How to Read an SGX Annual Report (What Actually Matters) -

Published: 20 May 2026 | Last data refresh: May 2026 (Frencken Group FY2025 and Micro-Mechanics FY2025 annual reports)

The SGX annual report is the cheapest, most thorough piece of due diligence a retail investor has access to. Free, filed with the exchange, 150 to 200 pages, required about an hour or two to read properly. Most retail investors either don’t read it at all, or stop somewhere around page thirty.

This is a framework for reading SGX annual reports in the order that extracts the most signal per minute. Not a stock recommendation — the point is to make you capable of forming your own view from primary evidence, rather than from broker summaries or social media commentary.

I use Frencken Group (SGX: E28) and Micro-Mechanics (Holdings) Ltd (SGX: 5DD) as worked examples, drawing from their FY2025 annual reports. Both illustrate what good disclosure looks like; they appear here as instructive examples, not recommendations. One practical note: SGX companies use different financial year-ends — Frencken’s closes 31 December, Micro-Mechanics’ on 30 June — which affects when reports appear and whether peer comparisons are meaningful on a given date.

What most readers do, and why it doesn’t work

The default approach is front-to-back. Chairman’s letter. CEO’s message. Financial highlights. A skim of revenue, net profit, and EPS. Done.

Three reasons this fails.

The chairman’s and CEO’s letters are marketing prose, drafted with full knowledge that most readers form their first impression there. Reading them before the evidence means you absorb the management narrative before you can evaluate it.

The “financial highlights” page is curated by the company. Management chose which numbers to feature. If revenue was up but margins compressed, the highlights page may lead with revenue.

Reading the headline statements before the notes means you don’t yet know what the numbers represent. Revenue is just a figure until you’ve read the revenue recognition policy. Net profit is just a figure until you’ve read the segment breakdown. The notes give the headline numbers their meaning — reading the answer before the question is reading in the wrong order.

My reading order

1. Auditor’s report and Key Audit Matters

This is where I start. The Key Audit Matters (KAMs) section discloses what required the most auditor judgement, where misstatement risk was highest. It is the most underread page in any annual report.

2. Notes to the financial statements

Before the income statement. Before the balance sheet. The notes contain the accounting policies, segment breakdown, related-party transactions, contingent liabilities, and capital commitments. Reading these first means the headline statements are legible when you reach them.

3. MD&A / Operating and Financial Review

Compare management’s narrative against what you’ve already seen in the notes. The gap between the two is often where the story lives.

4. Corporate Governance section — skimming

Look for the related-party transactions disclosure, director dealings, and audit committee composition.

5. Chairman’s and CEO’s letters — with scepticism

By now you know what the notes say. Read the management letters as positioning rather than primary disclosure.

6. Shareholding Statistics

Usually the last few pages. Are founders and directors still holding meaningful stakes, or quietly distributing? A stock where the top-20 shareholders hold 90%+ is thinly traded — small order flow can move it significantly.

This order is opinionated: notes first, management commentary last.

The buried-in-plain-sight sections

Seven sections that retail investors most commonly miss, and where the most useful signal tends to live.

Segment information

Usually near the end of the notes — Note 32 in Frencken FY2025; absent in Micro-Mechanics, which operates as a single reportable segment (Note 3.13).

The segment note tells you what the business actually is, decomposed by product line and geography — versus how management describes it in the chairman’s letter.

Frencken FY2025: Two divisions: Mechatronics and Advanced Plastics Solutions (APS). The prior-year name for APS was Integrated Manufacturing Services (IMS) — a rename worth noting, because relabelling a segment can quietly impair period-over-period comparisons. Mechatronics contributed S$778.4M of S$865.1M group revenue and essentially all of group profit. APS contributed S$83.1M (down from S$85.7M) and moved from a S$0.1M profit into a S$1.1M loss. The same note discloses that just three customers in Mechatronics account for S$412.7M, or 47.7% of total group revenue — a customer-concentration story invisible from the income statement alone.

Micro-Mechanics FY2025: Single reportable segment. The simplicity is itself informative: genuinely one business, doing one thing, across five plants.

Notes complexity should track business complexity. When it doesn’t, ask why.

Related-party transactions

Usually around Note 4–5 (Note 4 in Frencken; Note 20 in Micro-Mechanics), and in the corporate governance section. Look for: rentals from director-owned premises, services from family-linked entities, loans to or from directors, and transactions with companies controlled by directors or significant shareholders.

Frencken FY2025: Clean. Intra-group management fees only, plus standard key management personnel (KMP) compensation — S$5.342M in total (S$4.597M salaries, S$159K pension contributions, S$586K share-option expense), of which S$2.042M relates to directors of the Company. No related-party rentals, services, or director loans.

Micro-Mechanics FY2025: Note 20 discloses that the US subsidiary leases office premises and receives electrical services from Sarcadia LLC, a company controlled by one of its directors. Total annual payments run to approximately S$1M against group revenue of S$65.2M. The arrangement is fully disclosed, the auditor signed off, and the amounts are immaterial relative to the business. I flag it not because anything is wrong — it isn’t — but because this is precisely the kind of disclosure most retail readers skim past.

Normal: small disclosed amounts, clear business purpose, on commercial terms. What warrants scrutiny: material and growing amounts, vague descriptions, and especially a related party that is also a lender or guarantor.

Revenue recognition policies and changes

Found in the accounting policies section (Note 2(h) in Frencken; section 3.11 in Micro-Mechanics) and the revenue note itself.

Two broad categories. Point in time — revenue booked when control of goods transfers, usually on delivery and acceptance. Conservative when acceptance criteria are substantive. Over time — revenue booked progressively as work is performed; common for construction, tooling, certain services. Can be aggressive if completion percentages are management-estimated rather than milestone-verified.

Frencken FY2025: Three streams, three policies. Sale of goods: point in time, on control transfer plus customer acceptance. Moulds: over time, percentage-of-completion via engineer-certified milestones. Installation services: point in time, on functional testing and acceptance. All conservative; none changed from FY2024.

Micro-Mechanics FY2025: Single policy. Revenue recognised at point in time on delivery to the customer’s site with all acceptance criteria met. As conservative as it gets.

A year-on-year policy change — particularly from “upon acceptance” to “upon shipment” — something to keep an eye on. Quiet changes that inflate current-year recognised revenue are one of the oldest available tools.

Trade receivables and the receivables-vs-revenue check

Found in the trade receivables note (Note 20 in Frencken; the statement of financial position in Micro-Mechanics).

The check: compare year-on-year receivables growth to revenue growth. If receivables grow materially faster, the company is extending looser credit to hit sales targets or failing to collect on completed sales — either is a warning sign.

Frencken FY2025: Revenue up 8.9% (S$794.3M → S$865.1M). Net trade receivables up 0.8% (S$137.8M → S$138.9M). Collection improved. AR concentration also improved: 31% of receivables from four customers, down from 37% in FY2024. Credit terms run 14–120 days.

Micro-Mechanics FY2025: Revenue up 12.6% (S$57.9M → S$65.2M). Trade receivables up 14.8% (S$10.85M → S$12.45M), the AR-to-revenue ratio nudging from 18.7% to 19.1%. Not a flag in isolation; in a business with 30–60 day terms, a two-point gap might be caused by a single large invoice timed late in the period.

The aged-receivables breakdown matters more than the headline. A ballooning over-90-days bucket (depending on industry, same industry might be considered as a norm) is where a collection problem hides.

Operating cash flow vs net income

Compare net income to operating cash flow over three years. A profitable company that persistently fails to convert earnings to cash is either absorbing it into working capital or producing accounting earnings that don’t fully represent economic reality.

A quick approximation: cumulative operating cash flow divided by cumulative net income over three years. For a steady-state industrial, expect roughly 1.0. Persistently below 1.0 means earnings are sitting on the balance sheet. Above 1.0 usually indicates conservative accounting.

The cash flow statement’s operating section shows exactly where the gap is going — changes in receivables & payables, inventory.

Strong historical cash conversion only tells you about the past. Free cash flow is operating cash flow minus capex, and capex isn’t always smooth. The next subsection checks what’s contractually committed.

Contingent liabilities and capital commitments

Usually toward the end of the notes (Note 33 in Frencken; Note 21 in Micro-Mechanics). Contains signed contracts for capex not yet on the balance sheet, guarantees, litigation, and tax disputes. Most years this note is small. The years when it isn’t are exactly when it matters.

Frencken FY2025: Frencken’s historical cash conversion is fine. But the commitments note discloses that contracts for property, plant and equipment, not yet on the balance sheet, jumped from S$490,000 at end-FY2024 to S$60.1M at end-FY2025 — roughly 120 times larger. Invisible from the income statement, the cash flow statement, or the balance sheet. The careful reader registers: significant capex is contractually locked in, it will compress free cash flow over the next one to two years, and historical conversion ratios are no longer the right forward anchor.

Micro-Mechanics FY2025: Capital commitments of S$113,772 — essentially nil. The contrast is instructive: Frencken is building; Micro-Mechanics is harvesting.

Read these two notes together: operating cash flow conversion (backward-looking) and capital commitments (forward-looking). Either alone gives you half the picture.

Auditor’s Key Audit Matters and going-concern paragraphs

KAMs disclose which areas required the most auditor judgement — where misstatement risk was highest. Required for SGX-listed company audit reports. The single most underread page in any annual report.

A KAM does not mean something is wrong. It means the auditor paid extra attention here. Every KAM has two parts: the matter, and what the auditor did in response.

Frencken FY2025: Single KAM — impairment assessment of goodwill, carrying S$20,658,000 (2.80% of total assets). Deloitte & Touche LLP reviewed management’s value-in-use calculations, growth rates, and discount rates, concluding estimates fell within a reasonable range. Standard for an industrial company with prior acquisitions on the balance sheet.

Micro-Mechanics FY2025: Single KAM — impairment of property, plant and equipment and right-of-use assets of the US subsidiary, carrying S$20.7M (down from S$23.9M in FY2024). KPMG LLP noted the subsidiary had been loss-making for several consecutive years, though it returned to profitability in FY2025. More informative than Frencken’s: it names a specific subsidiary and a specific concern, not a group-level estimate on an acquired intangible.

How to read a Key Audit Matter (KAM) Example: Frencken FY2025 — Impairment Assessment of Goodwill ① The Matter Why the KAM was raised: Goodwill of S$20,658,000 (2.80% of total assets) arises from prior business combinations. Impairment testing requires significant management judgement on future cash flows, growth rates, and discount rates. ② How Addressed What the auditor did: Deloitte assessed the reasonableness of management’s value-in-use model, challenging growth rate assumptions and discount rates against industry benchmarks. Concluded estimates were within a reasonable range. ③ Cross-reference to Note See Notes 2(f)(i), 3(b)(i), and 18 for the goodwill carrying amount, accounting policy, and impairment disclosure. The note shows the cash-generating unit and key assumptions behind the valuation. What is flagged Specific area of highest misstatement risk What was done Auditor’s work to address the risk ¨C47C Where to read more Turn to the referenced note(s)

Going-concern paragraphs are separate from KAMs. The standard auditor’s opinion is clean boilerplate. When an auditor adds a “material uncertainty related to going concern” paragraph, it is the single most important paragraph in the document. Hyflux’s annual reports, read in retrospect, show how a deteriorating going-concern position telegraphed the eventual collapse. Most readers didn’t notice those paragraphs at the time. They were there.

Red flags as patterns

Patterns, not company names. Knowing them builds calibration — the internal model of good disclosure, so deviations register.

Auditor resignation or change mid-year. Routine explanations exist. It almost always warrants investigation before dismissing it.

A “material uncertainty related to going concern” paragraph appearing for the first time after years of clean opinions. Not boilerplate. The auditor’s assessment of viability is in doubt.

Frequent prior-period reclassifications across multiple consecutive years. Restating comparatives is often required by new accounting standards — normal once. Repeatedly is not, and it frustrates any attempt to build a consistent historical picture.

Related-party transactions that are large, recurring, and where the related party is also a lender or guarantor. The S-chips episode of the early 2010s featured several variants: large RPTs, related-party financing, and opaque holding-company chains above the listed entity. The pattern persists as a genre.

Revenue growth without corresponding operating cash flow growth. If revenue is being booked without converting to cash, the receivables note and cash flow statement will both show it.

Sudden segment reclassifications or renames. Sometimes legitimate. Worth asking about — particularly when the reclassification improves the look of the prior-year comparative.

Capital commitments materially larger than prior annual capex. Not bad in itself, but it signals a step-change in the asset base and alters the forward free cash flow profile.

Most companies are honest, most reports are normal. The reader trained on good disclosure feels when something deviates. The reader who has only read highlights pages doesn’t.

What changes for banks and REITs

The framework above is built around industrial and general commercial companies. Two sectors need a materially different lens.

Reading priorities by company type Industrial / General Bank REIT ¨C67C ¨C68C ¨C69C Segment note (customer conc., geo split) Trade receivables vs revenue growth Operating CF vs net income (3-yr ratio) Capital commitments vs annual capex Revenue recognition policy + changes RPT note (counterparties, amounts, terms) KAMs (often: goodwill / PPE impairment) NIM and trend (net interest margin) CET1 ratio (MAS Notice 637 minimums) NPL ratio; Stage 3 loans; ECL assumptions Non-interest income share Skip: DCF, EBITDA, FCF multiples RPTs less critical (tightly regulated) KAMs (almost always: loan provisioning) DPU — sustainable? one-offs? capital top-ups? Gearing ≤50%; ICR ≥1.5x (MAS, Nov 2024) WALE table (weighted avg lease expiry) Income support — find the expiry date NAV per unit; valuer’s cap rate assumptions Manager fee structure (base + perf fee) KAMs (often: property valuation)

Banks

The income statement hierarchy that works for an industrial company doesn’t apply to a bank. There is no gross margin in the conventional sense. Debt is the bank’s raw material, not a funding cost.

Net interest margin (NIM) is the most important single metric — interest earned on assets minus interest paid on liabilities, as a percentage of average interest-earning assets. The bank’s operating-margin equivalent. Trend matters more than absolute level.

Common Equity Tier 1 (CET1) ratio is loss-absorbing capital as a percentage of risk-weighted assets. Minimum requirements for Singapore’s Domestic Systemically Important Banks (D-SIBs) are set out in MAS Notice 637.

Non-performing loan (NPL) ratio is the proportion of gross loans flagged as unlikely to be repaid in full — the bank’s credit-quality scorecard.

Stage 3 loans and ECL provisioning. Under SFRS(I) 9, loans are staged by credit-risk deterioration. Stage 3 is the impaired bucket. Expected credit loss (ECL) assumptions are management judgement and can move earnings materially.

Non-interest income share — fee income, trading income, wealth management — measures how much earnings depend on the interest rate cycle. Higher and stickier is better.

What not to do for a bank: DCF, free cash flow analysis, EBITDA multiples. Debt is the bank’s input, not a cost to service.

REITs

Distribution per unit (DPU) is the starting point, but the right question is whether it’s sustainable and what it comprises. Check whether the headline includes one-off distributions, capital top-ups, or anniversary payments that won’t repeat.

Cash flow statement plus the WALE table. Weighted average lease expiry (WALE) tells you how long current rental income is contracted for. Short WALE means more re-leasing risk in the near term, particularly in soft markets.

Gearing and the MAS framework. MAS revised its leverage requirements on 28 November 2024 (Code on Collective Investment Schemes, Property Funds Appendix). The current framework applies a single aggregate leverage limit of 50% to all SGX-listed REITs, with a minimum ICR of 1.5x required of all REITs regardless of gearing level. The previous two-tier structure — under which the 50% cap was conditional on maintaining a 2.5x ICR — has been removed. [Source: MAS, “MAS Rationalises Leverage Requirements and Introduces Additional Disclosures for REITs,” 28 November 2024, mas.gov.sg]

ICR and the debt maturity profile. ICR is EBITDA divided by interest expense — the margin of safety on the interest bill. A REIT with a comfortable ICR and predominantly fixed-rate debt is in a fundamentally different position from one near the minimum with a floating-rate book heading into a tightening cycle. Both numbers are in the financial review section and the debt note. From FY2025 annual reports onwards, MAS requires REITs to publish an ICR sensitivity table in their results disclosures — look for it in the notes as a new standard item.

NAV per unit and the price-to-NAV ratio. Independent property valuations underpin the NAV. Read the valuer’s note for cap rate assumptions — they are a direct lever on the reported asset value.

Income support, earn-out arrangements, and rental top-ups — and their expiry dates. Common when a sponsor injects a property and supplements initial income to make the yield look attractive. Distributions run high for two to three years, then support expires and the underlying DPU falls off a cliff. The expiry date lives in the footnote, not the headline DPU figure. Always find it.

Two further cautions: DRPs used heavily can mask cash distribution shortfalls, and repeated equity-raising at discounts dilutes existing unitholders — even when each transaction looks small.

Quick checklist

Eleven questions to work through before closing an annual report:

Did the auditor flag any KAMs? What were they, and what did the auditor do about them?
Did the audit opinion change — clean to qualified, or a going-concern paragraph for the first time?
Did revenue and trade receivables grow at roughly similar rates?
Did operating cash flow track net income over three years?
Are there material related-party transactions? Who are the counterparties, and is any counterparty also a lender or guarantor?
Did any accounting policy change this year, particularly revenue recognition?
Are segment names, reporting structures, or prior-year comparatives restated? If so, why?
What is customer concentration in the largest segment, and how has it shifted?
What are the capital commitments not yet on the balance sheet, and how do they compare to last year’s capex?
Did the chairman’s letter address anything the notes don’t support — or omit something the notes do disclose?
From the shareholding statistics: are founders and directors maintaining meaningful stakes, and what does top-20 concentration imply about float?

Closing thought

Most companies are honest. Most annual reports are normal. Reading closely is about calibration, not paranoia. After enough reports with good disclosure, you develop an internal model of what that looks like — the specificity, the consistency between sections, notes and management commentary that reinforce each other. When something deviates, something registers. That discomfort is the prompt to look further.

The annual report is the cheapest, most thorough piece of due diligence available to any investor. Reading it well is a durable advantage against being wrong for avoidable reasons.

Position disclosure: Not applicable. This article is educational content. The Butt Man does not hold a position in any specific stock discussed as an example. Where Frencken Group (SGX: E28) and Micro-Mechanics (Holdings) Ltd (SGX: 5DD) are referenced, they are used solely as illustrative examples of annual report disclosure — not as recommendations.

Standard disclaimer: The information in this article is provided for educational and informational purposes only and does not constitute investment advice, an offer to sell, or a solicitation of an offer to buy any security. Past performance is not indicative of future results. All views are those of the author. Readers should conduct their own research and consult a licensed financial advisor before making any investment decision. The Butt Man and stockbutts.com accept no liability for losses arising from reliance on this content.

Disclaimer

The Butt Man

An ordinary man with extraordinary inspiration