Guide · Agent skill

How to use the ADR, H-Share and A-Share Analysis agent skill

Updated 2026-10-29 · written with AI from the project's own documentation and checked against our catalogue data

The adr-hshare skill gives an AI agent a set of analytical frameworks for comparing one Chinese company's share price across its US ADR, Hong Kong H-share and mainland A-share listings. This guide explains what the skill does, how to install it, how to ask for it in plain language, and where its limits are. After reading, you will know how to ask the agent for a premium calculation, how to read the result against the skill's reference ranges, and when to check its output against live data.

ADR, H-Share and A-Share AnalysisHKUDS/Vibe-Trading · MIT Download the skill

Quick start

1Download the skill. Download the adr-hshare ZIP file from this page.
2Unzip into a skills folder. Unzip it into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder for one project.
3Or upload it in the Claude apps. In the Claude apps, open Settings, go to Capabilities, then Skills, and upload the skill.
4Ask a specific question. Name the company, the listings and the exchange rates or prices you want analysed.
5Supply your own numbers. Give the agent current prices and FX rates, since the skill contains only static examples.
6Check the output. Compare the result with the skill's reference bands and confirm conversion ratios against the current listing terms.

What does the adr-hshare skill make the agent do?

Many Chinese companies are listed on several exchanges: A-shares in Shanghai or Shenzhen, H-shares in Hong Kong, and ADRs in the US. The skill teaches the agent how these structures work, how to calculate the premium or discount between listings, and how to read that premium against typical ranges.

The skill also covers delisting risk for US-listed Chinese companies, including the HFCAA (Holding Foreign Companies Accountable Act) audit framework. It is analytical guidance with Python formulas and worked examples. It does not pull live market data itself.

Explains cross-listing structures such as A + H, H + ADR and A + H + ADR
Calculates AH premium and ADR premium with the formulas in the file
Lists ADR conversion ratios for companies such as Alibaba, JD.com and Baidu
Reviews delisting risk factors such as PCAOB inspection status and HK listing type
Describes arbitrage strategy ideas and event-driven cross-listing effects

How do you install the adr-hshare skill?

Download the ZIP file for the skill from this page. Unzip it into your personal skills folder at ~/.claude/skills/, or into a project's .claude/skills/ folder if you want the skill available only in that project.

If you use the Claude apps instead, open Settings, go to Capabilities, and upload the skill under Skills. The skill package contains one file, SKILL.md, which holds the instructions the agent reads.

Personal install: unzip into ~/.claude/skills/
Project install: unzip into the project's .claude/skills/ folder
Claude apps: Settings → Capabilities → Skills, then upload

How do you trigger the skill with a natural request?

You do not need a special command. The skill's description covers cross-listing premium analysis, tracking pricing gaps between ADRs, H-shares and A-shares, arbitrage signals, dual-listing valuation and delisting risk. Ask about those topics directly and name the company and the listings involved.

Include the numbers you have. The skill contains only static example figures, so the agent needs you to supply current prices and exchange rates for a real calculation.

Example: "Calculate the AH premium for PetroChina using A-share 8.50 CNY, H-share 6.20 HKD, USDCNY 7.25 and USDHKD 7.82."
Example: "What is the ADR premium for Alibaba if BABA is $85.00 and 9988.HK is HKD 82.50, using the 8:1 ratio?"
Example: "Assess the delisting risk for a US-listed Chinese ADR with PCAOB access blocked and no Hong Kong listing."

How is the AH premium calculated?

The AH premium compares the A-share price with the H-share price converted into CNY. The H-share price is multiplied by the USD/CNY rate divided by the USD/HKD rate. The skill's example uses PetroChina: an A-share at 8.50 CNY and an H-share at 6.20 HKD give an H-share value of about 5.75 CNY and a premium of about 47.8%.

The skill then groups premium levels into interpretation bands, from above 50% (extreme) down to below 0% (H-share premium over A-share). These bands are the skill's own reference points and are not a verified model.

def calculate_ah_premium(a_price_cny, h_price_hkd, usdcny, usdhkd):
    h_price_cny = h_price_hkd * (usdcny / usdhkd)
    ah_premium = (a_price_cny / h_price_cny - 1) * 100
    return ah_premium

How is the ADR premium calculated?

The ADR premium compares the ADR's USD price with the USD value of the equivalent Hong Kong shares. The conversion ratio matters: for Alibaba, 1 ADR equals 8 HK shares. The skill's worked example uses BABA at $85.00 and 9988.HK at HKD 82.50 with a USD/HKD rate of 7.82, giving a premium of about 0.71%.

The skill's table lists ADR ratios for companies including Alibaba (8:1), JD.com (2:1), Baidu (8:1), NetEase (5:1) and Trip.com (1:1). Check the ratio against the current listing terms before using it, because the table is static.

def calculate_adr_premium(adr_price_usd, hk_price_hkd, adr_ratio, usdhkd):
    hk_equivalent_usd = (hk_price_hkd * adr_ratio) / usdhkd
    premium = (adr_price_usd / hk_equivalent_usd - 1) * 100
    return premium

How does the skill assess delisting risk?

The skill sets out a simplified HFCAA framework. Risk is rated Low when PCAOB inspection is complete with no issues, Medium when deficiencies are noted, High when PCAOB access is revoked or restricted, which starts a 3-year delisting countdown, and Critical when a company is on the SEC identified issuer list for three consecutive years.

The file also lists factors to weigh: PCAOB status, years on the SEC identified issuer list, whether a Hong Kong listing exists and whether it is primary, VIE structure, and state-owned status. The skill notes that a Hong Kong primary listing can act as a backup if US delisting occurs. These tiers are simplified summaries and should not be treated as legal advice.

Low: PCAOB inspection completed, no issues
Medium: PCAOB inspection completed, deficiencies noted
High: PCAOB access revoked or restricted
Critical: on the SEC identified issuer list for 3 consecutive years

What about the arbitrage strategies and data access?

The file describes three strategy ideas: an AH premium mean-reversion rule based on a z-score, ADR and Hong Kong intraday gaps, and event-driven effects such as new Hong Kong listings or index inclusion. The mean-reversion example uses a z-score threshold of 2.0 with illustrative inputs. The file gives no backtest and no risk controls.

The Data Access section includes example yfinance code for downloading prices and FX rates. The skill itself does not fetch data. If you run that code yourself, it needs network access and its example dates are fixed in the file, so change them to the period you need.

Mean-reversion thresholds are illustrative and not validated
Use your own current data for any real calculation
Run the example download code only in your own environment

What people use it for

Explaining why a dual-listed Chinese stock trades at different prices in Hong Kong and mainland China
Working through a student exercise on AH premiums and ADR conversion ratios
Reviewing the delisting risk factors for a US-listed Chinese ADR as part of a research note
Setting up a spreadsheet-style premium calculation using the skill's Python formulas

Frequently asked questions

Is the adr-hshare skill free to use?

The skill is published under the MIT licence, as stated on this page. Check the licence text in the source repository for the full terms before reuse.

What does the adr-hshare skill require to run?

The skill lists no requirements. It runs no scripts and uses no network access, so it works as guidance text the agent reads. Any calculation needs prices and exchange rates that you supply.

Does the skill give live stock prices?

No. It contains only static example numbers and no data source. The Data Access section has example yfinance code, but the skill does not fetch prices on its own.

How do I install the adr-hshare skill?

Download the ZIP from this page and unzip it into ~/.claude/skills/ or a project's .claude/skills/ folder. In the Claude apps, upload it under Settings, then Capabilities, then Skills.

Can I use this skill to make trading decisions?

It is not designed for that. The signal thresholds and mean-reversion rule are illustrative and not validated, and there is no backtest or risk control, so acting on them could lose money.

How does this compare with other ways to check AH premiums?

The skill is a set of analytical frameworks and formulas that an agent can apply to numbers you give it. It is not a data feed or a trading tool, so for live quotes you need a separate market data source.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

This skill suits investors, analysts, students and researchers who want a clear method for AH and ADR premium calculations and a simplified view of HFCAA delisting risk. It is not suited to trading decisions, real-time pricing, or legal or regulatory advice, so pick a market data service or a qualified adviser for those needs.

Download the skill