to well under a third by the close. Whether that reflects sellers holding back or simply fewer cars entering the channel, we cannot yet separate.

5. Redesign vs. Hardware: What Dealer Buyers Pay For

The 2024 model year is a useful natural experiment. Model 3 gained both the Highland redesign and Tesla's HW4 (AI4) computer that year. Model Y gained HW4 as well, but kept its existing body — the refreshed Model Y arrived as a 2026 model-year car, first delivered to Canadian customers in the first half of 2025, and the examples we see locally are Berlin-built.

Worth noting on build origin: 2024 Model 3s in the local pool are a mix of Shanghai-built and US-built cars. Production for Canada shifted toward US-built Model 3 in the closing months of that model year, after tariffs on Chinese-built vehicles were introduced late in that year.

The 2024 comparison lets us price the two changes separately. Comparing consecutive model years at a common 50,000 km, ordinary annual steps in these segments run from roughly flat to about 13% — Model 3 RWD posted −0.2%, +1.6%, +3.9%, +12.0% and +7.1% across 2018 through 2023, and Model Y AWD posted +7.8%, +3.8% and +12.8% across 2020 through 2023.

5.1 Buyers Pay for the Redesign, Not the Computer

Mileage-adjusted price step between consecutive model years | Q1 + Q2 2026 pooled

Model Y's 2023→2024 step — pure hardware, same car — came in at +6.7%, sitting inside its own ordinary range and slightly below its median step of +7.8%. Model 3's 2023→2024 step, which bundles the redesign with the same hardware change, was +25.3%: double its largest ordinary step and more than six times its median.

Since the Model Y comparison isolates HW4 and finds nothing unusual, the roughly 19 points of excess on the Model 3 is attributable to what buyers can see and feel — the exterior, interior, acoustic glass and rear screen — not to the autonomy computer behind the windshield.

The timing sharpens the point. Roughly four million HW3 vehicles worldwide sat frozen on FSD v12.6 from early 2025 until June 29, 2026, when Tesla began rolling out FSD v14 "Lite" to early-access HW3 cars — a distillation of the HW4 v14 behaviour onto a computer with about 15% of AI4's effective memory bandwidth. Through all of Q2, HW4 cars therefore held a clear and well-publicised software advantage over HW3 cars. The secondary Tesla market has not priced it yet.

5.2 Why This May Not Hold — and Why It Might

On the new-car side, FSD adoption is climbing steeply. Tesla's Q2 2026 results put active FSD (Supervised) users at 1.48 million globally, up 56% year over year, with roughly 200,000 added during the quarter — the largest quarterly gain the company has reported. About 55% of North American deliveries in the quarter had FSD active at handover. This is a North American phenomenon specifically, since neither China nor Europe has seen a comparable broad rollout.

It would be easy to read that as a signal that FSD-capable hardware must eventually command a used-market premium. We would be cautious, for a reason specific to how Tesla now sells the software. Tesla stopped offering FSD as a one-time purchase on February 14, 2026, and it is now sold as a monthly subscription on Model 3 and Model Y — outright purchase survives only on new Model S, Model X and the top Cybertruck trim. That distinction matters enormously for resale: a one-time purchase is attached to the vehicle and travels with it to the next owner, whereas a subscription follows the account and transfers nothing. Essentially all of Q2's new activations on Model 3 and Model Y therefore add nothing to the future stock of used cars carrying FSD.

The practical consequence is the opposite of what the adoption figures suggest at first glance. The population of used Teslas with FSD permanently attached is now close to fixed, and it will not grow with new-car attach rates. Rather than becoming commonplace, those cars become a shrinking, differentiated subset of the used pool — which is an argument for them eventually being worth more, not less, but only for that narrow group rather than for HW4 cars generally.

Two further caveats cut in opposite directions. The absence of a hardware premium today may partly reflect limited buyer awareness rather than genuine indifference: many owners could not say which computer their car carries, and a market cannot price a distinction it does not perceive. As FSD v14 Lite reaches HW3 cars at scale, that could change in either direction — wider exposure may make buyers more conscious of the generational gap and push them toward HW4, or it may narrow the functional difference enough that the gap stops mattering.

The upgrade path is also unsettled, and it moved during the quarter. On the Q1 2026 earnings call Tesla indicated that customers who had purchased FSD would receive free upgrades to HW4 or HW4.5 if required. On the Q2 2026 call the position shifted: rather than retrofitting HW4, Tesla suggested it may skip ahead to a next-generation computer, with an intermediate board and the AI5 processor potentially entering production around the middle of 2027. Tesla has separately confirmed that HW3 cannot support Unsupervised FSD. For a used buyer weighing an HW3 car today, the honest summary is that the retrofit question has no settled answer.

Our finding that hardware carries no meaningful premium describes this market as it stood in Q2 2026. It is not a safe assumption about 2027 and beyond, and we will keep testing it each quarter.

6. Channel Shift and What It Implies

The most striking non-price development of the quarter was where cars traded rather than what they traded for. Overall observed volume fell by close to half against Q1 — and yet, within that shrinking total, one route to market grew conspicuously.

Our read is that this reflects a change in where consumers are selling their cars, not simply a shrinking market. A growing share of owners are selling directly to online instant-offer buyers rather than trading in at a dealership or selling privately — and Clutch is the name we see most often behind that shift locally. Vehicles acquired that way are then remarketed into the dealer buying channel, which is where they enter our observations.

Alongside this, we observed a noticeable increase in used Teslas listed on Tesla's own website, adding another route through which pre-owned inventory reaches buyers without passing through an independent dealer at all.

Alongside this, conversations with people working at traditional new-car dealerships point to a related shift on the new-vehicle side. Their read of customer feedback is that softening new-car sales at legacy internal-combustion brands is tied directly to Tesla's continued strength in BC — buyers who would previously have replaced one gasoline car with another are leaving the category altogether. We pass this along as industry conversation rather than as something our own observations can measure.

7. Q3 2026 Factor E Outlook

Whether the Ceiling Finally Bites

The central question for Q3 is whether the compression that pinned 2024 Model 3 values spreads downward into the 2022–2023 cohorts as the cheaper new car becomes more widely available. With so little room left between a 2024 Model 3 and a new one, that cohort has no space to appreciate and some space to fall. We will also be watching whether no-sale rates rise, which is the form the pressure would take if sellers keep resisting lower bids.

Energy Prices Still in Play

Metro Vancouver pump prices held near historic highs through Q2 with little of the relief that easing crude prices might have implied. As long as local fuel costs stay near current levels, the running-cost case for an EV remains in front of consumers regardless of what the crude benchmark does. A genuine break in BC pump prices would be a more meaningful signal for used EV demand than any move in the crude price itself, and given how quickly the geopolitical picture has been shifting, we would treat any forecast here with caution.

Tesla Product Moves

Tesla opened US orders for the three-row Model Y L in early July at US$61,990 — which converts to roughly $100,000 on the road in BC once exchange, freight and tax are applied, though Canadian timing and pricing have not been announced and the US configurator shows October–November 2026 delivery. No near-term used-market impact is likely. More relevant for Q3 is the continued arrival of refreshed Berlin-built Model Y units into the local pre-owned pool, which will begin to give the Model Y a top-of-curve dynamic comparable to the one the Model 3 developed in Q2. Whether Berlin capacity loosens enough to shorten Canadian Model Y waits is worth watching, since delivery timing has been doing real work in steering buyers between the two models. Tesla also delivered its strongest second quarter ever globally — 480,126 vehicles, up 25% year over year — which sustains trade-in supply into the back half of the year.

FSD v14 Lite on HW3, and the Retrofit Question

The v14 Lite rollout to HW3 cars began too late in the quarter to affect Q2 pricing, and its effect on resale values remains to be seen. Given that the market assigned essentially no premium to HW4 even while HW3 was a full generation behind, we would not expect an immediate repricing. Two developments make this worth watching closely rather than assuming continuity, and they pull in opposite directions: wider exposure to v14 could either narrow the perceived gap between the two computers or make buyers newly aware that a gap exists at all. Meanwhile Tesla's own upgrade messaging moved during the quarter, away from retrofitting HW4 and toward a possible next-generation computer around the middle of 2027. Until that path is clearer, we would treat any HW3 retrofit expectation as unpriced rather than promised. Section 5.2 sets out the reasoning in full.

Macro & Competitive Factors

Bank of Canada rate decisions continue to shape both consumer financing costs and dealer floorplan economics. Chinese EV brands preparing for Canadian entry under the revised tariff arrangement remain a second-half consideration; the Model 3 repricing is the first concrete evidence of how quickly that trade shift can reprice a segment.