# Four public fund pages: detailed UX comparison

Research captured 21 September 2026, approximately 19:45–19:58 SGT. Public, signed-out pages in the Codex in-app browser. This is a research assessment and proposed product direction, not an implementation specification or investment recommendation.

**Read the [visual report](index.html)** for original screenshots, annotations, side-by-side comparison and a searchable field inventory. The [structured inventory](field_inventory.json) records individual labels and their context; the [capability matrix](comparison_matrix.json) aligns the sites without treating unobserved features as absent. Evidence IDs below resolve to original images and captured page text in the report.

## Executive assessment

**Endowus is my best overall visual/usability reference for FundPilot. Morningstar is the strongest analytical reference; FSM is the strongest operational factsheet reference; POEMS is the quickest source of compact platform-specific dealing tables.** These are subjective judgments for this four-page sample, not ratings of the firms, their investment quality or their complete products.

Endowus has the clearest reading sequence: understand the fund, inspect performance, examine exposures, understand cost, then see administrative facts. Plain-language return-basis text, expandable details and simple cards lower the effort of interpretation. Its substantial investment sidebar and promotional content consume space, and its narrow-screen metadata is imperfect. Morningstar provides substantially richer peer, risk, style and sustainability analysis, but advertisements, long tables, abbreviated labels and many nested controls make the first encounter harder. FSM comes closest to an all-purpose Singapore distribution factsheet: research, eligibility, charges, minimums, processing timelines, history and documents coexist. Its density and sticky navigation compete with chart space. POEMS organizes practical information into recognizable tabs and matrices, but the embedded chart and outer page feel like different interfaces, and the mobile table is cramped.

For FundPilot, combine **Endowus's hierarchy, Morningstar's explicit analytical context, FSM's dated operational/document detail and POEMS's funding-method matrices**. Do not copy all four sites' fields into one default screen. Put source dates and definitions beside the measures that require them, with deep analysis available on demand.

## Exact identities, locale and comparison limits

| Page inspected | Identity actually observed | Conclusion |
|---|---|---|
| [Morningstar](https://global.morningstar.com/en-ea/investments/funds/0P00006G01/quote) | LionGlobal China Growth Fund SGD; SG9999002463; Fund • SGD; identifier 0P00006G01. Asia English edition, Singapore home-market banner. Financial Professional role selected; no account created. | Exact ISIN/currency agrees with POEMS. MS02, MS08, MS26. |
| [FSM Global](https://fsm.global/sg/funds/factsheet/370015/LionGlobal-China-Growth-SGD) | LionGlobal China Growth SGD; Lion Global Investors Limited; Equity / General / Greater China; SGD price; page code 370015. Singapore locale. | Name, manager, currency and objective are consistent with the same fund. **No ISIN was exposed in the inspected rendered sections**, and its document click did not yield a verifiable document. Exact share-class match remains corroborated but not independently established by an FSM ISIN. FS01–FS07. |
| [POEMS](https://www.poems.com.sg/fund-finder/lionglobal-china-growth-fund-526050/) | LionGlobal China Growth Fund; SG9999002463; SGD; code 526050 in the embedded chart. Singapore English page. | Exact ISIN/currency agrees with Morningstar despite shorter title. PO02, PO09. |
| [Endowus](https://endowus.com/investment-funds-list/lionglobal-singapore-trust-fund-SG9999002406) | LionGlobal Singapore Trust Fund; SG9999002406; SGD (Acc); oldest-class ISIN identical; inception 3 March 1989; Singapore page. | Intentionally a different fund. Compare interface and disclosure patterns, not its returns, risk rating, holdings or fees with the China fund. EN03. |

Even the matching ISIN does not make every displayed figure directly comparable. Morningstar's related-investment inception is 8 April 1994; FSM and POEMS show launch on 10 March 1994. The report preserves these labels instead of deciding which date is correct. Fund launch, class inception and available history are separate concepts.

## Method and coverage

All eight Morningstar primary tabs were visited or attempted; Analysis redirected to account creation. All six FSM tabs and all six POEMS tabs were inspected. All five Endowus section links were inspected, including expanded description, risk/calendar returns, fund facts, three exposure views and cashback explanation. Representative chart modes, period controls, metadata, table alternatives, price-history search and calendar expansion were exercised. The POEMS chart was also opened directly at its exact observed iframe URL to inspect menus that were difficult to operate inside the host page; those screenshots are labelled standalone chart evidence.

The evidence register separates **observed content**, **observed control with unverified outcome**, **inaccessible content**, **conditional content**, and **inference**. A dash, an empty component, a sign-up gate and an untested feature are different findings. Captured DOM text includes off-viewport rendered sections; viewport screenshots alone do not prove everything in the text was simultaneously visible. Early accessibility captures sometimes contained diffs; later captures use full DOM snapshots. Loading/duplicate/mispositioned shots are retained as originals but excluded from the curated comparison where appropriate.

Original evidence includes tool-generated filename extensions. Desktop evidence uses the normal browser viewport (roughly 1280×720). The capture tool returns images at varying dimensions; the evidence register records their decoded pixel sizes. Original image bytes are preserved, including JPEG content saved under the capture helper’s .png filenames. A bounded inspection using a requested 390×844 viewport covered Morningstar summary/header, FSM overview and its mobile charges panel, POEMS overview/chart and performance table, and Endowus overview/performance. This is not a complete mobile or WCAG audit. No screen-reader session, exhaustive keyboard traversal, contrast measurement, actual touch device, signed-in workflow, transaction, chart export or comparison portfolio was tested. Hover-only point tooltips were not exhaustively tested; the supported browser interface did not expose a hover action. Accessible chart/table equivalents were used where available.

Document links were inspected, with representative opening attempts. POEMS's factsheet destination was blocked by the browser (`ERR_BLOCKED_BY_CLIENT`); Endowus's public document destination remained blank; FSM's factsheet click produced no new verifiable document tab. Therefore document labels/dates are observed, **document contents and share-class identifiers within those documents are not verified**. These are environment/session outcomes, not claims that the documents are broken for customers. No access controls were bypassed.

The indexed FSM page was consulted only as a secondary check. It returned older figures and one-day date differences relative to the live browser. It is not used to overwrite this session's live evidence.

## Morningstar: what appears and how it works

### Summary and performance

The masthead combines fund name, full ISIN, four-star image, Medalist unlock action, watchlist button and SGD. A horizontal eight-tab bar separates Summary, Chart, Analysis, Price, Risk, Portfolio, Parent & People and Sustainability. A location banner and advertising reserve significant vertical space. The primary content occupies roughly two-thirds of the desktop width, with a right advertisement rail. MS02–MS03, MS26, MS-summary-full.

Overview is a horizontal strip/carousel of compact label/value pairs: NAV, one-day return, category, total assets, adjusted expense ratio, prospectus net expense ratio, distribution fee level, load type, equity style, SEC yield, TTM yield, initial minimum, turnover and open/closed status. Several are dashes for this fund. The title currency does not remove the need to check a particular component's metadata. The strip's forward control makes some fields less discoverable than a small wrapped grid.

The Returns view stacks a growth-of-10,000 chart, manager-change/restructure markers, investment-flow subchart, annual-return table and trailing-return table. Investment, category and index have separate colored lines and comparison rows. Annual columns run 2016–2025 plus YTD; rows include return, quartile glyph, percentile rank, category population and category name/code. The category code `EU0500` is less understandable than a plain name. Long tables scroll horizontally. Trailing periods are 1 month, 3 months, YTD, 1/3/5/10/15 years and earliest available; Day End / Month End / Quarter End is a real selector, and Day End was selected. Quarter End initially showed 30 June 2026. Reversal and metadata controls sit beside the table headings. MS25–MS26.

Distributions is an alternate view, not a statement of current payout policy. It shows two old events (1996 and 1997), with distribution date, distribution NAV, income, short/long-term capital gains, return of capital and total, denominated SGD. The annual-distribution component explicitly says no data available. This illustrates why a currently non-distributing fund can still have historical events. MS24.

Strategy is a prose objective followed by a related-share-class table: investment name, ticker, inception, net assets, front/deferred loads, expense ratio, minimum initial purchase, 12b-1 actual, purchase constraint and share-class attributes. Both SGD and USD rows appear. This is **share-class comparison**, unlike FSM's peer-fund list.

### Chart workspace

The dedicated chart is much more configurable than the summary graphic. Controls include comparison search; Data Type, Events, Indicators, Fundamentals, Display, Drawings, image generation, clear and settings; period 1M/3M/YTD/1Y/5Y/MAX; start/end date pickers; Daily/Weekly/Monthly frequency; currency; and chart/table switching. A table alternative was opened. The plotted legend initially used `0P00006G01`, which is less helpful than the fund name. MS16–MS22.

Opened menus reveal price, price-with-dividend, NAV, NAV-with-dividend, post-tax NAV, percentage-change/growth variants and an initial-value input; dividend events; moving-average/Bollinger overlays and MACD/momentum/ROC/RSI lower indicators; dividend yield and rolling dividend; Mountain/Line/Dot/AboveBelow display, linear/log scale and earliest/common/main-ticker start choices. These are **exposed generic chart options**, not proof that every series is meaningful or populated for this fund. Drawings/settings/export and adding a second security were not exercised. The presence of an AI Summary button elsewhere was recorded without invoking it.

### Risk

Risk has 3/5/10-year tabs, a portfolio-risk gauge, category-relative risk/return scales, a comparative measure table, risk/return scatterplot with table alternative, capture ratios and drawdown detail. The observed 3-year view includes alpha, beta, R-squared, Sharpe, standard deviation, upside/downside capture, maximum drawdown, peak, valley and duration. Investment/category/index columns help put a figure in context. MS05–MS07.

The metadata is unusually valuable: portfolio-risk score is dated July 2026 and explicitly does not inherit the 3/5/10-year period; other risk figures are dated August 2026. The comparison index is Morningstar China TME GR USD while the calculation benchmark is MSCI Golden Dragon NR USD. The metadata also supplies category population and SGD context. FundPilot must not collapse those index roles into one “benchmark” field.

### Portfolio

Asset allocation has net/short/long columns for equity, fixed income, other and cash, with economic-exposure control. Stock style has a graphical map; Factor Profile shows style, yield, momentum, quality, volatility, liquidity and size against category and historical range. There are 5-year and relative-category controls, metadata and an uninvoked AI Summary. MS08.

Style Measures compares investment, category and index for P/E, P/B, P/S, price/cash flow, dividend yield, long-term/historical earnings growth, sales, cash-flow and book-value growth. Market Cap switches to average market caps and giant/large/mid/small/micro percentages; the switch was verified. Exposure switches among sector, country and region; Country was exercised. Sector rows are grouped into cyclical/sensitive/defensive. These are deeper look-through analytics than the other inspected pages expose. MS28–MS32.

Holdings includes equity/bond/other/total counts, top-ten concentration, reported turnover with its own date, active share, women directors/executives percentages, then ten visible positions with weight, first bought, market value/date, currency, share change, one-year security return, forward P/E and sector. Holdings and active-share dates and active-share index are stated separately. Horizontal scrolling is required. “75 holdings” does not mean 75 rows are publicly displayed: the page expressly limits display to the ten largest. MS08, MS-portfolio-full.

### Price, people and sustainability

**Price means fees here**, not the price-history chart. Maximum front/deferred/redemption loads and minimum purchase lead into ongoing-fee comparison and a hypothetical cost illustration for SGD10,000, 5% return, 1/5/10 years. The table separates purchase, ongoing, redemption and total costs. A conspicuous caution is warranted in interpretation: the same page shows a missing current expense ratio and zero ongoing costs in the illustration. Do not infer a zero-fee fund from that illustration. MS13.

More Fees / Expenses Data expands five years of annual-report expense ratios against category/comparison group, actual and maximum management charges, maximum 12b-1/admin fees, waivers, three-year tax-cost ratios and potential capital-gain exposure. Most tax fields are blank for this fund. MS15.

Parent & People contains largely missing parent aggregates and a no-data rating-breakdown component, while People has inception, adviser, number of managers, tenure, named manager/start date and a timeline with gender legend. The named-manager biography link was observed but not expanded. Do not equate a missing parent block with missing individual-manager information. MS09.

Sustainability shows ESG risk rating, historical/current corporate and sovereign score sections, low-carbon designation, carbon-risk score and comparison context. Product involvement and fossil-fuel charts have table alternatives, both opened. Tables include exclusion status, portfolio/category percentages and, for fossil fuels, coverage. Business-practice, weapons, nuclear, environmental, health/life and values categories are listed individually in the structured inventory. Unresolved template strings and an unnamed/“undefined” control appear in accessibility text; this is a concrete accessibility concern, not a tested screen-reader verdict. MS10–MS11.

Analysis redirects to free-account creation; no analyst report was accessed. Medalist unlock and watchlist are visible actions, not verified anonymous capabilities. MS12.

## FSM Global: what appears and how it works

### Overview and operational sidebar

FSM presents asset/sector/geographic chips, uppercase title and manager, a large SGD NAV, alert/favourite icons and Buy/RSP actions above six tabs. The shrinking sticky header keeps identity, price and navigation available while scrolling. The left main column holds objective and a 12-item summary grid; a right rail holds payment eligibility, recommendation status, buy/sell timelines, risk, holidays, tools and related articles. This is information-rich but leaves less room for charts. FS01, FS08–FS09.

Summary fields are AUM in SGD millions, FSM risk rating with colored segments, launch date/price, historical dividend, yield, frequency, CPF risk classification, initial minimum, RSP minimum, RSP special-list status and annual expense ratio. Missing launch/dividend fields use a dash. AUM does not have an equally prominent independent date in this grid. “Last Updated” at the top includes Singapore time; it must not be interpreted as every metric's valuation date.

Payment method states Cash/SRS/CPFIS-OA/CPFIS-SA yes/no. Buy and sell processing are illustrated as connected dated stages: placement, transact, holdings completion or payout. The sidebar shows 3-year annualised volatility/Sharpe and 1/3-year max drawdown. The holiday card contains both a “no upcoming” Singapore message and separately labelled fund-holiday dates; retain that distinction and do not simplify the card to “no holidays.”

Price/Performance switches the chart basis. The 5-year performance mode was selected and explicitly states SGD with dividends reinvested. Default periods are 1W/1M/3M/6M/1Y/YTD, with More exposing 2/3/5/10Y. Summary change is a currency amount in price mode and a percentage in performance mode. The green filled chart has a time axis and Highcharts menu; Compare and export-like controls are present. Clicking Compare did not produce a verified comparison selection/result in this session, so only the control is established. FS09–FS12.

Annualised Returns uses positive/negative columns across 1 week through 10 years, followed by calendar-year bars for 2021–2025. Short periods appear under the broad annualised heading; FundPilot should explicitly distinguish cumulative short-period figures. Historical price cards give high/low values and dates for 1Y/3Y/all time. Documents are dated PDF-style tiles: prospectus, semiannual, annual, PHS and factsheet. FS01, FS13.

### Information, review, history, related funds and charges

Information uses label/value pairs for asset class, sector, geography, RSP list, MAS registration, right to cancel, forward pricing, initial/subsequent/redemption/RSP minimums, minimum holding, charges flag and dividend flag. Fund-house narrative and address/phone/fax/email/website follow. That narrative contains a 2022 AUM date; it is older context, not a current manager AUM estimate. FS02.

Top-ten holdings combines names, numeric weights, in-cell horizontal weight bars and a sector column that is blank for these rows. The heading gives total top-ten concentration and 6 September 2026 date. Country/region and sector sections show pie-style visual breakdowns with textual legends. Each states that figures were extracted from the PDF and refers discrepancies to the factsheet. That source caveat is valuable; the date must remain attached to the specific snapshot. FS14, FS-information-full.

Fund Review is an editorial layer: objective, 2026/2025 heading, investment approach, market conditions where the strategy may fare better or worse, investment style, manager context and a fund-versus-peer table. The table includes three-year annualised return, max drawdown, downside deviation and Sortino. Its stated basis is monthly total returns including gross dividends in SGD as of 31 March 2026, unlike the latest sidebar figures. Recommendation/review content is fund-conditional, not proven for every FSM fund. FS04.

Price History offers last-three-month daily observations with quick search, sort affordances, export, page size and pagination. Columns are row number, NAV(SGD), currency change, percentage change and NAV date. A search for `18 Sep` reduced the displayed rows to the single matching price, while “Total of 62” remained visible; that total is not the filtered-row count. FS05–FS06.

Related Funds is a wide peer table with fund/manager, YTD, 1W/1M/3M/6M/1Y/2Y/3Y/5Y/10Y and action. It includes the subject fund and peer funds, with sort arrows and pagination. The displayed “Total of 9 entries” sits alongside ten visible rows including the subject; interpret the count cautiously. Transaction and favourite actions were not exercised. FS07.

Charges explicitly separates FSM fees, other charges/processing and fund charges. Sales/RSP/switching charges are 0%; CPF platform fee 0%; Cash/SRS platform fee 0.0875% per quarter; Diamond tier exception and GST caveat are stated. Buy/sell processing uses T+4/T+7 business days. Fund charge fields are actual management charge, mother-fund charge, dated annual expense ratio (31 December 2025), prospectus and other-significant-fees source. The page explicitly explains that fund expenses are already included in NAV. The operational rates are observations, not a current fee quotation for a transaction. FS03.

At 390px, the overview becomes a two-column metric grid. A fixed bottom Buy/RSP/minimum panel reduces reading space but exposes **Charges & Payment Method** as a mobile sheet; opening it produced a readable stacked fee and eligibility view. The desktop Charges tab was not present in the visible mobile tab set. That is changed navigation, not absent fee information. FS-mobile-loaded, FS-mobile-charges.

## POEMS: what appears and how it works

POEMS has a persistent blue utility/login bar, white navigation, breadcrumb, fund title and a linked PSPL risk classification (“High”). Six tabs are General Info, Performance, Dividend Info, Documents, Regular Saving Plans (RSP) and Fund Manager Info. Orange marks the selected tab and actions. The content is a white panel on a grey background; floating app/chat promotions overlap the right edge and, on mobile, lower content. PO02, PO04, PO-mobile-performance.

General Info starts with objective, a large third-party-style chart iframe and Buy Now / Compare with Peer Funds. The chart uses a blue area line and miniature navigator with range handles. It has a fund code/title, axes and latest-price marker. Its menu offers 1/2/3/5/10/15/20-year windows, Unit Price/Performance, crosshair and settings. Direct chart inspection verified 5Y and Performance. In performance mode the baseline is 100%, so a reading near 120% is an indexed level, not a 120% gain. This needs clearer wording in FundPilot. Settings expose theme, plot type, crosshair, snap-to-price, projection percentage, gridlines, navigator, legend and quarter lines. Image export and print are visible but untested. These are menu/control observations, not a claim of validated projection logic. PO-chart through PO-chart-performance.

Below the chart, Fund Information is a two-column label/value block: asset class, sector, geography, launch date/price, ISIN, currency, funding source and Singapore dealing cutoff. Funding and charges are matrices with Cash, CPF OA-IA, CPF SA-IA and SRS-IA columns. They show initial/subsequent investment minimums and redemption units; sales/switching, redemption/realisation, annual management and expense ratio. N/A is used explicitly. A neighbouring return-metrics table lists return, standard deviation, Sharpe and Sortino for 1/3/5/10 years. Annualisation and source disclaimers appear beneath. PO02, PO09, PO-general-full.

Performance replaces the content with one wide table: YTD, 1W/1M/3M/6M/1Y/3Y/5Y/10Y; columns performance, standard deviation, Sharpe and Sortino. Short-period risk cells are N/A. The explanation says returns beyond one year are annualised and risk measures require at least one year and may be unavailable depending on inception. No metric-specific as-of date was visible beside this table in the inspected state. It is compact and easy to scan, but four decimal places for return imply more precision than many readers need. PO04.

Dividend Info has 12-month yield, a crediting-time explanation and searchable/paginated history with ex-date, payment date and payout per unit. Two 1996/1997 rows are visible, with missing payment dates; yield is N/A. This is a genuine historical section, despite a current non-distributing interpretation elsewhere. Page-size choices are 10/25/50/100, with disabled navigation for the two-row result. PO05.

Documents gives four labelled view links—prospectus, PHS, semiannual/annual reports, factsheet—without visible document dates. The representative factsheet opening was blocked in this environment. RSP gives a fund row, funding columns, monthly/quarterly amounts, initial-investment-required field, one-year return and Subscribe. Empty CPF cells still contain `(M)/(Q)` labels; the inspected view does not explain these as eligibility. Do not treat them as valid zero-minimum offers. PO06–PO07.

Fund Manager Info contains logo, name, address, phone, fax, website and a narrative with a June 2024 AUM reference. It describes the management company rather than named fund portfolio managers. PO08.

On mobile, tabs wrap into several rows, reducing above-the-fold space. The embedded chart eventually renders and offers a simplified toolbar. The five-column performance table compresses into the narrow screen with wrapped headers/period labels and a floating promotional widget over lower rows. This is usable for checking a value but the weakest reading experience in this sample. PO-mobile-loaded, PO-mobile-performance.

## Endowus: what appears and how it works

The hero leads with a small ISIN and large serif fund name, asset/geography/funding/accumulation chips, and a compact funding/payout/risk panel. Five section links become sticky while the page scrolls. A card-based main column is paired with an investment CTA and transfer promotion. The document is one long reading flow, rather than six separate content panels. EN01, EN08, EN14.

About this fund contains a truncated objective with working Read more, four document links, and an **Our perspective** editorial carousel. Its six topics are Product, People, Performance, Process, Cost and How it fits in a portfolio. This supplies a reason for researching the fund, named management context, strategy and portfolio-role/risk discussion. The captured rendered text includes the cards; individual carousel/read-more states were not exhaustively exercised. The content is Endowus's opinion and fund-specific curation, not an independently verified endorsement or proof every fund has all six cards. EN13.

Performance clearly separates Returns and NAV with an explanatory sentence. For this accumulating class, the return description says dividends are reinvested; NAV mode explains price changes for the accumulating fund. The line chart uses All Time/1Y/3Y/5Y, and 1Y NAV was verified. Daily-data date is stated. There is no peer/benchmark overlay in the inspected chart controls; the editorial text discusses a benchmark separately. EN04–EN05.

Below the chart, Total returns has Cumulative/Annualised modes. Cumulative shows all-time, YTD, 1 month, 6 months and 1/3/5/10 years. Annualised suppresses the subannual periods and shows all-time and 1/3/5/10 years. Switching changed the displayed numbers; some underlying radio checked states were inconsistent with the visible result, so verification relied on content. Distribution text explicitly explains accumulation at fund level. EN02, EN11.

Show risk and calendar returns expands max drawdown, volatility and a month-by-year matrix with annual total, signed red/green values and dashes for unavailable months. Load more increased the initial five years to ten out of 38. The calendar needs horizontal scrolling inside the main column even on desktop. Maximum-drawdown help defines the greatest peak-to-trough loss over the portfolio's history and links to methodology; changing the chart to 1Y did not change the displayed historical risk values in this session. Do not silently interpret those metrics as one-year risk. Volatility's exact period/methodology was not verified. EN02, EN11–EN12, EN15.

Top exposures switches Holdings/Regions/Sectors. Each is a simple two-column name/percentage table rather than a pie chart. Up to 15 positions are shown; the explanatory note says weights need not total 100% and may include negative values. That is clearer than forcing a complete-looking pie from partial holdings. A holdings as-of date was not visible beside this table in the inspected state; the fund-size date elsewhere must not be borrowed. EN06–EN07, EN14.

Fund fees displays a struck-through 1.39%, 0.77% p.a. after cashback and “Save 44%.” The fee definition says it is the fund expense ratio, embedded in NAV. A cashback modal explains trailer-fee rebates. This fund-fee figure is not an all-in Endowus service-cost quotation. The adjacent generic “Save up to 65%” promotion describes another comparison and should not be confused with the fund-specific 44%. EN08–EN09.

Fund facts initially shows ISIN, house and fund size/date. Show more adds fund inception, class currency, share class, oldest share-class ISIN/inception and daily pricing frequency. This is the strongest explicit distinction between selected class and oldest class in the four-page sample. EN03.

At 390px, the large title wraps to three lines, the fund metadata is somewhat cramped, section navigation scrolls horizontally, investment CTAs move into the vertical flow, and chart periods become a dropdown. Text and cards remain more comfortable than POEMS's table. This is a bounded observation, not evidence all mobile states are polished. EN-mobile, EN-mobile-performance.

## Common, partial and distinctive features

| Capability | Across the inspected pages | What matters for FundPilot |
|---|---|---|
| Name, currency, objective, performance and risk context | All four, with different labels and granularity | Common baseline, but require exact identity and independent dates. |
| ISIN prominently available | Morningstar, POEMS, Endowus; FSM not observed | Never match solely on name. |
| Price and performance chart alternatives | All four; POEMS verified through exact iframe URL | State price versus reinvested return and indexed baseline. |
| Platform funding/minimums/dealing | Strong FSM/POEMS; Endowus has funding chips; Morningstar has investment minimum/constraints | Company/distributor terms are not universal fund attributes. |
| Fees | All four | Maximum loads, actual distribution charges, TER, rebates and platform fees are different. |
| Holdings/sector/region | Morningstar, FSM, Endowus; not observed in POEMS primary tabs | Partial exposure and as-of labels matter more than decorative charts. |
| Calendar returns | Morningstar annual peer table; FSM annual bars; Endowus monthly matrix; not observed on POEMS detail tabs | Use annual overview first, expandable monthly table. |
| Analyst/selection rationale | FSM and Endowus public fund-specific commentary; Morningstar Analysis gated | Distinguish evidence, opinion and inferred portfolio role. |
| Documents | FSM/POEMS/Endowus links; no document library observed on Morningstar's inspected public tabs | Dated source documents are an important practical feature. |
| Detailed peer analytics | Morningstar strongest; FSM related-fund and dated review tables; POEMS outbound compare control; Endowus no peer overlay observed | Comparing funds needs aligned period/currency/basis, not just columns. |
| ESG, factor/style, capture ratios, active share | Observed on Morningstar only in this sample | Specialist optional layer, subject to entitlement/methodology. |
| Buy/sell processing timelines and fund holidays | FSM distinctive | High operational value, but requires platform-owned data. |
| Funding-method charge/minimum matrices and RSP schedule | POEMS especially compact | Good adviser reference, below core research. |
| Portfolio-fit narrative and trailer-rebate explanation | Endowus particularly strong | Helpful explanation patterns; do not copy marketing opinion as neutral data. |

“Only observed” means exactly that. It does not mean a competitor has no comparable feature elsewhere, behind login, for a different fund or in another locale.

## Important mismatches and non-equivalences

1. **Fund expenses:** FSM shows annual expense ratio 1.37% dated 31 December 2025; POEMS shows 1.64% without an adjacent date; Morningstar historical table shows 1.64 for 2025 but a missing current ratio. This research does not establish the reason or the authoritative value. Preserve report basis/date and investigate before importing.
2. **Maximum versus actual fees:** Morningstar's front load 5% and redemption load 1% are maximum-sales-fee labels. FSM and POEMS advertise platform sales/switching 0%. That can reflect different semantics rather than contradictory prices.
3. **Eligibility:** FSM says CPFIS-OA Yes; POEMS lists Cash/SRS and N/A in CPF columns. Treat eligibility as distributor/context-specific pending confirmation, not something to normalize by majority vote.
4. **Return dates:** Morningstar initially presents quarter-end trailing returns but a daily NAV; FSM's latest NAV and return panels need separate source-date checking; POEMS's table lacks an adjacent as-of date; Endowus prominently says daily data/date. Equal labels alone do not establish equal windows.
5. **Historical payouts:** Morningstar and POEMS retain 1996/1997 events; FSM's current Dividend flag is No. Payout policy and historical event existence require separate fields.
6. **Launch/inception:** 10 March versus 8 April 1994 remains unresolved. Do not use metadata inception as a guaranteed chart start.
7. **Risk scales:** FSM 8, Endowus 6, PSPL High, Morningstar stars/category risk/portfolio-risk score are different scales with different purposes. Stars are not a universal risk scale; the different Endowus fund also prevents numerical comparison.
8. **Holdings:** FSM's September extracted holdings and Morningstar's June holdings have different weights and dates. The lists are partial and taxonomies differ. Endowus's separate Singapore fund cannot be compared as a holdings-data discrepancy.
9. **Editorial age:** FSM's March research table, its older manager narrative, POEMS's June 2024 company narrative and current quote panels coexist. A page-level refresh date does not refresh prose automatically.

## Visual and usability judgment

| Dimension | Best reference in this sample | Reason and tradeoff |
|---|---|---|
| First reading / hierarchy | Endowus | Clear cards, readable prose and progressive disclosure; long page and large CTA rail. |
| Specialist research depth | Morningstar | Rich comparative measures and metadata; steep terminology/control density. |
| Singapore operational usefulness | FSM | Eligibility, minima, settlement-style stages, holidays, fees and dated documents; crowded sticky layout. |
| Fast platform terms lookup | POEMS | Compact funding matrices and familiar tabs; small text and uneven styling. |
| Return-basis discoverability | Endowus | Plain sentence beside chart and distinct annualised toggle; risk lookback still needs investigation. |
| Detailed source/method context | Morningstar | Explicit dates/index/calculation benchmark in metadata; key context can be hidden behind icons. |
| Holdings overview | FSM / Endowus | FSM adds weight bars and concentration/date; Endowus is cleaner and explains partial exposure. |
| Deep holdings inspection | Morningstar | Currency, purchase date, market value and security metrics; wide scrolling table. |
| Mobile reading | Endowus, provisionally | Better reading flow in sampled states; metadata/tab issues remain. FSM's mobile charges sheet is a good specialist pattern. |
| Accessible alternatives | Morningstar, for charts tested | Named chart-to-table controls and explanatory table captions; also observed unnamed/template-string defects. Not a compliance ranking. |

If forced to select **one fund-detail design for an adviser product**, I would start with Endowus, then add an explicit, compact “Research details” layer. If the criterion is **most information available to a specialist without switching products**, Morningstar wins. For a Singapore adviser checking whether/how a client can place an investment, FSM wins. Those criteria should not be conflated.

## Proposed FundPilot detail structure and priorities

These are proposed interface design priorities, not a statement that all data is available or that the features are implemented.

**1. Persistent identity header.** Exact class name, ISIN/provider ID, class/quotation currency, accumulation/distribution status, provider and company eligibility. Show latest NAV with valuation date and stale/missing indicator. Primary actions: shortlist, compare and add to Model Studio, preserving research state. Keep transaction-style buttons out unless that becomes an actual product responsibility.

**2. Overview.** Short objective plus a compact grid: asset/category/geographic focus, inception labelled by type, risk measure with scale name, dated expense ratio and fund size only when evidenced. Use an explicit “not supplied” state rather than empty gauges or invented zeroes. Separate source fact from any later generated explanation.

**3. Performance.** Default reinvested total return only after validated mapping; otherwise retain honest NAV-only wording. Offer price/return modes with currency, frequency, start/end bounds, distribution treatment and source date. Add a small trailing-return table with subannual cumulative and multi-year annualised labels. Expand to calendar years/months and a chart/table alternative. Benchmarks should have roles and dates, not a single ambiguous dropdown label.

**4. Risk.** Initial scope: documented three-year volatility, relevant drawdown and available risk classification, each with period, frequency and basis. Add Sharpe/Sortino/capture ratios only when their risk-free rate, downside convention and benchmark contracts are resolved. No averaging fund-level risk to manufacture portfolio risk.

**5. Portfolio exposures.** Dated top holdings with weights, top-ten concentration and partial-data warning; sector/region tables with optional bars. Separate holdings-report date from quote date. Style/factor/ESG tabs are later specialist features, not first-release essentials.

**6. Costs, eligibility and dealing.** Two clearly named blocks: fund expenses included in NAV; platform/adviser/account costs applied separately. Preserve actual versus maximum, annual versus quarterly, rebate and tax basis. Funding-method minimums, cutoffs and dealing calendars belong to company/distributor data, not an assumed Morningstar universal field.

**7. Income, documents and manager.** Current payout policy plus historical distribution events, with ex/payment/reinvestment dates kept separate. Date-stamped document links and source revision. Provider description and named manager information should not be merged. Optional editorial rationale should state its author, date, evidence and limitations.

| Priority | Recommended contents | Current evidence/dependency |
|---|---|---|
| P0: trustworthy core | Exact identity/currency, source dates, objective where available, NAV, category/provider, shortlist and portfolio-building actions, explicit gaps | Validate identity, coverage and freshness before display. |
| P1: useful research | Reinvested history, trailing/calendar returns, period-defined risk, dated fees, documents and distributions | Confirm calculation conventions, source coverage and display rights. |
| P1: operational context | Funding eligibility, minimums, dealing and fund/platform cost separation | Use approved distributor sources; terms differ by platform. |
| P2: deeper differentiation | Dated holdings/exposures, size, named managers and peer comparison | Require reliable source coverage and compatible peer definitions. |
| P3: specialist / editorial | Factors, ESG, capture ratios, valuation, active share and sourced explanation | Additional methodology and licensing validation required. |

The interface must work well with partial or differently dated data. Public competitor pages illustrate presentation patterns; they do not establish another product's data access or redistribution rights.

## Access and follow-up register

- Morningstar Analysis account-gated; no analyst text read. Watchlist, unlock, AI Summary and account actions untested.
- FSM exact ISIN unresolved in rendered evidence; PDF click/Compare outcome unverified. No export or account/transaction action.
- POEMS factsheet blocked in this browser; chart menus verified on the exact embedded public URL. Compare navigation destination observed; no peer comparison execution.
- Endowus factsheet blank in this browser; exact volatility methodology and exposure-report date unresolved; no Add fund/Invest workflow.
- Not an exhaustive permutation test of every chart option, every date or every explanatory icon. No hover/keyboard/screen-reader completeness claim. Mobile coverage is bounded and labelled.
- Research was based on public pages and did not involve account transactions or private provider data.

This public edition preserves the findings and documented public-access limits. Research captured 21 September 2026; published 22 September 2026. Screenshots and brand marks belong to their respective publishers and are reproduced here as evidence for comparative interface commentary. This independent research is not affiliated with or endorsed by those publishers.