Six points where the conventional answer and this one part company — then the same exercise for a fund factsheet and a stock screener.
| On this point | Typical report | FinNexus Terminal |
|---|---|---|
| The score | A proprietary number. The method is a paragraph of description. | Every weight, formula and input printed beside the result — and sliders to re-weight it yourself. |
| Missing data | Quietly omitted, or the metric simply doesn't appear. | Reported as "data unavailable". A check with no data returns N/A, never PASS. |
| Sources disagreeing | One value is shown. You never learn there was a second. | Both values shown with their derivations, and the size of the gap. |
| Technical signals | "MACD is bullish." No indication of whether that has ever meant anything. | Each signal backtested on that stock's own history, with sample size, confidence interval, and the baseline it must beat. |
| The AI section | Generated text. No way to tell which numbers are real. | Written from a sealed evidence pack, every bullet footnoted, then every generated number audited back against its source. |
| The conclusion | "STRONG BUY. Target 1673, 27% upside." | No buy, sell, hold or target anywhere. Analyst consensus is shown as a fact about the market, not as our view. |
| On this point | Typical factsheet | FinNexus Funds |
|---|---|---|
| What it holds | The top ten, by name and weight. The other eighty are a sector pie chart. | Every disclosed line, each matched to a company and scored by the same rules the terminal applies to a stock. |
| Overlap with your other funds | Not addressed. Each factsheet describes one fund as though you own nothing else. | The weight any two funds hold in common, name by name — the usual reason four funds behave like one. |
| Your return | The fund's return since launch, or over the last year. Not yours. | XIRR on your own dated instalments, so a SIP started at the wrong time looks like what it was. |
| How current it is | An as-of date in small type, read as though it were today. | The date, its age in days, and a check that fails outright once the photograph is too old to describe the fund. |
| What was not disclosed | Percentages that quietly sum to less than a hundred. | Undisclosed weight stated and excluded from every percentage, rather than counted as zero. |
| How a holding was identified | A name in a table. | Whether it matched on an ISIN or on a normalised company name, reported separately — name matching is inference. |
| On this point | Typical screener | The quality screen |
|---|---|---|
| The result | A list of survivors. Nothing about the ones that went. | A funnel: how many entered each step, cleared, failed, and could not be tested. |
| A missing figure | Usually treated as passing, or the row is dropped without saying so. | Counted in its own column and dropped, with the count on screen. Treating the unmeasured as clearing the bar lets through exactly the names nobody has figures for. |
| An empty result | Zero rows, and you guess which filter was too tight. | The step that emptied the funnel is named, and near misses — one or two tests failed — are listed beside it. |
| The moving average | A number you cannot reproduce, over an unstated window. | 200-day EMA over the whole stored series, seeded on the first 200 closes and smoothed at 2/201 — so it matches a charting package. |
| A size or liquidity floor | An amount in one currency, so the same filter means a different company in each market — or no floor at all, and the list quietly includes names you could not buy. | A rank inside the stock's own market. No exchange rate to go stale, and a floor that means the same thing in rupees, dollars and pounds. The amount is printed beside the rank. |
| How many filters are on | Every box ticked by default, which looks thorough and returns a list nobody chose. | Eight on, twelve more off until you switch them on — because eight already cut eleven hundred names to a handful. |
| Where a filter's figure came from | A number in a cell. | Each row says whether it was reported or computed, and the two filters reading a stored analysis say when that analysis was run. |
Search any listed name and watch twelve tools run, step by step — or search a fund and see what it holds.
