Methodology

How Evident computes the monthly total cost of ownership (TCO). All figures are estimates with an effective date — no guarantee. Every number is either an official value, a sourced market figure, or a clearly labelled model — and we say which.

List prices: gross list price (UVP incl. 19% VAT, excluding delivery/registration fees) — treated identically across all makes, including Tesla (fees noted in the source). Checked prices carry "price checked on …" with a source; unchecked ones are labelled "price unconfirmed" and enter no ranking.

Depreciation (the biggest item): where published 3-year residuals exist we use them as anchors (table below, with source and mileage basis). All other models use a segment approximation labelled as an estimate. In parallel we are building our own residual model on weekly-collected market data — it replaces the approximation once it demonstrably beats the anchors.

Battery degradation: fleet studies show modest degradation (Arval, 24,000 certificates: ~93% capacity at 70,000 km) — so battery health is a minor factor in our used-car residuals.

Energy: electricity or fuel from consumption × price, weighted by your charging mix (home/public). BEV real-world consumption comes from the gap model (per segment, fixed mild/−10 °C reference scenarios, DWD-weighted annual factor — details below). PHEV: real fuel use ≈ 3× WLTP for private usage (ICCT/Fraunhofer ISI 2022, n≈9,000; EU OBFCM fleet figure 3.5×) — the measured net including electric share; charged electricity is not billed separately here (conservative, in favour of the PHEV).

Real-world range: the WLTP range divided by the gap factor for the scenario (mild or −10 °C, below). The gap model is calibrated grid-side (charging losses included) while range comes out of the battery, so we take 8% charging losses off first (ADAC wallbox measurement). Cars with only an NEFZ figure (registered before ~2018) are scaled by ×0.88 to a WLTP equivalent first — otherwise older cars would read systematically optimistic. The same conversion applies inversely to NEDC consumption (÷0.88) before the gap model multiplies it; the input is always the homologation figure, never a measurement — the gap comes from the model, or it would be counted twice.

Long trip (600 km): a fixed reference run so cars stay comparable — leave full, arrive with 10% left, and charge only what the rest of the trip needs (never past 80%, where the curve goes flat). Motorway consumption = real-world consumption × 1.35; charging time comes from a modelled DC curve per cell chemistry (LFP holds a broad plateau, NMC peaks earlier and tapers longer) — where the maker states a 10–80% time (ADAC Autokatalog, for 400-odd cars) the curve is scaled to take exactly that long, never above the peak power. The figure is time at the plug only — pulling off, parking and paying are not modelled.

Tax: Kfz-Steuer from displacement + CO₂ above a 95 g/km allowance — the statutory formula, not a model; BEVs are temporarily exempt (first registration through 2030, up to 2035).

Insurance: estimated from list price (a comprehensive-cover proxy) × a class factor × Regionalklasse (postcode). 413 models already use their real GDV Vollkasko type class (official register 09/2025); the class→factor curve is a documented assumption pending quote calibration. Remaining models: power-band approximation.

Maintenance: a flat rate per drivetrain (EVs cheaper) — a sourced table, not a model.

Bonuses: the THG quota (BEV only) and the income-tested 2026 purchase grant are subtracted — statutory rules, not a model.

Financing (optional): if you enter the share of the price you finance + your own effective APR, we show the interest as a separate line (an amortising loan over the holding period). It is your rate, never an invented one — information only, not credit brokerage. Left blank, the line does not appear.

Total cost over the holding period: the €/month total × holding months — the "all in" figure beside the rate. Pure multiplication, not an extra model.

In short: depreciation, PHEV energy and insurance are modelled approximations, not quotes. Use the number to compare, not as a guarantee.

Price checks: 491 of 492 shown cars carry a check date — oldest 11 Jul 2026, newest 27 Sept 2026. A check date does not decay into "unconfirmed"; it states when the list price was last verified.

Published residual-value anchors (3 years) (As of 11 Jul 2026)

These models use published residuals instead of the segment approximation. "Basis" is the source's mileage assumption; listing studies are market observations, not forecasts.

Market benchmarks

• bev-all: 49 % — autohaus.de 20.01.2026 (DAT, Stand Dez 2025) (Erosion: 50,3 % (Jul 2025) → 49 %; Benziner 63,5 / Diesel 62,2)
• bev-premium: 49 % — ecomento.de 27.03.2025 (DAT/AutoScout24) (von ~70 % Anfang 2023 auf 49 % Ende 2024)
• bev-small-suv: 51.5 % — kfz-betrieb.vogel.de 11.07.2024 (Schwacke) (Stand Mitte 2024)
• bev-volume-range: 50 % — ecomento.de 03.06.2026 (Bähr & Fess) (publizierte SPANNE 0,45–0,55; 0,50 = Mittelwert)
• chinese-brands-used: 47.2 % — ecomento.de 27.05.2026 (Automobilwoche) (BEV+PHEV gemischt, kein fester 3-Jahres-Kohortenwert)

Our own market data (used listings)

Since week 28, 2026 we collect, weekly and ourselves, the prices used EVs are listed at in Germany: per model line the median of 20–40 fresh listings with their median age and mileage — 13 weeks (to week 40, 2026), 20 model lines. These medians appear as “Market ↺” on car pages, cost pages, head-to-heads and in the comparator. Rules: listings are asking prices, not transaction prices; the comparison is always against our estimate for exactly the listings' median age, never a blanket 3-year figure; trends only from two full 4-week windows (means of weekly medians), because a single weekly median swings by double digits on composition alone. The same data trains our residual model (release status: model health). Market Radar → · Model health →

Key assumptions (2026-01-01)

• Home electricity: €0.34/kWh
• Public AC / DC: €0.50 / €0.60/kWh
• Petrol / diesel: €1.80 / €1.70/L
• THG quota (BEV only): €300/yr
• Real-world vs WLTP (BEV, gap model energygap-v1-2026-07-26): mild ×1.202 (per segment ×1.173–×1.22), −10 °C winter standard scenario: additive heating penalty +5.5 kWh/100 km on real-world consumption (efficiency-sensitive — a constant heating load hits frugal cars relatively harder; falls back to ×1.35 without a consumption figure) — fixed, country-independent reference conditions. Annual € factor unchanged: DWD cold-season share 42 % × measured winter uplift 1.28 (ADAC long-term test). ICE/PHEV: ×1.2 (Ecotest constant).
• 2026 purchase grant, BEV (max, income-tested): €6,000

These values are read straight from the live cost model and the anchor table (no duplicated copy). Each figure carries its source and date on the car page; the model gates are on Model health.