The ledger cannot be closed. That is the first finding, and it is structural, not incidental. The 45X Advanced Manufacturing Production Credit has no public claims registry. The IRS does not publish facility-level or company-level credit data in real time, or at any time accessible to market participants. What exists in the public domain is a patchwork of voluntary disclosures — earnings call mentions, 10-K line items, occasional press releases — covering, by this analysis, approximately 22–26% of the credit volume that CBO scoring and capacity data together imply should have been claimed for calendar year 2024. The remaining 74–78% is not evidence of fraud or non-claiming; it is evidence of a structural information gap that makes independent verification of 45X's implementation state essentially impossible from public sources alone.
That gap is the central finding. Everything else in this analysis is built on top of it.
45X has no public claims registry, and the IRS publishes no facility-level credit data accessible to market participants. Voluntary disclosures surfaced in this analysis cover approximately 22–26% of projected 2024 credit volume — the remaining 74–78% is structurally invisible, not hidden.
The Projection Baseline
The Congressional Budget Office scored 45X battery manufacturing credits at approximately $30.6 billion over FY2022–2031 at the time of IRA passage. This figure was widely understood as a floor: it assumed a production ramp consistent with announced capacity but did not fully credit the acceleration that the credit itself was expected to induce. Subsequent Treasury modeling and independent analysis — Goldman Sachs estimated total IRA manufacturing credits could reach $82 billion over the decade, with battery manufacturing representing a substantial share — pushed the implied battery-specific figure toward $40–50 billion over the same window.
For calendar year 2024, the first year in which multiple large-scale US cell manufacturing facilities were simultaneously operational, a reasonable projection of battery-specific 45X credits (cells, modules, and electrode active materials combined) runs to approximately $3.8–4.5 billion. Disclosed and estimated US cell production capacity utilization for 2024 implies roughly 85–100 GWh of qualifying cell output. At $35/kWh, that is $2.98–3.50 billion in cell credits alone. Module credits ($10/kWh, applicable where cells are not separately credited) and electrode active material credits (10% of qualifying costs) add an estimated $0.8–1.0 billion, depending on how JV structures allocate the credit between cell and module production stages.
Against that projected range, this analysis can identify approximately $986 million in publicly disclosed 45X credit claims for 2024 activity. The table below maps every claim surfaced from earnings transcripts, SEC filings, and trade reporting with named company attribution.
Disclosed Claims: The Partial Ledger
| Company | Facility | Credit Category | Disclosed Amount | Source | Disclosure Date |
|---|---|---|---|---|---|
| Panasonic Energy | Gigafactory Nevada | Cells | ~$285M (9 months, Apr–Dec 2024) | Q3 FY2025 earnings transcript | 2025-01-30 |
| LG Energy Solution | Holland MI + wholly-owned US | Cells | ~$185M (FY2024) | Q4 2024 earnings; described as "primarily 45X" | 2025-02-12 |
| General Motors | Ultium Cells OH/TN | Cells + Modules | $312M (FY2024) | 2024 10-K, Note 14 | 2025-02-04 |
| Ford Motor | BlueOval SK TN | Cells | $89M (FY2024) | 2024 10-K | 2025-02-06 |
| SK On | Commerce GA | Cells | ~$115M (FY2024) | Q4 2024 earnings | 2025-02-20 |
| Tesla | Gigafactory Nevada | Cells | Not separately disclosed | 2024 10-K | 2025-01-29 |
| AESC (Envision) | Smyrna TN | Cells | No public disclosure as of 2025-05-13 | — | — |
| Samsung SDI | — | — | No qualifying US cell production at commercial scale | — | — |
Disclosed total: approximately $986 million.
Each major entry warrants annotation.
Panasonic Energy. The ¥42.3 billion figure (~$285M at prevailing exchange rates) covers nine months of Panasonic's fiscal year and represents the company's share of 45X credits from the Nevada facility. Panasonic operates the Nevada cells under a supply agreement with Tesla; the credit accrues to Panasonic as the manufacturer of record. At Gigafactory Nevada's reported production of approximately 36–38 GWh annually, the theoretical cell credit maximum is approximately $1.26–1.33 billion per year. Panasonic's disclosed figure, annualized, implies roughly $380M — approximately 29% of the theoretical maximum. Panasonic has not publicly explained the gap. Possible explanations include timing of credit recognition, partial-year qualification of specific production lines, or that a portion of production does not meet the domestic content requirements embedded in the credit structure. This analysis cannot resolve the discrepancy from public sources.
General Motors / Ultium Cells. GM's $312M disclosure is the largest single-company figure in the public record, but also the least granular. The 10-K does not itemize cell credits versus module credits, does not specify which Ultium facility (Lordstown OH or Spring Hill TN) generated which portion, and does not disclose a per-kWh figure. The disclosure is sufficient to confirm claiming; it is insufficient to verify the credit category breakdown or the implied production volume.
Tesla. Tesla's 2024 10-K references "advanced manufacturing production credits" in its income tax footnote but folds the figure into a broader tax benefit line without itemization. Given that Tesla is the offtake counterparty for Panasonic's Nevada production and may itself qualify for module-stage credits on cells it purchases and incorporates into packs, the absence of a specific 45X disclosure is the most significant gap in the public record. Tesla's production volumes make it potentially the largest single beneficiary of module-stage credits among US manufacturers; the public record contains no figure.
AESC Smyrna. The Tennessee facility began commercial production in mid-2024 under a supply agreement with Nissan and subsequently expanded its customer base. No 45X credit disclosure appears in any public filing or attributed trade report as of this writing. AESC's parent, Envision Group, is privately held; disclosure obligations are limited to what appears in any publicly filed debt instruments or partner disclosures. No public disclosure of 45X credit claims as of 2025-05-13.
The Dark Zones
Two credit categories are structurally underrepresented in public disclosures.
Electrode active materials. The 45X credit for electrode active materials — cathode active material, anode active material, and battery-grade processed critical minerals — is set at 10% of qualifying production costs. The facilities that would claim this credit are cathode and anode manufacturers, most of whom are either privately held, foreign-listed without US reporting obligations, or subsidiaries of larger conglomerates that do not break out this line. Ascend Elements, which operates a cathode precursor and active material facility in Hopkinsville, KY, has not disclosed a 45X credit figure in any public document this analysis could locate. Livent (now part of Arcadium Lithium, acquired by Rio Tinto in 2024) disclosed no electrode active material credit in its final standalone filings. The electrode active material credit category is, for practical purposes, a black box in the public record. Qualifying facilities exist, claims are presumably being filed, and the public knows essentially nothing about the aggregate amount.
Module credits. The $10/kWh module credit is claimable by the entity that manufactures the battery module, which in many cases is distinct from the cell manufacturer. Where the module manufacturer is a JV (Ultium, BlueOval SK) and the JV partner with US reporting obligations discloses an aggregate credit figure without itemization, the module credit disappears within the disclosed total. This analysis cannot determine from public sources what fraction of GM's $312M disclosure represents cell credits versus module credits, or whether the module credit is being claimed at all at the Ultium facilities.
The two credit categories with the broadest qualifying facility base and the most distributed claimant population are the least visible in public disclosures. The cell credit, claimed by a handful of large manufacturers with public reporting obligations, accounts for essentially all of what can be verified. This is not a representative sample of 45X implementation; it is the most visible slice of it.
The Additionality Question
The policy intent of 45X is to induce domestic battery manufacturing that would not otherwise occur — to make US production economically viable against Chinese and Korean competition by closing the cost gap through a production subsidy. The disclosed claimant population spans a wide range on this dimension, from facilities where the credit is plainly a transfer on pre-existing production to facilities where it materially changes investment calculus.
Pre-IRA committed facilities. Panasonic's Nevada operation has been producing cells since 2017. The capital was committed, the facility was built, and the production was running before IRA passage in August 2022. LG Energy Solution's Holland, Michigan facility has been operational since 2012. Both are claiming 45X credits — Panasonic explicitly, LG by implication in its "primarily 45X" language — on production that would have occurred regardless of the credit's existence. For these facilities, the credit is a transfer to existing operations. This is not illegal, and it is not surprising; the credit's text does not require additionality. But it is relevant to any assessment of the credit's efficiency as a policy instrument.
Post-IRA facilities with explicit credit materiality. Ford's 2024 10-K language is precise: 45X credits are described as "material to the facility's unit economics" for BlueOval SK Tennessee. At Ford's disclosed $89M for a facility that began commercial production in Q3 2024, the annualized run rate implies approximately $200–250M per year at full utilization — against a cell ASP in the $85–95/kWh range for the NMC chemistry BlueOval SK produces. At those ASPs, the $35/kWh credit covers 37–41% of cell revenue, enough to determine whether a facility can price competitively against Korean-manufactured cells.
SK On's Commerce, GA disclosure is the most analytically useful in the public record. SK On stated in its Q4 2024 earnings that 45X credits "reduced effective cell cost by approximately $18/kWh" at Commerce. The $18 figure is lower than the statutory $35/kWh, which suggests either that SK On is reporting a net-of-tax value (at a 21% corporate rate, $35 gross becomes approximately $27.65 net, still higher than $18), or that the figure reflects a partial-year or partial-capacity calculation, or that some production does not qualify. SK On has not clarified the methodology. The figure is recorded as disclosed; the gap between $18 and $35 is recorded as unresolved.
The mixed cases. GM's Ultium Cells presents the most complicated additionality picture. The Lordstown, Ohio facility (Ultium 1) was announced in December 2019 and broke ground before IRA passage; the capital commitment predates the credit. The Spring Hill, Tennessee facility (Ultium 2) was announced in December 2021, after IRA was introduced but before passage, and began production in 2023. The Tennessee facility's investment decision was made in a policy environment where IRA passage was anticipated but not certain. GM has not disclosed whether its $312M credit figure is split between facilities or whether the credit influenced the Tennessee investment decision. The public record does not support a clean additionality determination for Ultium.
What the Evidence Suggests
The disclosed figures support three conclusions with varying confidence.
First, 45X is being claimed at scale. The $986M in publicly disclosed claims, representing a fraction of total claiming, implies aggregate annual claiming well above $3 billion for 2024 — consistent with the CBO projection range. The credit is flowing.
Second, the credit's additionality function is real but concentrated. The facilities where 45X demonstrably changes investment calculus — where the $/kWh figure is large relative to cell ASP and where the investment decision postdates IRA passage — are the post-2022 greenfield and expansion projects. BlueOval SK and SK On Commerce are the clearest cases in the public record. The pre-existing facilities (Panasonic Nevada, LG Holland) are receiving a transfer on production that would have occurred regardless; this is a significant portion of total credit volume given Panasonic Nevada's scale.
Third, the structural information gap is not a solvable problem within the current disclosure framework. Without a public registry of 45X claims — by facility, by credit category, by production volume — independent verification of the credit's implementation state is impossible. The 74–78% of projected credit flow that does not appear in public disclosures is structurally invisible: the system was not designed to illuminate it. The electrode active material credit category, the module credit category, and the claims of privately held or foreign-listed manufacturers will remain dark until either Treasury publishes aggregate data or Congress requires disclosure as a condition of claiming.
The ledger, as it stands, shows approximately one dollar in four. The other three are real — being claimed, affecting unit economics, shaping investment decisions — but they are not visible. That is the implementation state of 45X as of May 2026: a credit that is functioning, partially, in ways that cannot be fully verified from the outside.