What does automated NAV reconciliation look like? NavRecs, screen by screen
Automated NAV reconciliation turns a morning of logins, downloads and pasted tabs into four screens. The files come in, every check runs, and your reviewer opens the breaks ranked by their impact on the net asset value (NAV), each one traced to the cell it came from. In NavRecs, every check re-performs in seconds: on our demo book, 371 checks across 6 funds run in under 3 seconds, measured. Your reviewer signs, and the whole morning lands on a tamper-evident record.
In short
- The files come to NavRecs: each source is set up at onboarding to be collected on a schedule, and anything not yet connected is dropped in.
- The overview puts the funds with open exceptions first; funds that agree wait.
- Each fund's checks are calibrated with the firm's own tolerances and switched on or off per fund.
- Reviewers see only the breaks, ranked by their impact on the NAV.
- Every figure is one click from the cell it came from.
- Every sign-off lands in an append-only, tamper-evident (hash-chained) audit trail.
How do the morning's files get in?
Collection is the first job to go. At onboarding we set NavRecs up, per source, to collect the files itself on a schedule: secure file transfer, a direct connection or a mailbox pickup into your own tenant, whatever each custodian, broker or pricing feed offers. For every connected source, your team stops logging in to portals and downloading statements before the review can even start.
A source that is not connected yet is simply dropped in, and it is read the same way. NavRecs reads the exported files your systems already produce, so there is nothing to migrate and nothing changes in your accounting system.
What does the overview show first thing?
The first screen is the overview board for the morning. On the synthetic demo book it shows a mixed morning: one NAV already released, one ready to release, and four funds with open exceptions. Funds with work sit at the top. Funds that agree wait.
That ordering is the point. A spreadsheet review works through the funds in tab order. The overview is built so the reviewer starts where judgement is needed, not where the file list happens to sit. It shows the state of the review across the book after the checks have run.
If you already run a NAV review in Excel, this is where the spreadsheet morning would be after the pasting and the comparing, without either. It is the same job, NAV reconciliation: the prepared NAV compared to its sources before release, done in seconds.
How are the checks set up for each fund?
Each fund has its own page. It lists every check for that fund with this morning's outcome, clean or open, and the fund's bands on one line under its name. Each check is calibrated per fund with the firm's own tolerances.
The product ships the checks you need: 18 shipped, switched on or off per fund, new checks built on request. The full list, in order, is on the checks page: cash, positions, prices, FX, trades, fees, accruals, capital activity and the NAV movement itself, among others. Every check is mapped to the GFSC pricing controls guidance (November 2021) and CSSF Circular 24/856.
That configuration is set once per fund and re-used every dealing morning. A check that does not apply to a fund is switched off.
What does the reviewer see when something breaks?
When the checks have run, the reviewer does not walk every agreeing row. Reviewers see only the breaks, ranked by their impact on the NAV. A break is a difference above the tolerance the firm set for that check on that fund. The opened break shows where each figure came from, report and cell, any escalation note, and the sign-off form.
Every figure is one click from the cell it came from. That is the contrast with initials in a spreadsheet cell: the workbook proves someone was there; this screen shows what was compared, to what, and why the outcome was accepted or escalated. Your reviewer makes the call with the evidence already in front of them, and nothing on this screen makes it for them.
What does the audit trail record after sign-off?
After investigations and sign-offs, the morning's history sits in the audit log. It is an append-only, tamper-evident (hash-chained) audit trail; nothing is deleted. For one morning it records file reads, comparisons, exceptions, sign-offs and releases, entry by entry. An auditor asking what was compared, at which tolerance, by whom and when is answered from this record, not from a reconstructed email pack.
The trail runs inside your own Microsoft tenant, with keys in your own Azure key vault and retention that follows your policy. There are no outbound internet calls, nothing leaves the tenant, and we never see your data. Sign-in is Microsoft Entra with your own two-factor, and users are unlimited. The detail is on the security page.
What does the morning look like with NavRecs?
The mechanical morning goes: the logins, the downloads, the rolled-forward workbooks, the re-keyed PDF lines and the scanning of rows that already agree. It runs in seconds, so your senior people start on the breaks instead of the build-up.
That can be up to 90% of the morning's review time. On one desk we studied closely, our own estimate was around ten senior hours a day across its funds. Roughly nine of those went on collection, logins, re-keying, roll-forwards and comparing numbers that already agreed.
About one hour was judgement. That is one desk, our own count, not a survey. What stays is the hour worth a senior person: your reviewer makes the release call with every figure evidenced, your accounting system keeps producing the NAV, and NavRecs never writes to it.
| Each morning | In NavRecs | In a spreadsheet |
|---|---|---|
| Getting the files | Collected per source on a schedule once set up at onboarding, or dropped in | Logged into, downloaded and pasted in, source by source |
| Comparing | Every check re-performed in seconds | Formulas across tabs rolled forward from yesterday |
| Finding the breaks | Only the breaks, ranked by their impact on the NAV | Scanning every row, including the ones that agree |
| Evidence | Every figure one click from the cell it came from | Initials in a cell |
| The record | An append-only, tamper-evident (hash-chained) audit trail | Workbook versions and emails |
| Sign-off and release | Your reviewer signs, with every figure evidenced | The reviewer signs, from a workbook |
How do you see these screens on sample data?
Book a 20-minute demo to see these screens on sample data, with your questions answered in your own terms. Bring whoever signs off the NAV. Beforehand we ask three questions, so the session matches your morning: which fund accounting system you run, how many funds you strike and how often, and which custodians and brokers your files come from.
After the demo, a pilot is a parallel run against your own morning, with pass criteria in writing and corrections at no cost. Setup takes weeks, not days, for a client on a common accounting system, and pricing is by fund count, not by seat. For the check list or the tenancy detail first, start at the checks or security. For the business case, read why automate the NAV review.
Frequently asked questions
Does NavRecs calculate the NAV?
No, and that is the point. Your accounting system calculates the NAV; NavRecs checks it independently before release, against the exported files, and records every step. Your reviewer signs the release.
Do the files arrive on their own?
Yes, once each source is set up at onboarding: NavRecs collects its files on a schedule, by secure file transfer, a direct connection or a mailbox pickup. A source that is not connected yet is dropped in. Either way, it reads the exported files your systems already produce.
Is there generative AI in the numbers?
No. The engine is deterministic. Same files, same answer, every time. No generative AI in the numbers.
Where does the data live?
Inside your own Microsoft tenant. Nothing leaves it, and we never see your data.
See the checks NavRecs runs every morning → The checks
How it stays inside your own tenant → Security