Gold Verification Audit Report July 2025

Executive summary

The existence and accuracy of the Reserve Bank of Australia’s (RBA) gold holdings were verified by RBA Audit in June 2024 at the RBA’s Sydney Holding Point and in June 2025 at the Bank of England (BoE). All gold bars selected for verification were satisfactorily verified. No discrepancies or inconsistencies were identified during the ultrasound and weighing of selected bars. Also, based on our discussions and observations, the gold custody arrangements and segregation of duties at the BoE were assessed as appropriate.

This physical verification of the RBA’s gold bar holdings at the BoE follows previous reviews in 2022, 2019 and 2013. No issues have been identified in this review or previous reviews.

Background

The RBA holds 2.664 million fine troy ounces (FTO) of gold, valued at A$9.940 billion, as at 31 January 2025. Almost all of the RBA’s physical holdings are stored at the BoE; four bars (0.0016 million FTO) are stored at the RBA’s Sydney Holding Point. Some of the RBA’s holdings are used for gold swaps with carefully selected counterparties to enhance returns. See Table 1 below for details.

All bars in the RBA’s account at the BoE must meet the standards for ‘Good Delivery’ set by the London Bullion Market Association (LBMA).¹ These standards include the list of acceptable refiners of gold bars in the London bullion market, technical specifications of bars, quality assurance provisions, and compliance and risk management rules. Quality assurance provisions require refiners to comply with the Proactive Monitoring Programme (which can include dip sampling from a commercial melt), while compliance and risk management rules require refiners to have an effective governance framework and to maintain appropriate systems and controls.

Table 1: RBA’s Gold Holdings at 31 January 2025

Objective

The objective of the review was to verify the RBA’s gold holdings at the BoE and Sydney Holding Point and to obtain a high-level overview of the gold custody arrangements at the BoE.

Observations and findings

Holdings at the BoE

The RBA had a total of 6,372 bars, as at 2 February 2025, held at the BoE. The physical verification of gold bars was performed by examining a sample of 460 bars, being 7.22 per cent of the RBA’s total holdings. The sample was chosen by the auditor on a random basis, which gave bars an equal chance of being selected. The sample excluded bars that have been previously sampled and verified. The sample comprised a pre-identified selection of bars and a sample selected on-site (see below).

As part of the review, Audit also observed and was provided with a high-level overview of the BoE’s gold custody arrangements, segregation of duties and physical security arrangements, with no issues noted.

Pre-identified sample

A random sample of 450 bars (around 7 per cent of the RBA’s holdings) was selected by Audit and notified to the BoE prior to Audit’s arrival on-site; as such, the location of these bars and their retrieval from the vaults was not witnessed. Audit physically verified each of the sampled gold bars, confirming the refiner’s identification, the refiner bar number, the gold assay rating (the fineness of the bar) and the BoE bar number (checking that these all matched the RBA’s holdings).

A randomly selected sample of 60 of the 450 bars were subject to ultrasound checks and weighing. Ultrasound checks and weighing were performed by BoE staff using their equipment with Audit present.² These checks confirmed that the weight of the gold bars was consistent with the details on the listing (within an acceptable tolerance of ±0.025 fine ounces, which is in line with the LBMA Good Delivery List Rules).

On-site sample selection

A further sample of 10 bars was randomly selected upon Audit’s arrival at the BoE; this sample was not provided to the BoE prior to the visit. Audit was escorted to physically verify each of these bars within the BoE’s gold vaults. Audit verified each of the selected gold bars, confirming the refiner’s identification, the refiner bar number, the gold assay rating and the BoE bar number. These 10 bars were not subject to ultrasound checks and weighing and were sighted in the vaults only.

No discrepancies were identified between the RBA’s gold listing and the 10 gold bars that Audit sampled in this way.

Sydney Holding Point

The RBA holds four gold bars at the Sydney Holding Point. Audit physically verified the gold bars noting their details, including the refiner’s identification, refiner bar number, assay rating and weight. No issues were noted. The weight of these gold bars is checked every five years and is not due to be checked again until 2029.

Conclusion

Based on the procedures performed at the BoE:

• All gold bars selected (460 bars) were satisfactorily verified by Audit confirming the refiner’s identification, the refiner bar number, the assay rating and the BoE bar number. This included the 450 gold bars pre-selected and the 10 bars that were selected only upon Audit’s arrival at the BoE.
• No discrepancies or inconsistencies were identified during the ultrasound checks and weighing of the 60 bars pre-identified. In line with the LBMA Good Delivery List Rules, the gross weight of each bar is expressed in multiples of 0.025, rounded down to the nearest 0.025 of a troy ounce.
• Based on Audit’s observations and discussions, the gold custody arrangements, segregation of duties and physical security arrangements at the BoE are assessed as appropriate.

Audit also sighted the four gold bars stored at the Sydney Holding Point. No issues were noted.

Appendix A: Sample summary by refiner

(a) All bars produced by Russian refiners were refined prior to 7 March 2022.

Endnotes

See LBMA (2025), ‘Good Delivery Rules’, January. Available at <https://www.lbma.org.uk/publications/good-delivery-rules>. 1

Audit obtained the certificates of calibration from an independent provider confirming the condition of the ultrasonic flaw detector used for the ultrasound checks and the condition of the gold bullion scale(s) used for weighing. 2