Hermès Index — Methodology

Universe

Components include luxury handbag variants meeting our liquidity, source coverage, and price thresholds:

• Minimum 100 sales over the trailing 12 months
• Minimum 4 data sources confirming prices
• Median price above $1,000
• Brand from our covered universe (14 houses)

Weighting

Weights are computed from trailing 12-month sales volume using square-root scaling:

weight_i = sqrt(volume_i) / Σ sqrt(volume_j)

Maximum 10% per component (cap), minimum 0.3% (floor). Excess from capped components is redistributed proportionally to uncapped components. Final weights normalized to sum to 1.0.

Index value formula

Index = 100 × Σ (weight_i × current_price_i / base_price_i) / divisor

Current prices use the trailing 90-day median per component, aggregated across all platforms. Base prices are fixed on the rebalance date and updated only on composition changes.

Rebalancing

• Weight rebalance: quarterly (re-weighting only, composition unchanged)
• Composition review: semi_annual (admissions and removals)
• Buffer rules: ±20% threshold around inclusion/exclusion criteria to prevent flip-flopping
• Announcement: Rebalances are published 30 days before effective date

Chaining

When composition changes, the divisor is recomputed to ensure index value continuity across the transition:

divisor_new = divisor_old × (old_basket_value / new_basket_value)

This guarantees the index value on the rebalance date is identical before and after the change. Standard practice across major equity indexes (S&P 500, Dow Jones).

Audit and changelog

Every rebalance event is logged with before/after composition, weight changes, and the updated divisor. Historical compositions are preserved indefinitely so the index value can be recomputed at any past date — guarding against survivorship bias.

Methodology updates are versioned with effective dates. The current version is 1.0.