This InsightMeter guide is for readers in Hong Kong and the wider Asia region who watch US-listed names—often through American Depositary Receipts (ADRs)—while also reading Form 13F institutional tables and Form 4 insider filings. Your reader job is practical and narrow: stop mixing ADR tickers, ordinary share counts, depositary ratios, and currency conversions when you interpret those disclosures. The existing HK/Asia filings guide covers timezone, EDGAR search, and general research hygiene. This article goes one level deeper on the unit-of-account problem that quietly invents fake accumulation stories.
Before you screenshot a “smart money buy,” write three labels on the same note: (1) which legal security you mean (ADR program vs ordinary/local share vs another share class), (2) which unit the filing uses (ADR shares, ordinary shares, or another deposited unit), and (3) whether any dollar value you see is USD market value, local-currency value converted at an FX rate, or a broker app’s display preference. If any of those three is blank, you do not yet have a comparable observation—you have a collage.
Asia-based readers hit this collage constantly. A Hong Kong broker may show the ADR ticker you trade. A US 13F information table may list a CUSIP that maps to the ADR, the ordinary, or a related class depending on what the manager holds. A Form 4 may report transactions in the class the insider is required to report—often not the ticker on your phone home screen. FX headlines (“HKD strength,” “JPY move”) can then make a flat USD stake look like it “grew” or “shrank” in local purchasing-power chatter that has nothing to do with share accumulation.
Soft product note: normalized screens can speed lookup across filings, but they do not replace checking the ADR ratio and identifier on the primary document. Treat secondary tables as pointers back to EDGAR, not as the final court on “how many shares.”
In a broker app, you usually see a trading ticker, a last price in a display currency, and a quantity of “shares” you can buy or sell. That quantity is almost always ADR units if the product you trade is the ADR. The app rarely forces you to ask: “how many ordinary shares does one ADR represent today?” Ratios change; programs change; unsponsored vs sponsored programs differ in documentation quality. The app optimizes for order entry, not for filing literacy.
On EDGAR, Form 13F information tables identify holdings primarily through issuer name, title of class, CUSIP, and share (or principal) amount—not through the marketing ticker your broker prefers. Managers report the securities they hold that fall within the 13F reporting framework. If the manager holds the ADR, the line should reflect that ADR security. If the manager holds ordinary shares (when those are the reportable US-listed or otherwise 13F-eligible instrument in context), the line reflects that instrument. Assuming “same company name = same line as my ADR ticker” is how double-counting and fake “new positions” begin.
Form 4 filings center on the issuer and the class of equity securities in which the reporting person has a reportable interest. Transaction tables show amounts in that class. Footnotes may explain ADR programs, conversions, or depositary arrangements—but many readers never open footnotes. A code P purchase of 10,000 securities is not automatically “10,000 of the ADR I trade at 9:30 US Eastern,” nor “10,000 ordinary shares on the home exchange,” until the class and any ratio language say so.
Practical habit: keep a one-line identity card for each name you follow—local ticker (if any), ADR ticker(s), CUSIP(s) you have seen in 13Fs, current ADR ratio with date checked, and a note on whether dual-listed ordinary shares are commonly held by US 13F filers. Update the card when ratios or tickers change. Stale identity cards create confident wrong notes.
The classic trap is converting without saying so. Suppose one ADR represents two ordinary shares. A Form 4 reports 50,000 ordinary-equivalent units in a class that is not your ADR. Your thread says “insider bought 50,000 ADRs.” You just doubled economic size in the narrative—or halved it, depending on the direction of the mistake. Either way, the conviction story is now fiction.
The second trap is comparing across periods when the ratio changed. If a program consolidates from 1:5 to 1:1, naive share deltas look like massive buys or sells. Corporate actions and depositary ratio changes are not “smart money discovered the name.” They are unit redefinitions. Adjust history to a common ordinary-share equivalent (or a common ADR equivalent) before you label accumulation.
The third trap is mixing ADR share counts with ordinary free-float percentages from a local exchange presentation. Ownership percentages only make sense when numerator and denominator share the same share class basis. Quoting a US 13F ADR share count as a percent of local ordinary shares outstanding without converting is arithmetic theater.
Prefer this sentence structure in notes: “As of [date], [filer] reported [N] [ADR units / ordinary shares / other], CUSIP […], which under a [R]:1 ADR ratio checked on [date] equates to roughly [M] ordinary-share equivalents; USD value [V] if stated.” Long sentences beat wrong verbs.
CUSIP matching is powerful and brittle. Two lines with similar issuer names can be different depositary receipts, preferred classes, or legacy CUSIPs after reorganizations. Always read the title of class field. “ADR,” “ADS,” “Ordinary,” “Common,” and class letters are not decoration—they are the matching key beside the CUSIP.
Vendor aggregators sometimes roll ADR and ordinary into one “company exposure” tile for storytelling. That tile can be educationally useful as a prompt and disastrous as a share-count source. If your claim is “Manager North added ADR shares,” you need the ADR line’s share amount, not a blended economic exposure estimate. If your claim is “North’s economic long interest in Issuer X rose,” you must say you are combining instruments and show the conversion math.
Amendments matter here too. A 13F-HR/A can correct CUSIP, share amount, or title of class. Asia readers who screenshot a first print during US night hours and never refresh can freeze the wrong security identity into a group chat. Prefer the latest amendment for the report period before you argue about size.
Remember structural lag: 13F is quarter-end residue filed later. Even a perfectly identified ADR line is not a live blotter of what the manager holds today. Pair ratio hygiene with lag honesty so “North is loading the ADR this morning” never silently means “North’s table as of last quarter-end showed more ADR shares than the prior quarter.”
Form 4 codes still mean what they mean—P is not magically bullish theater—but Asia readers add two distortion layers: unit conversion and FX. A purchase reported in one class can be restated by a blog into ADR units incorrectly. Separately, a local-currency headline (“insider spent HK$X million”) may use an FX rate from a different day than the transaction price in the filing, inventing a round-number drama.
Mechanical codes remain mechanical across borders. Awards (A), option exercises (M), and tax withholding (F) can populate share columns without open-market buying conviction. Ratio mistakes on mechanical lines create especially loud false clusters: “three insiders bought huge ADR blocks” that were mostly grants restated with a wrong multiplier.
Footnotes are where ADR program language often hides. If a footnote explains conversion into ADSs or references a deposit agreement, capture that language before you compare the share column to your broker quantity. Skipping footnotes is how “buys” survive contact with paperwork for about twelve seconds in a careful review and forever on social media.
Worked fiction: CFO Form 4 shows code P for 20,000 shares of ordinary class; ADR ratio is 1 ADR = 5 ordinary. Careless restatement: “CFO bought 20,000 ADRs.” Careful note: “CFO purchased 20,000 ordinary shares (accession …); ~4,000 ADR equivalents at 5:1 ratio checked [date]; not a recommendation.” Same filing, opposite narrative temperature.
Fiction A — Fake 13F add from ratio change. Prior quarter ADR shares 100,000 at 1:2. Program changes to 1:1; manager’s economic interest flat; reported ADR share count halves or doubles depending on how the depositary restates. Naive badge: “trimmed 50%.” Adjusted view: ratio restatement—flag provisional until common-basis shares compared.
Fiction B — Broker ticker ≠ 13F line. You trade ADR ticker WIDG. Manager table shows CUSIP for WIDG ordinary-related class with title “Ordinary Shares.” Thread claims “North bought your WIDG ADR.” Corrected claim: “North’s 13F shows ordinary-class line; ADR equivalence requires ratio math; do not paste broker ticker as identity.”
Fiction C — FX-only value story. ADR share count flat quarter to quarter; USD market value up because ADR price rose; a local newsletter converts value to HKD at a weaker USD/HKD print and screams “stake ballooned in HKD terms.” Educational verdict: share count flat; price and FX explain value optics; not accumulation.
Fiction D — Form 4 cluster with mixed classes. One officer buys ADRs (code P); one director receives ordinary-class award (code A) restated online as ADR buys; one 10% holder rebalances. Cluster thesis of “aligned ADR buying” dies once class and codes are separated.
Copy this checklist before you publish or forward an ADR-related 13F or Form 4 claim.
Checklists are educational process tools. They do not create profitable trades.
Most ADR filing errors are identity errors wearing confidence.
Fixing them makes Asia↔US research slower and much clearer.
Even perfect ADR/ordinary hygiene does not reveal shorts, many derivatives, intra-quarter trading, or private context. Filings are incomplete by design. Clean unit math reduces avoidable false stories; it does not produce guaranteed outcomes.
ADR ratios and program details can be hard to verify quickly across issuers. When you cannot confirm the ratio, mark the note provisional: “class matched; ratio unchecked.” Provisional is an honest research state. Inventing a ratio to finish a thread is not.
Some managers and insiders hold economically related instruments that never appear together in one tidy table. Absence of an ADR line is not proof of zero economic interest in the issuer’s story.
Nothing here is legal advice or a recommendation to buy or sell any security, ADR, or ordinary share.
Teams serving HK/Asia readers can standardize a shared identity card template: fields for tickers, CUSIPs, ratio, last verification date, and “known pitfall” notes (e.g., dual-class votes, multiple ADR programs historically). Attach the card to any internal alert that mentions the name.
When a vendor tile disagrees with your card, investigate before you update either. Silent disagreement is how two analysts quote incompatible share counts in the same meeting.
Revisit cards after depositary announcements, stock splits, and ticker migrations. Burns are curriculum.
Prefer: “North’s 13F ADR-class line (CUSIP …) shows N shares as of [as-of]; prior M; ratio R:1 checked [date]; filing date …” Avoid: “North is chasing the same stock you buy on your phone.”
When you must speak in local currency, put FX in a subordinate clause and keep share counts in the main clause. Value-only headlines travel faster than corrections.
If class identity is uncertain, say “issuer-related line; ADR match not confirmed” in the first sentence. Burying uncertainty is how wrong tickers spread through WhatsApp forwards.
Use the HK/Asia US-filings guide for EDGAR search mechanics, timezone habits, and general cross-border workflow. Use share-count vs dollar-value whenever value optics tempt you. Use Form 4 codes and footnotes when insider lines look like conviction. Use 13F lag so quarter-end residue never becomes “live flow.”
This ADR article is the unit-of-account layer those guides assume. Without it, Asia readers can follow every other checklist and still invent size.
Methodology and glossary pages help you keep vocabulary stable—CUSIP, ADR, as-of, accession—so teammates argue about evidence instead of labels.
HK/Asia investors reading U.S. filings · Why share-count changes beat dollar headlines · Form 4 內部人代碼解析 · 13F lag: quarter-end ≠ trade timing
Match the filing’s security class and CUSIP to the instrument you mean, convert with a dated ADR ratio, separate FX value optics from share accumulation, and read Form 4 codes/footnotes before you narrate buys. Clean units beat dramatic tickers. Continue with related guides, methodology, and glossary. Nothing here recommends buying or selling any security.