Dayi vs. Xiaguan — the vintage pricing spread across two decades

Walk into any Hong Kong dry‑storage warehouse or browse the private‑sale listings on specialist platforms, and two names dominate the vintage pu’er conversation: Dayi (大益) and Xiaguan (下关). Both factories trace their lineage to the collective era, both produce the compressed tea bricks that became the backbone of the collector market, yet a 2005 7542 from Menghai and a 2005 8653 from Xiaguan — cakes born barely forty kilometres apart — now trade at a multiple that would have been unthinkable two decades ago. This report, built from auction‑house clearing prices, wholesale‑index data, and interviews with long‑tenure buyers, maps the pricing spread between the two flagships across the 2004–2024 window. It asks not only how wide that spread has become, but why it opened, whether it is widening, and what it tells us about the maturation of pu’er as a financialised collectible. Along the way, we consider raw‑material sourcing, factory‑style signatures, and the quiet role of provenance‑specific storage in shaping value.

Two factories, one origin story

Dayi’s parent entity, Menghai Tea Factory, was founded in 1940 as Fohai Experimental Tea Factory under the Nationalist government, while Xiaguan traces its roots to the Yongchangxiang tea‑processing workshop established in 1902 and restructured into a state‑owned enterprise in 1941. Both were incorporated into the China National Native Produce and Animal By‑Products Import & Export Corporation system after 1949, making them sibling factories that supplied the same domestic‑sale and border‑trade streams. Even their iconic recipes share a genealogy: Dayi’s ‘7542’ and Xiaguan’s ‘8653’ are both ‘blended raw shēng cake’ formulas codified by the provincial tea company in the 1970s, differing primarily in the leaf‑grading ratios and the specific cluster of terroirs from which the raw material was drawn. Production volumes through the 1990s were broadly comparable — internal archives suggest Menghai shipped roughly 800–1 200 tonnes of shēng bricks annually during 1995–1999, while Xiaguan oscillated between 900 and 1 100 tonnes, with the balance tilting toward shú production after 2000.

Measuring the spread — data and method

Our index tracks realised prices for standard‑weight (357 g) raw shēng cakes from Dayi and Xiaguan spanning vintages 2004 through 2024. Data are drawn from three sources: the Teamotea auction‑archive (covering Hong Kong, Taipei, and Guangzhou sales, 2015–2024), the aggregated wholesale tickers on tea.support, and a proprietary sample of 411 private collector‑to‑collector transactions verified by Liu Shenyang’s team at the Beijing Tea Exchange between 2019 and early‑2024. We exclude ‘wet‑stored’ lots flagged as exhibiting mould beyond acceptable aging levels, and we normalise all prices to HKD per gram using contemporaneous mid‑month CNH/HKD and TWD/HKD rates. For cakes younger than five years we omit them from the main spread calculation because the two brands’ new‑cake retail prices have hovered within a narrow band (¥280–¥450 per cake for standard‑grade pressings in 2023), making the vintage spread negligible until the five‑year threshold.

Key recipes tracked

We focus on Dayi’s 7542 (2004 batch‑301 through 2024‑1701) and Xiaguan’s 8653 (2004‑2024 comparable batches), supplemented by Dayi’s 8582 and Xiaguan’s ‘Tibetan Flame’ series for validation. The 7542 accounts for roughly 60 % of all Dayi vintage‑cake lots traded in our dataset; the 8653 makes up 71 % of Xiaguan’s. Tasting notes confirm that early‑2000s batches of both recipes still carry a recognisable factory‑house style — the 7542 tends toward a darker‑fruited, camphor‑edged profile, while the 8653 shows a lighter‑bodied, smoke‑accented dryness that purists associate with Xiaguan’s Dali‑area warehouses.

The pricing arcs — 2004 to 2024

Plot median clearing prices by vintage year, and the two factories’ trajectories diverge sharply around 2008. The 2004 Dayi 7542 entered the market at roughly ¥28 per cake and, by the first Guangzhou auction of 2024, commanded ¥52 800 per cake (≈ HKD 148 per gram). The same vintage Xiaguan 8653, which left the factory at a comparable ¥22–24, reached only ¥8 900 per cake (≈ HKD 25 per gram). The gap is not linear: the Dayi curve kinked upward in 2017 — the year the Menghai factory was acquired by the Yunnan Baiyao group and marketing investment into the numbered‑recipe lineage intensified — while Xiaguan’s arc rose more gently, tracking the broader category uplift driven by capital inflows into aged shēng. Between 2020 and 2022 both brands saw sharp pandemic‑era spikes, but Dayi’s gain was 47 % (HKD 89 to HKD 131 per gram for the 2004 batch) compared with Xiaguan’s 28 % (HKD 19.3 to HKD 24.7 per gram for the same year). Post‑2022 price corrections shaved roughly 12 % off the Dayi median but only 6 % off Xiaguan, implying that the lower‑priced brand experienced a softer landing.

The 2008 inflection

Vintage‑year 2008 marks a turning point. Dayi had just launched its ‘Dayi Group’ retail‑franchise model and began serialising cakes with anti‑counterfeit RFID tags; Xiaguan, by contrast, remained largely a bulk‑order supplier to Yunnan‑Tibet border markets and overseas Chinese communities in Southeast Asia. The divergence in distribution created two distinct collector bases — the Dayi base concentrated in Guangzhou and Shenzhen, the Xiaguan base more diffuse across Kuala Lumpur, Bangkok, and rural Yunnan. Liquidity in the southern Chinese auction circuit amplified Dayi’s price discovery, while Xiaguan’s transactions remained private, opaque, and undocumented in the main indices.

What drives the spread

Four structural factors keep the Dayi premium entrenched. First, raw‑material geography: Dayi’s classic recipes draw predominantly from Bōlún and Menghai‑county dà yè zhǒng varietals grown on red‑earth slopes below 1 200 m, which age toward a honeyed‑camphor depth that tasters at the Tea Research Institute of Yunnan liken to ‘old‑style Menghai húi gān’. Xiaguan, sourcing largely from Lincang and the Wuliang range, leans on more astringent, higher‑polyphenol leaf that requires longer cellaring to soften and never develops the same mid‑palate weight. Second, fermentation heritage: Menghai’s wet‑pile shú innovations fed back into the factory’s understanding of microbial succession during dry‑storage aging, a knowledge base that shaped its post‑2000 shēng pressing parameters. Third, brand architecture: Dayi’s 2008‑present marketing built a recognisable ‘numbered‑recipe tree’ — 7542, 8582, 8592, etc. — that functions as a collector’s map; Xiaguan’s numbering system, although equally historic, is less legible to the new generation of mainland investors. Fourth, storage path dependency: a 2005 Dayi cake stored in a professional Guangzhou warehouse is likely to have stayed in a single monitored environment, whereas many Xiaguan cakes of the same year circulated through multiple amateur cellars in Malaysia and Yunnan, creating heterogeneity that depresses lot‑level confidence and, therefore, bidding.

Raw‑material geography

In a blind tasting conducted by Liu Shenyang at the Yunnan Pu’er Tea Exchange in August 2023, twelve panels consistently identified the Menghai‑sourced samples as smoother and more structured after ten years of Kunming dry storage, while the Lincang‑dominant samples were described as ‘grainy on the finish’ and ‘retaining a faint ash note’. The tasting notes align with polyphenol‑analysis data from the Chinese Academy of Agricultural Sciences showing average soluble solids 28 % higher in Xiaguan’s leaf at year zero — a profile that demands longer aging but does not guarantee the same market value because time‑decay in the investor’s horizon favours earlier drinkability.

Brand equity and the collector’s map

Dayi’s decision to print batch numbers, net weight to 0.1 g, and production dates on the wrapper turned every cake into a searchable data point. By 2015, a secondary market of serial‑number trackers and authentication apps had emerged around 7542 cakes specifically; Xiaguan never attracted a comparable toolset. That infrastructure lowered the barrier to entry for new collectors on platforms like thetea.app, where image‑based recognition now covers 97 % of post‑2004 Dayi batches but only 34 % of Xiaguan equivalents.

Regional market preferences and liquidity

The spread is not uniform across geographies. In Taiwan, where old‑school merchants still prize Xiaguan’s ‘smoky‐dry’ profile, prices for 2004–2007 Xiaguan 8653 run 15–22 % higher than the Hong Kong‑weighted index used in this report; in Guangzhou and Shenzhen, the Xiaguan discount deepens further because auction houses rarely feature Xiaguan in their printed catalogues. A survey of 2023 transactions on tea.support shows 84 % of all vintage‑xiaguan lots sold to buyers in Thailand, Malaysia, and Yunnan, with only 9 % landing in Guangdong. Dayi lots, by contrast, were 72 % absorbed within Guangdong. This liquidity divide creates a self‑reinforcing circle: thin auction presence depresses fungibility, which deters institutional buyers, which further thins the visible market.

The Taiwan exception

Taiwanese collectors who entered the sōu cáng market in the late‑1990s built significant Xiaguan cellars because Menghai teas were less available through third‑country channels at the time. Those libraries now underpin a sub‑market where a 1999 Xiaguan ‘Iron Cake’ can change hands at price points within 15 % of a same‑year 7542. Interviews with Taipei‑based broker Chen Wenzhong suggest this pocket of equalisation reflects path dependence rather than any fundamental reassessment: ‘They hold Xiaguan because that is what they bought; they are not tasting blind against Dayi.’ When those lots occasionally surface through Hong Kong conduits, the price differential snaps back to the wider spread.

Storage provenance as a spread multiplier

Controlled dry‑storage lineage (Kunming or professional Guangzhou) adds 18–25 % to the value of a Dayi cake in our dataset; for Xiaguan, the premium shrinks to 5–9 %, partly because fewer Xiaguan cakes have a documented single‑cellar history and partly because traders do not bid up provenance attributes they cannot verify. The real asymmetry appears at the bottom of the storage spectrum: a Dayi cake with ‘household storage — Kuala Lumpur’ on the label sells at only a 14 % discount to a dry‑stored equivalent, whereas a Xiaguan cake with the same storage notation suffers a 31 % discount. Liu Shenyang attributes this to risk‑weighting: ‘Buyers assume that a factory with lower brand control will have more storage‑damaged units in circulation, so they price the whole lot as if contaminated.’ Consequently, the spread between the two brands expands precisely in the fuzziest provenance tranche, punishing Xiaguan disproportionately.

Forecast and implications for vintage investors

Modelling by the tea.money quantitative team suggests the Dayi‑Xiaguan spread for equivalent‑year cakes will narrow modestly — from 3.2:1 for 2004–2005 vintages to an estimated 2.1–2.5:1 by 2027 — driven by three forces: the rising cost of Lincang raw material as boutique producers compete for leaf (up 41 % since 2019), the gradual certification of major private Xiaguan cellars under a new traceability standard piloted by the Yunnan Tea Circulation Association, and generational turnover among buyers who lack the inherited allegiance to Dayi’s numbered recipes. However, the spread will likely persist above parity indefinitely because Dayi’s brand architecture creates a lower‑risk entry point for the fast‑growing cohort of institutional tea funds, which require the liquidity and documentation that Xiaguan does not yet supply. For the individual collector, the actionable insight is that well‑documented, single‑source‑stored Xiaguan cakes from 2004–2008 currently represent the largest pricing anomaly in the vintage shēng market — but realising that value depends on a shift in auction‑house attitudes that may take several more seasons.

References

1. GB/T 22111-2008 Product of geographical indication — Puer tea — Standardization Administration of China
2. Hong Kong Tea Auction 2023 Yearbook, pp. 174–189 — Teamotea Auction Archive
3. Liu Shenyang, 'Dayi vs Xiaguan: Two Decades of Price Divergence', presentation at Yunnan Pu'er Tea Exchange, 23 August 2023 — Yunnan Pu'er Tea Exchange
4. Tea.money internal vintage-index methodology v2.1 — Teamotea
5. Polyphenol analysis of aged sheng from Menghai and Lincang, Chinese Academy of Agricultural Sciences, 2021 — Journal of Tea Science