Contrarian Corpus
short seller research note follow up
2016-12-19 · 13 pages

China Huishan Dairy Holdings 6863.HK

Huishan's revenue is fabricated: tax-bureau data, cow-farm photos, and e-commerce scrapes all show the reported top line is dramatically overstated — the stock is a zero.

N 4 Narrative
V 2 Visual
C 2 Craft
Source URL unavailable

Thesis

Part 2 of Muddy Waters' China Huishan Dairy (6863.HK) short thesis argues Huishan's revenue — not just its margins and CapEx — is fraudulent. Investigators obtained State Administration of Taxation (SAT) VAT data for Huishan's four finished-product entities showing CY15 sales of RMB 2,352m, 29.4% below the company's adjusted reported figure and 36.2% below SAIC filings. The implausible 9.0 tons/cow yield underpinning Huishan's story is unsupported by the pre-IPO farms, where MWC investigators photographed rusting roofs, crumbling concrete, poor ventilation and cheap construction incompatible with world-class yields. Huishan's much-hyped 'Internet+' online sales miracle collapses under scraping: T-mall/Suning revenue reads as effectively zero while Mengniu and Yili dwarf it by ~1,500x. Muddy Waters concludes Huishan is 'a fraud through and through' and should struggle to avoid default.

SCQA

Situation

China Huishan Dairy is a Hong Kong-listed vertically-integrated dairy company claiming premium raw-milk yields, rapidly growing branded liquid milk sales and a thriving Internet+ e-commerce channel across northeast China.

Complication

Part 1 established margin/CapEx/cash fraud; Part 2 shows the top line itself is fabricated — SAT tax records, decrepit cow-farm conditions, and third-party e-commerce scrapes all contradict Huishan's reported revenue and yield narrative.

Resolution

Muddy Waters publishes granular tax-bureau, on-the-ground and web-scraping evidence to force the market to reprice the shares and to warn creditors given Huishan's heavy debt load and collapsing real cash flow.

Reward

Given fake revenue, fake margins and a massive debt stack, Muddy Waters expects Huishan to struggle to avoid default — implying equity that trades toward zero for short-sellers positioned against it.

The three reasons

  1. 1

    SAT tax data shows reported revenue is 29-36% higher than actual finished-product sales

  2. 2

    Claimed 9.0 tons/cow milk yield is implausible vs China average of 5.8 tons and pre-IPO farms are dilapidated

  3. 3

    Huishan's e-commerce 'sales miracle' is fiction — T-mall/Suning sales outsold by Mengniu/Yili by 1,500x

KPIs cited

SAT vs adjusted reported CY15 finished-product sales
RMB 2,352m SAT vs RMB 3,333m reported — 29.4% lower
SAT vs SAIC CY15 finished-product sales
RMB 2,352m SAT vs RMB 3,685m SAIC — 36.2% lower
Milk yield per cow
Huishan claimed FY14 yield 9.0 tons/cow vs China 2012 average of 5.8 tons
Share of external sales covered by SAT data
~79% of Huishan external sales are finished products
Online sales (Huiji subsidiary, CY15)
VAT/SAIC show max ~RMB 1m online vs ~RMB 7.5m offline and ~RMB 8.4m total
T-mall monthly unit sales vs peers
Huishan 826 units; Yili 15,722; Mengniu 35,151 — ~1,500x revenue gap
Reported 1H17 sales growth vs peers
Huishan liquid milk +29.1%, overall +17.7% vs Mengniu 6.6%, Yili 0.2%, Modern -8.5%
Realized raw milk price
Huishan ~RMB 4,000-5,000/ton vs MoA national index ~RMB 3,300-4,200/ton

Pattern membership

Where this document fits across the library's 12 rhetorical / structural patterns.

Notable slides (5)

Notes

Part 2 of a two-part Muddy Waters short thesis on China Huishan Dairy (6863.HK) published 2016-12-19. Part 1 focused on margin/CapEx/cash fraud; Part 2 layers on revenue fraud using (a) State Administration of Taxation VAT records, (b) on-the-ground farm photography and expert commentary on cow-shed design, and (c) algorithmic e-commerce scraping of T-mall/Suning. Document is formatted as a Word-style research note with embedded photos and basic Excel charts — not a designed deck. Quotes Huishan 2016 Annual Report verbatim ('sales miracle' language) to set up the e-commerce teardown. Huishan shares collapsed ~85% in March 2017, validating the thesis. campaign_phase set to follow_up since this is explicitly Part 2 of the same short campaign.