Funds Never Read a 56-1 From Page One — Here Is the Playbook for Where They Actually Open (a 4-Part, 9-Item Map + the 6 Places Everyone Skips)

You have done this. Pull the Form 56-1 for a name you hold, open it, hit 300 pages, and tell yourself "tonight I finish it."

Pages 1 through 10 go fine — vision statement, corporate structure. By page 20 you are into shareholding charts and rosters of directors and your eyes drift. Page 35 you scroll faster. Page 50 you close the file.

And you end up in the same place every time: somebody else's one-page summary. A post in a group, a note passed around, a 10-minute earnings video. Two problems with that, and nobody tells you about either. One, it has already been filtered by someone else — they decided for you what matters. Two, it is always late — by the time anyone writes the summary, the price has moved.

Here is the thing that shocked me when I started out on the risk side at a bank:

No institutional analyst reads a 56-1 from page one. Not one of them.

They do not read faster than you. They are not more diligent than you. They know where the material is. They open straight to that page, read it against the one item it must always be read with, and close the book — roughly half an hour per company. The military word for it is surgical strike, not taking the whole city.

This article is that map. And I will tell you up front that it ends on something bitter: even with this map in hand, it is not enough. We will do the arithmetic later.

Why this document outweighs anything you read free on the internet

Every 56-1 One Report ends with a statement almost nobody reads, and it is the reason this document is not in the same category as a research note. The One Enterprise Public Company Limited puts it under certification of the accuracy of information in its FY2021 One Report: "บริษัทฯ ได้สอบทานข้อมูลในแบบแสดงรายการข้อมูลประจำปีฉบับนี้แล้ว ด้วยความระมัดระวัง บริษัทฯ ขอรับรองว่าข้อมูลดังกล่าวถูกต้องครบถ้วน ไม่เป็นเท็จ ไม่ทำให้ผู้อื่นสำคัญผิด หรือไม่ขาดข้อมูลที่ควรต้องแจ้งในสาระสำคัญ" [named individuals are personally on the hook if this book misleads you — that is what makes it evidence instead of opinion] — and the directors sign it.

That is the whole difference. Nothing you read free on the internet carries a signature attesting that it is not false. One Report does.

What One Report is — the short answer

Form 56-1 (One Report) is the annual registration statement/annual report that the SEC (Thailand's Securities and Exchange Commission) requires listed companies to produce. It merges what used to be two separate documents — Form 56-1 (the annual registration statement) and Form 56-2 (the annual report) — into a single book. It must be filed within 3 months of the end of the accounting period (SET Investnow)

The format went market-wide in 2022 — InfoQuest reported it at the time: "ในปี 2565 จะพิเศษกว่าทุกปีที่ผ่านมา เพราะนับตั้งแต่ปีนี้เป็นต้นไป แบบแสดงรายการข้อมูลประจำปีและรายงานประจำปีจะรวมกันเป็น '56-1 One Report'" [the entire market changed document shape on the same date — anything you built on the old format breaks at that line]. The first documents to carry the name therefore cover FY2021, filed through early 2022.

The most common confusion is treating "One Report" and "56-1" as two different things. They are the same document. The real cover reads "แบบแสดงรายการข้อมูลประจำปี/รายงานประจำปี 2568 แบบ 56-1 One Report" [one cover, both filings, FY2025 — if you are hunting for a separate annual report, there isn't one]. The MD&A is not a separate document either — it is Item 4, inside the book.

Why it matters: Everything a company writes in this book sits under that certification. Getting it wrong carries legal consequences. That is why I weight language in a One Report above news coverage or a management interview — not because it is always truer, but because the cost of writing something misleading is much higher.

1. The real structure: 4 parts, 9 items, plus attachments

I put the tables of contents of three real One Reports side by side (AppliCAD FY2025, Lohakit Metal FY2022, The One Enterprise FY2021). The structure matches all the way through, because the SEC dictates the items. Only the thickness differs. This is the map to memorize:

Something from the real books that most write-ups leave out: some printed volumes never label a "Part 4" — they simply put the certification at the back under the heading certification of the accuracy of information (2 of the 3 books I opened do it this way). And the number of attachments is not the same at every company — the FY2022 book I opened has only Attachments 1–4; the FY2025 book has the full 1–6. Write a scraper that assumes every book is identical and it breaks right here.

From the model-validation seat: When anyone hands me data, the first question is not "what is the number" but "which item of which document did this come from." The item is the context. The same figure sitting in Item 2 (risk) and in Item 4 (discussion of results) does not translate the same way. The first is what could happen. The second is what already has.

2. Where the MD&A sits, and why it carries the book

MD&A = Part 1, Item 4, management discussion and analysis. The SEC splits it into three sub-items:

• 4.1 Analysis of operations and financial position, and material changes — management explaining in its own words why the numbers came out the way they did
• 4.2 Factors or events that may materially affect financial position or operations — what they see coming
• 4.3 Financial statement data and key financial ratios

The weight of 4.1 sits here: the financial statements say what happened; the MD&A says why. And "why" is the only part of the entire book that requires judgment from whoever writes it — which makes it the part that tells you the most about how management sees the world.

Two companies, the same gross profit decline, two completely different explanations:

AJ · FY2025 Q2: "gross profit decreased by THB 97.45 million or 141.26% due to the average selling price declining more significantly than the decrease in costs YoY." — source filing at the SEC (page 3)

PTTEP · FY2025 Q1: "For the First quarter of 2025 compared with the First quarter of 2024 A decrease of 36 MMUSD or 7% from a net profit in Q1/2024, was primarily from lower average selling prices, together with higher depreciation, depletion and amortization, as well as higher operating expenses." — source filing at the SEC (page 4)

These two paragraphs describe structurally different businesses. The first is a business where the selling price fell faster than costs could follow it down — margin squeezed from the price side. The second is a business where selling prices fell alongside rising depreciation — a capital-heavy cost structure. Information at that level is not in the financial statement tables. It exists only in the words.

3. The playbook: the 6 places institutions open, in the order they open them (copy it)

This is the core of the article. Read a 200–300 page book from page 1 to the last page and you run out of energy before you reach anything good. The sequence below is what I actually use when I have to know a company inside half an hour, and it is not ordered by page number. It is ordered by information density per minute burned.

300 pages  →  6 places actually opened · ~35 min total

  ①  Part 4 · certification + signatures ──────────────   2 min
         └─ confirm the filing is complete before spending time in it
  ②  Item 4.1 · MD&A, results of operations ───────────  15 min  ★ start here, not page 1
         └─ let the company name its own earnings driver
  ③  Item 2.2 · risk factors ──────────────────────────  10 min
         └─ ⚡ always read it "against ②" — this is the whole playbook
              written in ② but absent from ④ = not yet live
              present in both               = already in the numbers
  ④  Item 5.3 · legal disputes ────────────────────────   3 min
         └─ shortest section in the book, costliest to skip
  ⑤  Item 9.2 · related-party transactions ────────────   5 min
         └─ who connected to the major shareholders is the money going to
  ⑥  Notes to the financials (only what ② cites) ──────  as far as you get
         └─ let the MD&A point the way, do not hunt on your own

  The other ~270 pages = read when there is a reason to, not because they are in the book

Detail on each stop:

1. Part 4 — the certification (2 min) Who signed, and whether anything is attached alongside it. This is a check that the document is complete before you spend time reading it.
2. Item 4.1, MD&A (15 min) Read the discussion of results first. Do not start with what the business does — the MD&A will tell you itself what drives the results.
3. Item 2.2, risk factors (10 min) Read it against Item 4.1 immediately. A risk written into Item 2 that never surfaces in Item 4 = a risk that has not materialized. One that shows up in both = already happened.
4. Item 5.3, legal disputes (3 min) Short, and the most expensive thing in the book to skip.
5. Item 9.2, related-party transactions (5 min) Who, among people connected to the major shareholders, the money is flowing to.
6. Notes to the financial statements (as far as you get) Only the notes the MD&A points at — you do not read all of them. Let the MD&A tell you which note to open.

Step 3 is the one people skip most often and the one that pays me most, because it changes the question from "what are this company's risks" — which gets you a long list every company writes more or less the same way — to "which of these risks has started landing in the results", which narrows to one or two, and those are the ones that mean something.

A rule from measuring risk: The longest risk list is usually the most useless one, because it is written to cap liability, not to communicate. The real signal is the point where the language in the risk section and the language in the results section start using the same words.

4. Six places readers skip — with the filed text for every one

All of this comes out of an index I built myself on documents actually filed with the SEC. Every box links back to the original.

Skip 1 — the sentence that says which side compressed the margin

Most readers retain "profit fell." The full sentence names the mechanism:

CITY · FY2025 (full year): "The higher net loss was attributable to several factors including the increase in cost of sales and services which rose in accordance with the sales and service incomes, and from the minimum wage hike in January 2025." — source filing at the SEC (page 1)

Skip 2 — labor cost wired directly to government policy

When the minimum wage moves, labor-intensive companies write about it immediately, and they write about it very differently:

CPH · FY2024 (full year): "increased labor costs from the adjustment of the minimum wage rate on January 1, 2024" — source filing at the SEC (page 1)

QLT · FY2025 (full year): "Policies to increase the minimum wage directly affect the Company's business operations, as the Company's primary cost is personnel wages." — source filing at the SEC (page 6)

The first says "it happened, we booked it." The second says "our cost structure is wages." The second is worth far more, because it also tells you the sensitivity to the next policy round.

Skip 3 — FX gains and losses that are not operating profit

WICE · FY2022 Q2: "resulting from foreign exchange gains from the weak baht situation most of which the company have foreign trade receivables denominated in USD resulting in a relatively high exchange rate profit" — source filing at the SEC (page 2)

That profit came from the currency, not from selling better. Run growth off net profit without reading this paragraph and the trend you get is wrong.

Skip 4 — receivable quality signals

AFC · FY2022 Q2: "There are long standing debtors ... with combined outstanding balance of 124.32 million Baht. ... a reserve for bad debt was set at 74.67 million Baht, unchanged, to cover possible loss due to the difference between the realized value of the collateral and the outstanding balance" — source filing at the SEC (page 2)

unchanged is the heaviest word in that sentence — the reserve did not move, which means management judges the collateral value still covers the exposure. If that word disappears next quarter and the number moves, that is a change of assumption, not a change of number.

AEONTS · FY2025 Q2: "The Company continues to implement debt restructuring measures for persistent debtors and participates in the Bank of Thailand's "You Fight, We Help" program, to maintain asset quality at an appropriate level." — source filing at the SEC (page 4)

I built Early Warning Systems at a bank, and a sentence like that is what the model catches before the NPL numbers move, because entry into a restructuring program always precedes reclassification.

Skip 5 — capex that has not reached the statements yet

NEO · FY2024 (full year): "Furthermore, the Company is in the process of expanding its production capacity to support future growth. The investment plan will be adjusted based on the Company's performance, such as dividing the construction of factory buildings into phases and evaluating new machinery investment annually" — source filing at the SEC (page 2)

"dividing the construction ... into phases" is the company telling you the investment plan flexes if results disappoint — information that could never appear in last year's CAPEX figure.

YONG · FY2026 Q1: "resolved to approve the investment in the construction of one concrete products factory and three ready-mixed concrete plants, including investments in machinery and equipment to support the production of new products." — source filing at the SEC (page 1)

Skip 6 — legal disputes (Item 5.3, the shortest section in the book and the most expensive to skip)

TNITY · FY2023 Q1: "The management of the Company and the subsidiary have determined that this event is an abnormal transaction, as it was carried out with fraudulent intent. Such conduct is prohibited by law and runs counter to public order and good morals." — source filing at the SEC (page 1)

KC · FY2023 (full year): "7 of 24 above-mentioned defendants, former executives, and related persons, are in the process of the Company's lodging a complaint with the Director-General of the Department of Special Investigation, to proceed the criminal prosecution for participation in defraud, fraudulent understandings from the sale of bills of exchange, and consent to record the account incorrectly." — source filing at the SEC (page 5)

Both paragraphs sit inside a document the directors certified as not false, and both are text that will never appear in a one-page earnings summary.

5. Reading across years: the real case where one company tells the same story for five of them

The highest-value use of One Report is not reading one book. It is reading the same company across consecutive years and watching how the explanation changes.

I pulled every CITY cost explanation held in the index and lined them up by date. The result is a complete narrative arc, and I had to interpret nothing:

Every line in this table clicks back to the original at CITY's filings on market.sec.or.th

The 2024 sentence — "unable to cut down expenses any further" — is the one I want you to look at. The company itself says the toolkit is empty. One year later the minimum wage rises again, right on cue. Read only the FY2025 book and all you see is "a loss because wages went up." Read across years and you see the end of a process that started in 2022.

I wrote up a case like this in more detail at Reading MD&A across FY2021–FY2026: why one company tells a different "story" every year

Why it matters: Five years of financial statements are available everywhere. Five years of explanations, lined up and read in one sitting, almost nobody does, because it means opening five books of 200 pages each. That is the gap I built a tool to close.

6. Why Ctrl+F inside the PDF is not enough

Say you want to know who in the market is writing about "ความเสี่ยงจากการพึ่งพาลูกค้ารายใหญ่" [the risk that a handful of buyers can walk off with the revenue line]. Hit Ctrl+F on that string and you miss all of this:

AH · FY2023 Q1: "significantly higher volume from our major customers" — source filing at the SEC (page 4)

MBAX · FY2025 Q3: "the sales of the new product from the major customer starting from the 4th quarter of 2024 onwards" — source filing at the SEC (page 2)

Neither sentence uses "concentration risk" or "การกระจุกตัว" [the phrase a Thai filing would reach for if it wanted the risk flagged — and neither of them reached for it], and both are the same thing. Companies do not write their reports in the words that happen to occur to you. Full reasoning at Why Ctrl+F cannot read the whole market's filings

7. Five traps I walked into myself

This is the part a generic SEO article will not write for you, because it makes the subject look harder. Not knowing it is how you get caught on the day you are most confident.

1. "Not found" does not equal "the company did not write it" — some filings are scanned images with no text layer to search. In a live run on the minimum-wage topic on August 5, 2026, my system cut 11 rows out of a single result set: not_located 5 rows, no_text_layer 4 rows and no_source_pdf 2 rows. A cut row does not mean the data is wrong. It means I cannot verify it, so I do not show it.
2. The MD&A is management's angle, not neutral fact — they choose what to emphasize. The fix is reading it against Item 2.2 (risk factors) and the notes to the financial statements, not swallowing it whole.
3. Numbers in the discussion and numbers in the statements can sit on different bases — especially anything described as adjusted for extraordinary items. Check the definitions match before you compare across companies.
4. The Thai and English versions do not always track each other exactly — the passages quoted here are in English because the index is built on the English versions the companies file. For a formal citation, anchor on the version filed with the SEC.
5. Quoting is not summarizing — I deliberately do not let the system summarize on a company's behalf. The moment it summarizes, it becomes the author, and you start believing something the model made up without noticing. The mechanism that lets fabricated numbers escape I wrote up separately at AI invents numbers more convincingly than you think: the 3 gates I use to stop it

The line you do not cross: If you take text from a One Report into a research note, into a client conversation, or onto a board table, you have to be able to point back to the document page in one click. A result you cannot trace is worth less than nothing, because it looks credible enough to get used again.

8. Questions I get asked

Is One Report different from 56-1? No. Same document. The full name is the annual registration statement/annual report, Form 56-1 (One Report), which merged the old Form 56-1 with the annual report (Form 56-2) starting in 2022.

Which page of One Report is the MD&A on? Part 1, Item 4, management discussion and analysis. The page number varies by company — the last book I opened had it on page 70, another on page 67. Work from the table of contents. Do not memorize a page number.

Where do I download One Report? The SEC's disclosure system at market.sec.or.th, and the company's own investor relations page. Both are free.

When does a company have to file? Within 3 months of the end of the accounting period — a company closing its books on December 31 files around late March.

Is reading only the MD&A enough? No. But if you only have time for one item, read that one, then open Item 2.2 and Item 5.3 in the order set out in Section 3 above.

Do I have to read every book myself if I follow many companies? That is the point where doing it by hand stops working. One company across 5 years = 5 books × ~250 pages. Twenty companies in a portfolio = past what one person can do. See the next section.

9. The bitter lesson I promised at the top: the map alone is not enough

Everything in this article can be done by hand, free. The documents are public, the SEC lets you download them, and I recommend doing it by hand on the two or three companies you care about most first, because that is how you learn what you are looking for.

Now multiply the playbook above by your actual portfolio.

Assume 20 names in the book and you want the 5-year read from Section 5

   20 companies × 5 years          =   100 filings
   100 filings × ~250 pages        =   25,000 pages
   100 filings × 35 min (playbook) =   ~58 hours

   = 2 hours a day, every day, without missing one  →  just under 1 month
     and that is a single pass. No notes taken.
     No cross-year comparison. No cross-company comparison.

   The month ends just as the new quarterly numbers land → start over at one

This is where doing it by hand stops working. Not because you are not good enough, but because the arithmetic does not favor one human. And it is the answer to why the institutional side invested in data teams and NLP a long time ago — they do not read better than you. They just do not read with the naked eye.

I hit this wall myself when I wanted to know who in the market was writing about labor costs, and there is no way to do it one book at a time. So I built Boom Leverage Terminal: a semantic index over management discussion and analysis and Form 56-1 filings for 916 listed companies, covering FY2021–FY2026, roughly 479,663 indexed rows of text — the near-month of work in the box above collapses into one question answered in seconds.

The rules I set for myself and enforce in code:

• Every row passes a citation gate — the system takes that sentence and finds it in the real source file. No position found = cut, not shown. Measured live off the index built on August 14, 2026: 362,524 rows checked, 359,413 rows = 99.14% traced back to a document page, and the remaining 3,111 rows were kept out of the results (by domain: auditor's reports 99.48% · financial statements 99.06% · risk 99.23% — all inside the FY2021–FY2026 window that is open to search).
• It quotes what the company wrote and does not summarize on its behalf — what you get is verbatim plus a page number.
• No buy or sell recommendations — the system answers who wrote about what. It does not answer whether to buy or sell, and it will not.

If you would rather not think up search terms yourself, the Scan page takes one topic and runs it across the whole market (how it works). If you already have a watchlist, upload a CSV/Excel and filter down to your own names (how to do it).

The whole picture

One Report is not a hard document. It is just long, and most people do not know where the material is. What to keep from this article:

• Fixed structure: 4 parts, 9 items, plus attachments — learn the map once, use it on every company.
• MD&A = Part 1, Item 4 — the financial statements with "why" attached.
• Always read Item 4 against Item 2.2 — a risk that has started surfacing in the results is a risk that has already happened.
• Reading across years beats reading one year — the CITY case above tells its own story five years running.
• Everything you cite has to click back to the source page. If it cannot, do not use it.

Run it yourself tonight — a free Boom Leverage Terminal account normally gets 10 credits/day with no card on file, but for the launch logging in alone unlocks the full DELTA pack, free until September 30, 2026 — pulling 4 full years of history · 150 credits/day, again with no card, because billing is switched off entirely during the promotion. When it ends the account drops back to the free pack on its own. No automatic charge. (why I am giving it away, and the full terms · a straight comparison of how deep each pack goes)

Go back to the arithmetic box in Section 9. Four years of history is the depth that makes playbook step ③ work, because reading risk against results only means something when you can see how it moved across years, not a still frame of the latest quarter.

For teams and institutions — bulk data pulls, file exports to work with downstream, or wiring it into internal systems through an API: details at enterprise plans, or talk to me directly at contact@boomleverage.com. I worked model risk in a bank, and I know a risk team needs more than an answer that sounds good.

To be explicit: This article is educational and explains how to read public disclosure documents. It is not investment advice. It does not direct anyone to buy or sell any security, and it makes no guarantee or prediction about price. Every company name here is used only to demonstrate how to read text in documents filed with the SEC. The quoted passages are verbatim from the source documents, with no interpretation or rewording — before making an investment decision, read the full documents yourself and consult a licensed professional as appropriate to your situation.