TREMOR's US insurance-investment data comes from the Federal Reserve Z.1 Financial Accounts: table S.128.1.s for Life and S.128.2.s for P&C. The Fed published these as L.116 and L.115 until the March 2026 release, when it renamed and renumbered them, so every line number on this page is the current one and will differ from any older L.116 reference. This page documents how that data is validated: tree-closure checks, a line-by-line cross-check against the Fed's own published table, and how the Z.1 figures relate to the insurance industry's own NAIC statutory numbers. For data sources and concepts, see the companion insurance methodology page.
Open the interactive US Insurance Assets tool on TREMOR for the live validation tables.
Dollar figures on this page are a snapshot as at Q2 2026, the latest quarter in the data when this page was written. The interactive tool computes the same checks live against whatever the latest quarter is.
Tree closure
For every parent in the asset tree, the sum of its children should equal the parent total. This check runs across all parents at the latest quarter and flags any gap: PASS = gap under 1% of parent (exact closure); PARTIAL = 1% to 15% (a known Fed publication gap, explained below); FAIL = over 15% (would suggest a misclassified series, investigated and fixed). The Total Financial Assets parent closes exactly at 0.00% for both Life and P&C at the latest quarter. The live per-parent closure table is in the interactive tool.
Known gaps explained
A few parents do not close to zero. In each case it is a known Fed publication quirk rather than a data error, but the quirks are not all the same kind, and the difference matters:
- Corporate & Foreign Bonds (Life and P&C): the Z.1 publishes the "ex-MBS/ABS" portion as a separate line but does not publish the MBS/ABS portion on its own at all. The gap equals the implicit MBS/ABS holdings within insurer corporate-and-foreign bonds. Nobody publishes that slice, so nobody could carry it.
- Life Misc Assets: the gap is one Fed line we cannot pull - Deferred and Unpaid Life Insurance Premiums, $43.8B at Q2 2026, line 26 of the current table. The Fed prints it in the Z.1 every quarter. What it does not do is distribute it as a FRED series, and FRED is where this tool gets its data. We leave the gap visible rather than filling it with a residual, because a residual would also swallow the next line the Fed adds and nothing would show.
- P&C Loans: the Z.1 parent is "Commercial Mortgages including reinsurer holdings" while the child tracked is "Commercial Mortgages (direct)"; the small gap (~$1B) equals mortgages held by reinsurers. Both are real, separately published series.
How the check works
The Z.1 publishes a hierarchy: Total Financial Assets, then asset-class lines, then sub-lines. Every parent and child that this tool carries in its tree is a published FRED series; children are summed and compared to the parent total at the latest quarter. A near-zero gap confirms the tree is consistent with the Fed's own breakdown. A non-zero gap is always something outside the tree, and it is one of the two cases above: either the Fed does not publish that slice separately, or the Fed publishes it in the Z.1 but does not distribute it on FRED.
Cross-check vs the Fed's published table
A line-by-line comparison of our FRED-sourced data against the Federal Reserve's own published Life Insurance Companies table - S.128.1.s, which the Fed published as L.116 until March 2026 - for each published line at the chosen quarter. All tracked items match.
Nothing in this comparison is transcribed by hand. The "Fed Published" column is read straight out of the Fed's release CSV bundle (Life is csv/S1281_s.csv), with the line numbers taken from the Fed's own data dictionary, and the file is recorded by SHA-256 so the exact bundle behind a given run is identifiable. The old per-table HTML pages were retired after the March 2026 release and now return 404, so the CSV bundle is the only machine-readable source. The comparison currently covers the 28 published lines of that table, against the release of 11 September 2026, with six quarters selectable (Q1 2025 through Q2 2026).
The columns are: Line, Fed Line Description, Fed Published ($B), Our FRED Data ($B), Diff ($B), Diff %, Status, and Our Series Name. Because the Fed renumbered this table when it became S.128.1.s, these line numbers are the current ones and will not line up with older L.116 references - Debt Securities, for instance, is now line 3 where it used to be line 5.
Two of the 28 lines are marked "not tracked", meaning they are published Fed lines not carried as separate FRED series here: line 9 (total loans), which is an aggregate whose components are already in the tree, and line 26 (deferred and unpaid life insurance premiums), which is the Life Misc Assets gap explained above. The live per-line comparison table, selectable by quarter, is in the interactive tool.
Fed Z.1 vs NAIC statutory - understanding the numbers
This tool's data comes from the Federal Reserve Z.1. The insurance industry's own regulator (NAIC) publishes different numbers using different accounting rules. Both are correct - they measure different things.
Year-end 2024 comparison (Life insurance)
| Metric | Fed Z.1 (this tool) | NAIC Statutory | Difference | Why |
|---|---|---|---|---|
| Total | $10.3T (Financial Assets) | $9.3T (Net Admitted Assets) | +$1.0T | Market-value uplift + broader definition |
| General Account | $7.1T | $6.0T | +$1.1T | MV vs amortized cost; non-admitted assets |
| Separate Account | $3.2T | $3.3T | -$0.1T | Close match (both near market value) |
This comparison is held at year-end 2024 on both sides, because that is the latest year the NAIC statutory source reports. The Z.1 column is therefore year-end 2024 and not the Q2 2026 figures used elsewhere on this page. NAIC statutory numbers sourced from the ACLI 2025 Life Insurers Fact Book (80th edition, Nov 2025, reporting year-end 2024).
Framework differences
| Dimension | Fed Z.1 (this tool) | NAIC Statutory |
|---|---|---|
| Accounting basis | Fed's own framework (market value for equities / some bonds) | Statutory Accounting Principles (SAP) - amortized cost for bonds |
| Asset inclusion | All financial assets (broader) | Net admitted assets only (excludes non-admitted: software, furniture, etc.) |
| Coverage | Fed's estimate for the entire sector | ~850 life insurers filing with state regulators (~99% of industry) |
| Frequency | Quarterly (~10-week lag) | Annual + mid-year commentary |
| Source | FRED (free, API, quarterly) | NAIC reports (free PDF; company-level data paywalled) |
| Best analogy | Like GAAP market-value financials | Like regulatory/tax-basis financials |
Key takeaway
The ~10% gap between Z.1 ($10.3T) and NAIC ($9.3T) is structural and expected, not a data-quality issue. The Separate Account numbers match closely ($3.2T vs $3.3T) because both frameworks value SA assets near market value; the gap concentrates in the General Account ($7.1T vs $6.0T) where bond valuation (market vs amortized cost) and non-admitted asset exclusions compound.
NAIC sources (free, public)
- NAIC Industry Snapshots - quarterly summaries
- NAIC Capital Markets Bureau - asset-mix research reports
- ACLI 2025 Life Insurers Fact Book - comprehensive industry data
- III.org Life Insurance Facts - clean HTML tables
See it live
This page is the static, readable companion to the validation views in TREMOR's US Insurance Assets tool (live tree-closure and S.128.1.s cross-check tables). Open the interactive tool on TREMOR, or read the insurance methodology page. TREMOR is part of tigzig.com - AI for analytics, databases and macro signals.