Artmetria Index Methodology

The art market is not a single transparent market. It is fragmented across auction houses, galleries, private sales, fairs, regions, categories, periods and reputational networks.

Artmetria starts from this complexity.

The Index is built to help users compare signals without pretending that every signal has the same reliability, depth or meaning.

A high score should not be read as a recommendation. A low score should not be read as a dismissal. Each score is a structured signal that must be interpreted within its coverage, data quality and category context.

The Artmetria Index may combine several families of signals:

• market activity;
• auction liquidity;
• price visibility;
• presence across selected houses and segments;
• category and period context;
• object-level characteristics;
• provenance and literature signals when available;
• condition and medium-related signals when available;
• confidence level and data completeness;
• qualitative interpretive factors.

These signals are organised into a proprietary framework inspired by the way specialists assess artworks, markets and collecting contexts.

The purpose is not to replace expert judgement, but to make the first layers of analysis more legible.

The Artmetria Index value shown on each cockpit hero is a weighted composite of three market terms, rebased to 100 on a per-vertical anchor date.

Index_t = 100 × (
   0.50 × (median_price_t / median_price_baseline) +
   0.30 × (log_volume_t / log_volume_baseline) +
   0.20 × (quality_t / quality_baseline)
)

Why median, not average:a single 100 M€ outlier should not drive a vertical's index. The median is robust to extreme values by construction.

Why log volume: raw transaction counts grow exponentially with ingestion coverage. The logarithm smooths volume bursts while preserving directional signal.

Why a weighted quality term: a market can be cheap (low median) but composed of premium lots, or expensive but composed of entry-level pieces. The quality factor (estimation-weighted mean of the analytical score) captures the mix.

Baseline per vertical: Fine Art anchors at January 1, 2008(long-cycle pivot). Niche sub-verticals with thin historical coverage anchor at their own densest credible window — 2018for Books & Manuscripts and Jewellery, 2020 for Design.

The two pills shown next to the value (↗ +X.X% 90d and ↗ +X.X% 1y) are real rolling deltas — not cumulative motion versus the 2008 baseline.

For each delta, the engine computes the index value at the comparison date (t − 90 j or t − 365 j) using its own 90-day window, then takes the percentage change. Both endpoints share the same baseline so the delta isolates the directional motion.

Deltas are suppressed when the reference window has fewer than 50 sold lots — at that point the rolling comparison is noise.

Three primitives keep the displayed value honest about its own sampling error and composition.

Winsorization on the quality term.The quality factor is a weighted mean of analytical scores, weighted by lot estimation. A single mega-lot with a 50 M€ estimate could dominate the aggregate. Each lot's weight is capped at the in-window 99th percentile of estimations before the weighted mean is computed. The median and count terms are unaffected — they are already robust by construction.

Confidence interval on the value. Each snapshot carries a 95 % confidence interval (the ± X.X shown next to the value) derived from the Bonett–Price asymptotic formula on the median:

CI_{value, 95%} ≈ 79 × IQR / (median × √n)

The 79 constant folds the median's 1.58 × IQR / √n distribution into the composite formula (price term carries 0.5 of the weight). The interval is wider on small samples — a 350-lot vertical reads ± 8, a 6 000-lot vertical reads± 3.

Adaptive window. When the rolling 90-day window has fewer than 50 sold lots — the empirical floor below which median + volume become unstable — the engine automatically widens the window to 180 days and surfaces the change in the hero footer.

Every hero displays a coverage badge. There are four states :

• Canonical index — the current 90-day window has at least 1 000 sold lots and the baseline window is fully scored. The value is computed across a dense market.
• Under construction — between 100 and 1 000 sold lots in the current window. The value is directional but precision will improve as coverage grows.
• Provisional value — the baseline has price and volume signal but no quality scores yet (the analytical scoring backlog has not caught up to the anchor date). The displayed value reflects price + volume motion; the quality dimension will engage automatically as scoring fills in historical lots.
• Insufficient data — the baseline is genuinely empty (zero sold lots or zero median price). No value is displayed; the hero shows an explainer instead.

The blended global index averages markets that don't behave alike. Three regional sub-indices isolate the motion :

• Americas — sales held in US salesrooms (New York, Los Angeles, Philadelphia, Chicago, Dallas, Boston, Miami) + Canada.
• Europe — UK rooms (London, Edinburgh), continental Europe (Paris, Monaco, Geneva, Zurich, Cologne, Munich, Berlin, Vienna, Milan, Rome).
• Asia — Hong Kong, Tokyo, Beijing, Shanghai, Seoul, Taipei, Mumbai, Singapore.

Per-sale geography, not per-house.Sotheby's, Christie's, Bonhams and Phillips are international auction houses that run sales across all three regions — assigning each house to a single region by headquarters would systematically miscount Europe and Asia. The geographic slices therefore filter by the auction's actual salesroom country (parsed from each sale's URL or page metadata), not the parent house's domicile.

Gazette Drouot is an aggregator.The Drouot entry on the regional house mix covers ~70 individual French auction houses that sell at the Hôtel Drouot (Millon, Piasa, Ader, Osenat, Cornette de Saint-Cyr, etc.). It reads as a single name in the breakdown because Drouot's catalogue surface doesn't expose per-house granularity uniformly.

Coverage caveat.The regional mix reflects only the houses currently indexed: Christie's and Bonhams contribute most of the volume today, while some sources (e.g. Drouot, Sotheby's) are sparsely indexed or temporarily unavailable, so the breakdown under-weights them for now. The per-sale country mapping is also more complete on recent and upcoming sales than on the historical archive (some older URLs no longer resolve to a page that carries the location), so the geographic sparklines should be read with that in mind on dates before 2024.

Artmetria uses a C1–C10 framework to organise analysis across multiple dimensions.

At a high level, these dimensions can include identity, period, quality, subject, composition, technique, format, condition, provenance, literature, visibility and market context.

The C1–C10 scoring is currently a preliminary pipeline (an early version): credible as a reading grid, but still being calibrated as Artmetria expands its coverage and improves its data normalisation.

C1–C10 should be understood as a reading grid. It helps break down an artwork or market signal into interpretable components, rather than compressing everything into a single opaque number.

Artmetria distinguishes between market-level signals and object-level signals.

Market-level signals may relate to activity, liquidity, auction history, regional presence, sale frequency, price visibility or institutional/market attention.

Object-level signals may relate to the characteristics of a specific artwork, such as medium, format, period, subject, condition, provenance, exhibition history or literature references when available.

This distinction matters because a strong artist market does not automatically make every artwork equally strong, and a compelling object may exist within a market segment that remains thin, emerging or difficult to read.

Every analytical signal depends on the quality and depth of the underlying data.

When Artmetria has stronger coverage, the interpretation can be more stable. When coverage is partial, the interpretation must remain more cautious.

A score may therefore be accompanied, now or in future versions, by confidence indicators, data-quality notes, coverage warnings or source-context explanations.

This is a core part of the methodology: uncertainty should be made visible rather than hidden.

An Artmetria score is not a certificate, appraisal, authentication, investment recommendation or guarantee of future value.

It is an analytical signal produced from available information and methodological assumptions.

Scores may help users compare, question, prioritise or contextualise. They should not be used alone to make acquisition, sale, legal, tax or financial decisions.

Artmetria is designed to support judgement, not replace it.

Art markets are affected by taste, scarcity, provenance, macroeconomic conditions, institutional narratives, collector behaviour, geography, liquidity and timing.

For this reason, Artmetria does not present its Index as a price-prediction engine.

The purpose is to make signals more readable, not to claim certainty about future performance.

The platform favours structured interpretation over speculative forecasting.

The methodology is designed to evolve.

As Artmetria expands coverage, improves normalisation, refines scoring layers and adds more contextual signals, the Index can become more precise and more useful.

This evolution does not mean that earlier outputs were “wrong”; it means that the platform is built as a progressive analytical system.

The beta version should therefore be read as the first public expression of a method, not as a finished authority.

Artmetria is intended for educational, analytical and research-oriented use.

Users should treat the Index as one layer among others: direct viewing, expert advice, provenance research, condition review, legal checks, tax considerations and independent judgement all remain essential.

The Index helps ask better questions. It does not remove the need to ask them.