The checks behind the India Capital Flows tool. This page is a snapshot taken on 15 Nov 2026: monthly data to Oct 2026, daily trades to 12 Nov 2026. The tool's Validation tab runs again on every data refresh and has the row-by-row tables. A new mismatch with no explanation stops the refresh until it is looked at.
For what each source and series means, see the companion India Capital Flows methodology page.
Open the interactive India Capital Flows tool on TREMOR for the live Validation tab.
Summary
| Checked against | What is checked | How much | Result |
|---|---|---|---|
| Another publisher | FPI equity on the stock exchanges: NSDL against NSE, day by day | 19 trading days | Same direction on 19 of 19. Typical gap Rs 674 crore |
| Another publisher | FPI flows each month: NSDL against RBI | 107 months | As NSDL reports it: same direction in 94 of 107, typical gap US$ 793 million. By trading month: 96 of 107, US$ 238 million |
| Another publisher | Mutual funds each month: AMFI against SEBI | 89 months | 512 of 526 figures agree to within Rs 1 crore. 14 differ, each explained |
| Another publisher | The rupee to dollar rate NSDL uses against the market rate | 5,259 days | Typical gap 0.101% since 2020 |
| Two reports, one publisher | NSDL's reports against each other: daily and monthly, country and custody, sectors | 1,898 comparisons | 1,885 agree. 13 beyond rounding, listed |
| Totals | Parts against the total the publisher prints | 2,657 comparisons | 2,652 agree. 5 beyond rounding, listed |
1. Checked against another publisher
NSE, RBI and the currency market against NSDL. SEBI against AMFI.
Foreign investors on the stock exchanges: NSDL against NSE
NSE publishes a provisional figure the same evening, from what trading members report during the day. NSDL publishes the confirmed figure a working day later, from custodians. They are separate counts of the same trades, so they should move together and stay close.
14 Oct 2026 to 12 Nov 2026: same direction on 19 of 19 trading days. Typical gap Rs 674 crore, largest Rs 2,115 crore.
NSE publishes only the latest day, so this comparison starts on the first day it was collected and grows by a row every trading day.
FPI flows each month: NSDL against RBI
RBI publishes its own monthly figure for net foreign portfolio investment, in US dollars, on a balance-of-payments basis. The tool's month can be counted two ways, and both are compared here.
| NSDL month counted as | Months | Same direction as RBI | Typical gap, US$ million | Correlation |
|---|---|---|---|---|
| Trading month (the option) | 107 | 96 | 238 | 0.96 |
| As NSDL reports it (the tool's default) | 107 | 94 | 793 | 0.95 |
Added up over twelve months, the one-day date shift washes out:
| Twelve-month totals, NSDL month counted as | Periods | Same direction as RBI | Typical gap, US$ million | Typical gap, % | Within 10% of RBI |
|---|---|---|---|---|---|
| As NSDL reports it | 96 | 89 | 1,484 | 8.2% | 53 |
| Trading month | 96 | 90 | 1,336 | 8.6% | 52 |
Oct 2017 to Aug 2026. Month by month the trading month is closer to RBI's figure. Over twelve months the two are level.
Why the two differ. Neither RBI nor NSDL publishes a reconciliation of the two that we could find. Part of the gap is timing, and we measured it. NSDL counts each day in the month it reported it, a working day after the trades. Counted by trading month, the monthly figures are much closer. Over twelve months the timing washes out and about 8% remains. We believe the rest comes from how each figure is built. RBI's figure is on a balance-of-payments basis, from reports of money crossing the border. NSDL's figure comes from custodians' reports of FPI trades, converted to dollars at each day's rate.
Mutual funds each month: AMFI against SEBI
The tool's mutual fund figures come from AMFI's monthly report. SEBI publishes its own monthly statistics for the same industry, built from the same returns the fund houses file. Each month is compared for all funds, equity funds and debt funds: assets at month end and money put in, net.
Apr 2019 to Sep 2026: 512 of 526 figures agree to within Rs 1 crore. 14 differ by more, in 13 months, each explained below. 6 figures could not be compared because SEBI's page leaves them out.
SEBI publishes about a month after AMFI, so the newest AMFI month is checked once SEBI's page for it appears.
The 13 months that differ or could not be compared (all funds, net in)
| Month | All funds, net in: AMFI | SEBI | Gap | Gap, % | Why |
|---|---|---|---|---|---|
| Sep 2026 | -1,33,375.57 | -1,33,375.56 | -0.01 | -0.000% | A figure revised after AMFI's monthly report. The gap reverses within two months. |
| Aug 2026 | +41,353.60 | +41,383.43 | -29.83 | -0.072% | A figure revised after AMFI's monthly report. The gap reverses within two months. |
| Jul 2026 | +2,35,902.39 | +2,35,872.59 | +29.80 | +0.013% | A figure revised after AMFI's monthly report. The gap reverses within two months. |
| Jul 2025 | +1,78,794.31 | +1,77,935.34 | +858.97 | +0.483% | A figure revised after AMFI's monthly report. The gap reverses within two months. |
| Jun 2025 | +49,094.61 | +49,953.62 | -859.01 | -1.720% | A figure revised after AMFI's monthly report. The gap reverses within two months. |
| Dec 2024 | -80,354.57 | -20,059.36 | -60,295.21 | -300.584% | Follows from SEBI's November 2024 page: December is worked out as the year to December minus the year to November. |
| Nov 2024 | +60,295.30 | 0.00 | +60,295.30 | SEBI's page for November 2024 repeats October's Grand Total row. Its category rows are November's. | |
| Oct 2024 | +2,39,828.69 | +2,39,828.78 | -0.09 | -0.000% | SEBI's page for the month before has no equity sub-total, so the equity net figure cannot be worked out. |
| Sep 2024 | -71,114.08 | -71,114.15 | +0.07 | +0.000% | SEBI's page has no equity sub-total this month. |
| Aug 2024 | +1,08,123.40 | +1,08,123.48 | -0.08 | -0.000% | SEBI's page has no equity sub-total this month. |
| May 2024 | +1,10,516.97 | n/a | SEBI has no page for the month before, so the month's net figure cannot be worked out. | ||
| Dec 2019 | -61,809.70 | -61,497.08 | -312.62 | -0.508% | A figure revised after AMFI's monthly report. The gap reverses within two months. |
| Oct 2019 | +1,33,481.91 | +1,33,481.91 | 0.00 | 0.000% | A figure revised after AMFI's monthly report. The gap reverses within two months. |
Rs crore. Gap = AMFI minus SEBI. Gap % is on SEBI's figure. SEBI's figures run from 1 April of each fiscal year, so a month is SEBI's year to that month minus its year to the month before.
The rupee to dollar rate NSDL uses
NSDL converts each day's flows to US dollars at a rate it prints. That rate is compared with the market close for the same trading day (Yahoo Finance, USD/INR).
| Period | Days | Typical gap | 19 days in 20 within | Largest gap |
|---|---|---|---|---|
| 2005 to 2011 | 1,686 | 0.243% | 0.92% | 8.1% |
| 2012 to 2019 | 1,916 | 0.165% | 0.643% | 5.41% |
| 2020 to date | 1,657 | 0.101% | 0.469% | 2.26% |
A market close and an official daily rate are set at different times of day, so a small gap is normal.
2. Checked between two reports of the same publisher
NSDL publishes the same money in more than one report. 1,898 comparisons, 13 beyond rounding.
| What must agree | Comparisons | Exact | Within rounding | Beyond rounding |
|---|---|---|---|---|
| Daily flows added up for each month = NSDL's monthly report, every route (to Rs 5 crore) | 689 | 22 | 667 | 0 |
| Total FPI holdings agree between the country report and the custody report | 348 | 327 | 19 | 2 |
| Sector holdings at each month end add up to total FPI equity holdings in the country report | 177 | 37 | 129 | 11 |
| US dollars: daily flows added up for each month = NSDL's monthly US dollar report, every route (to US$5 million) | 684 | 22 | 662 | 0 |
Within rounding = off by no more than NSDL's rounding to whole crores (daily against monthly: Rs 5 crore, or US$ 5 million).
Daily against monthly, FPI equity: across all 262 months from January 2005 the largest gap is Rs 3.8 crore.
The 13 comparisons beyond rounding
| Check | Month | Figure | Compared | First figure | Second figure | Gap | Gap, % |
|---|---|---|---|---|---|---|---|
| Country report against custody report, total FPI holdings | Nov 2022 | FPI equity holdings | Country report against custody report | 49,86,480 | 48,34,447 | +1,52,033 | +3.14% |
| Country report against custody report, total FPI holdings | Nov 2022 | All FPI holdings | Country report against custody report | 53,98,303 | 52,45,010 | +1,53,293 | +2.92% |
| Sector holdings against FPI equity holdings | Jan 2012 | FPI equity holdings | Sectors added up against country report | 8,93,652 | 8,93,326 | +326 | +0.04% |
| Sector holdings against FPI equity holdings | Feb 2012 | FPI equity holdings | Sectors added up against country report | 9,52,697 | 9,54,249 | -1,552 | -0.16% |
| Sector holdings against FPI equity holdings | Mar 2012 | FPI equity holdings | Sectors added up against country report | 9,69,450 | 9,53,514 | +15,936 | +1.67% |
| Sector holdings against FPI equity holdings | Apr 2012 | FPI equity holdings | Sectors added up against country report | 8,95,402 | 9,43,196 | -47,794 | -5.07% |
| Sector holdings against FPI equity holdings | Jun 2012 | FPI equity holdings | Sectors added up against country report | 9,35,783 | 9,34,143 | +1,640 | +0.18% |
| Sector holdings against FPI equity holdings | Aug 2012 | FPI equity holdings | Sectors added up against country report | 9,53,392 | 9,53,541 | -149 | -0.02% |
| Sector holdings against FPI equity holdings | Sep 2012 | FPI equity holdings | Sectors added up against country report | 10,56,562 | 10,56,537 | +25 | 0.00% |
| Sector holdings against FPI equity holdings | Mar 2013 | FPI equity holdings | Sectors added up against country report | 11,49,883 | 11,49,864 | +19 | 0.00% |
| Sector holdings against FPI equity holdings | Apr 2013 | FPI equity holdings | Sectors added up against country report | 11,98,458 | 11,98,771 | -313 | -0.03% |
| Sector holdings against FPI equity holdings | May 2013 | FPI equity holdings | Sectors added up against country report | 12,40,186 | 12,40,127 | +59 | 0.00% |
| Sector holdings against FPI equity holdings | Jun 2013 | FPI equity holdings | Sectors added up against country report | 11,83,305 | 11,83,450 | -145 | -0.01% |
Rs crore, or US$ million for the dollar check. Gap is the first figure minus the second, and the gap % is taken on the second. Each of these is explained under known issues, and the figures are shown as published.
3. Totals that must add up
Parts against the total the publisher prints. 2,657 comparisons, 5 beyond rounding.
| What must agree | Comparisons | Exact | Within rounding | Beyond rounding |
|---|---|---|---|---|
| Each month: equity + debt + hybrid + mutual funds + AIFs = NSDL's published total | 299 | 244 | 55 | 0 |
| Each month: debt = general limit + VRR + FAR | 299 | 299 | 0 | 0 |
| Holdings by fund domicile: top 10 countries + Other = total, every month | 534 | 211 | 323 | 0 |
| Custody holdings of every client type add up to the custody total, every month | 356 | 112 | 244 | 0 |
| FPI holdings by SEBI category add up to their totals, every month | 464 | 168 | 292 | 4 |
| US dollars, each month: equity + debt + hybrid + mutual funds + AIFs = NSDL's published total | 299 | 254 | 44 | 1 |
| US dollars, each month: debt = general limit + VRR + FAR | 299 | 299 | 0 | 0 |
FDI (RBI): FDI net equals FDI into India minus FDI by India in all 107 months from Oct 2017 to Aug 2026, to within US$ 0.01 million.
The 5 comparisons beyond rounding
| Check | Month | Figure | Compared | First figure | Second figure | Gap | Gap, % |
|---|---|---|---|---|---|---|---|
| FPI categories against their printed total | Jun 2014 | FPI equity holdings | Parts added up against printed total | 17,28,568 | 17,15,429 | +13,139 | +0.77% |
| FPI categories against their printed total | Jul 2014 | FPI equity holdings | Parts added up against printed total | 17,71,282 | 17,55,734 | +15,548 | +0.89% |
| FPI categories against their printed total | Aug 2014 | FPI equity holdings | Parts added up against printed total | 18,27,134 | 18,13,273 | +13,861 | +0.76% |
| FPI categories against their printed total | Sep 2014 | FPI equity holdings | Parts added up against printed total | 18,52,372 | 18,36,008 | +16,364 | +0.89% |
| US dollar routes against the printed total | Apr 2014 | All routes, net flow | Parts added up against printed total | 74 | 418 | -344 | -82.30% |
Rs crore, or US$ million for the dollar check. Gap is the first figure minus the second, and the gap % is taken on the second. Each of these is explained under known issues, and the figures are shown as published.
4. Known issues in the source data
16 items. Figures are shown as the source published them, and each item says what the tool does about it.
Seen, no explanation found, shown as published (5)
| ID | Area | Item |
|---|---|---|
| DQ1 | Who holds: FPI categories | Jun to Sep 2014: categories add up to more than the total In Jun 2014 to Sep 2014, the months when FIIs were moved into the new FPI categories, the category rows add up to more than NSDL's printed total. Largest gap Rs 16,364 crore. What the tool does: Shown as published. |
| DQ2 | Holdings | Nov 2022: two NSDL reports disagree on total FPI holdings The country report and the custody report give total FPI holdings Rs 1.53 lakh crore apart in Nov 2022. Every other month they agree to within Rs 1 crore. What the tool does: Holdings charts use the country report. Shown as published. |
| DQ3 | Sectors | 2012 and 2013: older sector list does not add up to FPI holdings The sector list for 2012 and 2013 includes individual issuers alongside sectors. From 2014, sector holdings add up to total FPI equity holdings within a few crore every month. What the tool does: Shown as published in the older sector list. |
| DQ4 | Who holds: custody | Insurers' custody holdings swing in 2014 to 2016 Custody holdings of insurance companies read Rs 9.59 lakh crore (Mar 2014), Rs 3.72 lakh crore (Mar 2015), Rs 12.27 lakh crore (Mar 2016). We believe the swings follow which clients each custodian reports. What the tool does: Shown as published. Avoid reading single months of the custody split in this period. |
| DQ16 | Flows | Apr 2014: the US dollar total does not add up NSDL's April 2014 US dollar page prints a total of US$418 million, the same figure as its rupee total. Its own parts add up to US$74 million. What the tool does: Shown as published. The equity and debt lines are not affected. |
Seen, and how the tool handles it (9)
| ID | Area | Item |
|---|---|---|
| DQ6 | Sectors | NSDL changed its sector list on 30 Apr 2022 The list went from about 35 sectors (plus some issuers) to 24, with no published mapping between the two. What the tool does: The two lists are shown separately and never joined. The current list has 109 fortnights of history. |
| DQ7 | Who holds: FPI categories | SEBI changed the FPI categories in 2014 and 2019 Category names and groupings differ before 2014, from 2014 to 2019, and after 2019. What the tool does: Shown as published. Compare a category only within one set of rules. |
| DQ8 | Countries | Country means the fund's home, and only the top 10 are listed NSDL reports the country where the fund is domiciled. Mauritius, Singapore, Luxembourg and Ireland are fund hubs. Only the top 10 countries are listed each month. What the tool does: A country outside the top 10 in a month is counted in Other for that month and shows a gap in its line. |
| DQ9 | Flows | The current month is month to date NSDL updates the current month (Nov 2026) every day until it ends. Completed months do not change (checked against a July 2026 copy: no cell changed). What the tool does: The current month is left out of monthly charts and 12-month sums. Daily charts show every day reported so far. |
| DQ10 | Flows | NSDL dates each day by when it was reported Custodians report a day's trades on the next working day, and NSDL's daily page carries that later date. A few entries hold more than one trading day: a depository holiday while the exchanges traded, a Saturday session or the Muhurat session. NSDL's monthly report counts each day in the month it was reported. What the tool does: Daily and monthly charts show NSDL's own dates by default. The Date and Month controls switch to the trading day. An entry that holds more than one trading day is then shown on the first of them. |
| DQ11 | Flows | US dollar figures NSDL converts each day's flows to US dollars at that day's rate and publishes a monthly report in US dollars as well as rupees, from January 2002. Its monthly dollar figures match the daily dollar figures added up. What the tool does: Monthly dollar figures are NSDL's monthly US dollar report. The current month to date adds up the daily report, the same days as the rupee figure. |
| DQ12 | Flows | NSDL and RBI measure FPI flows differently RBI's FPI series is on a balance-of-payments basis in US dollars and comes out about two months later. NSDL's is custodian-reported data by trade, in rupees and US dollars. The two do not match. What the tool does: FPI figures in the tool are NSDL's. RBI's monthly figure is used only as a check, on the Validation tab. |
| DQ15 | Flows | FAR and mutual funds are reported separately only from Sep 2024, AIFs from Sep 2025 Before September 2024 NSDL's monthly report puts FAR money inside the general-limit debt line, and has no separate mutual fund lines. AIFs have their own line from September 2025. What the tool does: Shown as published. Total debt is comparable across the whole history; the split by route is not. |
| DQ14 | All | Rounding NSDL prints whole crores in its monthly and holdings reports, so a total can differ from the sum of its parts by a few crore. What the tool does: Differences within rounding count as a match in our checks. |
Seen and fixed (2)
| ID | Area | Item |
|---|---|---|
| DQ5 | Who holds: FPI categories | Category breakdown from Aug 2024 From August 2024 NSDL splits the category report into two tables. Our first parse read only the smaller one. Fixed on 5 Oct 2026: both tables are read and add up to the grand total. What the tool does: Expired FPIs that have not been re-categorised are shown as one line. |
| DQ17 | Sectors | 30 September 2012 and 31 October 2014 sector reports NSDL's 30 September 2012 page has a header that does not match its columns, and our first parse read the wrong columns: the fortnight's net flows were shown as holdings. NSDL's 31 October 2014 page opens with a blank row, and our first parse skipped the fortnight. Fixed on 15 Nov 2026: both are read from NSDL's own pages and checked against the country report. What the tool does: Shown as NSDL published them. |
5. What has a second source and what does not
Where a figure has one publisher only, it is checked within that publisher's own reports.
| Data | Publisher | Checked against |
|---|---|---|
| FPI equity on the stock exchanges, daily | NSDL | NSE provisional figure, every trading day (section 1) |
| FPI flows, monthly | NSDL | RBI's monthly figure (section 1). NSDL's daily report against its monthly report (section 2) |
| FPI flows in US dollars | NSDL | The market exchange rate (section 1). NSDL's dollar report against its rupee report (sections 2 and 3) |
| FPI holdings | NSDL | NSDL's country report against its custody report and its sector report (section 2) |
| FPI sectors and countries | NSDL | Sector holdings against total FPI equity holdings (section 2). Parts against totals (section 3) |
| Mutual fund flows and assets | AMFI | SEBI's monthly statistics, every month from April 2019 (section 1) |
| SIP contributions | AMFI | One publisher |
| Mutual fund trades | SEBI | One publisher |
| FDI | RBI | The three lines against each other (section 3) |
See it live
This page is the static, readable companion to TREMOR's interactive India Capital Flows tool. Open the interactive tool on TREMOR for the Validation tab with its row-by-row tables, or read the India Capital Flows methodology page. TREMOR is part of tigzig.com - AI for analytics, databases and macro signals.