Farm Machinery Depreciation Data: Where to Find It

That definition rules out most of what ranks for this query. A used-tractor listing tells you one asking price for one machine at one moment — it says nothing about what the seller's business actually expensed. An official stock count such as Eurostat's EF_MP_MACH table counts tractors in four power bands but attaches no euro values at all. Only accountancy sources answer depreciation questions, and inside Datadory's agricultural-farm-machinery pool exactly one does: the EU FADN Public Database, whose variable pane exposes machinery asset values, gross investment and depreciation straight from audited farm accounts.

One primary record owns this question and three others frame it. The comparison below lines them up against the counting sources people reach for by mistake.

FADN publishes depreciation in three editions, and knowing which one you are reading matters more than any filter choice. The Standard Results edition runs from 2004 onward on standard-output weighting, the SGM edition carries the older standard-gross-margin methodology across 1989–2009, and a Medians view re-expresses eleven layouts as medians rather than weighted means.

Every figure sits in a cell cut by country, NUTS region, TF8 or TF14 farm type, one of six SIZ6 economic-size classes and accounting year, with machinery asset values, gross investment and depreciation reported in EUR alongside the rest of the accountancy variables. Coverage is EU-wide with data final through 2023 and 2024 still preliminary, and the underlying sample aggregates roughly 75,000 audited farms weighted to represent about 3.5 million holdings.

Two caveats belong in your method notes. Figures are subject to continuous improvement, with documented breaks in series where national size thresholds changed — Hungary moved its economic-size threshold from EUR 4,000 to EUR 8,000 in 2023. And the record stops at weighted aggregates: no farm-level microdata exists, so a depreciation figure describes a cell, never a machine.

FADN gives you values but only from 2004, and only for the EU. Three counting datasets extend the picture backwards and outwards, and together they let you sanity-check any depreciation rate you assume.

FAOSTAT – Agricultural Machinery tracks tractors, combine harvester-threshers and milking machines in use across about 258 countries annually from 1961 to 2009 — roughly 10,700 rows, densest in 2007 at 8,204 observations. Its sample rows show why stock series matter for depreciation work: United States tractors in use drift from 4,690,000 in 1961 to 4,503,625 in 2000, Germany's total falls from 833,200 in 2005 to 681,200 in 2009, and China climbs from 52,239 tractors in 1961 to 3,515,757 by 2009. A fleet can shrink while it ages, which is precisely the signal a residual-value model needs.

The World Bank WDI – Agricultural machinery, tractors panel (AG.AGR.TRAC.NO) covers 292 countries, territories and aggregate groupings, of which 262 report at least once; year columns run 1960–2025, but observations end between 2007 and 2009 for most countries — 178 stop in 2007, 36 in 2008 and 10 in 2009.

Closer to home, Eurostat's EF_MP_MACH table counts tractors by power band (≤40 kW up to >100 kW), tillage machinery, seeders, sprayers and combine harvesters across reference years 2005, 2013 and 2023, split by ownership, farm size class and NUTS 2 region — roughly 100,000–200,000 observations. Pair its 2023 power-band mix with FADN's 2023 asset values and you can see whether a country's fleet is concentrated in the bands that depreciate fastest.

TractorHouse carries the richest schema — a 19-field detail page including year, manufacturer, model, full serial/PIN number and asking price displayable in 60-plus currencies. MachineryPete adds historical auction price results beside about 16,000 used-tractor listings, bucketed into horsepower ranges from Under 40 HP to 300+ HP. On the European side, Agriaffaires' UK edition listed 29,167 used tractors, and Mascus held 81,388 agriculture ads within roughly 498,000 active listings across about 40 portals.

For specification baselines, TractorData.com documents 18,310 tractor models across roughly 150–200 brands with production dates, serial numbers and original list prices — the last item being the anchor every straight-line depreciation schedule starts from.

Four audiences recur around this slice, and each leans on a different corner of it:

Persona fit follows the same split: financial users live in FADN's accountancy variables, market users in the listing panels. Either way the entry point is the EU FADN Public Database product page, where the field dictionary shows which variables each workflow needs.

Honest limits, because three of them sink naive projects.

First, geography. FADN covers EU Member States only — the UK appears through accounting year 2020 — and there is no US equivalent in the slice. American depreciation behaviour has to be inferred from marketplace residuals and the manufacturing-side cycle, not observed in farm accounts.

Second, continuity. The stock series stop where FADN starts: FAOSTAT's machinery domain froze at 2009, most World Bank tractor observations end between 2007 and 2009, and EF_MP_MACH offers just three farm-structure-survey reference years. A continuous value-and-count panel therefore exists only from 2004 for Europe.

Third, granularity. FADN publishes weighted averages by cell, never individual machines, and its own disclaimer warns of breaks in series where national thresholds changed. Depreciation schedules by make and model — the thing an appraiser often wants — exist nowhere in this pool as structured data; the nearest proxy is pairing TractorData.com's list prices and production dates with live asking prices from the marketplaces.

This page is one thread of a broader guide. Start with the agricultural farm machinery data guide for the pillar view of all 17 pooled datasets. For the fleet-size side of the same question, the walkthrough of agricultural machinery stock by country compares FAOSTAT and World Bank panels in detail, and tractor specs by model database shows how to build the specification baseline behind a residual model. The dataset-level record for the EU FADN Public Database documents every dimension and edition, and the agricultural-farm-machinery data hub indexes all 14 primary sources.