Deloitte LLP 2 new Street Square London EC4A 3BZ   Phone:+44 (0)20 7936 3000 Fax +44 (0)20 7583 1198 www.deloitte.co.uk

Independent Accountant’s Report on Applying Asset Agreed Upon Procedures

The Board of Directors of

Harvest Funding PLC

10th Floor, 5 Churchill Place

London

E14 5HU

(the “Issuer”)

The Board of Directors of

Bank of Scotland plc

The Mound

Edinburgh

EH1 1YZ

(the “Seller”)

Lloyds Bank Corporate Markets plc

33 Old Broad Street

London

EC2N 1HZ 

(“LBCM”)

Citigroup Global Markets Limited

Citigroup Centre

Canada Square, Canary Wharf

London

E14 5LB

(“CGML”, and together with LBCM, the “Joint Arrangers” and the “Joint Lead Managers”)

and the other Managers (as defined in the Engagement Letter)

10 February 2026

Dear Sirs/Madams,

PROPOSED ISSUE BY HARVEST FUNDING PLC OF RESIDENTIAL MORTGAGE-BACKED NOTES (the “Issue”)

We have performed the procedures enumerated below on certain residential mortgages (the “Loan Pool”), which were agreed to by the Issuer, the Seller, the Joint Lead Managers, the Joint Arrangers, the Managers and the other Managers (as defined in the Engagement Letter). The procedures were performed to assist you in evaluating the validity of certain characteristics of the Loan Pool. This agreed upon procedures engagement was conducted in accordance with attestation standards established by the American Institute of Certified Public Accountants. The appropriateness of these procedures is solely the responsibility of the Issuer, the Seller, the Joint Lead Managers, the Joint Arrangers and the Managers. Consequently, we make no representation regarding the appropriateness of the procedures described below either for the purpose for which this report has been requested or for any other purpose.

This report (the “Asset Agreed Upon Procedures Report”) is addressed to the Board of Directors of the Issuer, the Board of Directors of the Seller, the Joint Lead Managers, the Joint Arrangers and the Managers identified in the Engagement Letter who have agreed to participate in the proposed Issue.

Deloitte LLP is a limited liability partnership registered in England and Wales with registered number OC303675 and its registered office at 1 New Street Square, London, EC4A 3HQ, United Kingdom.

Deloitte LLP is the United Kingdom affiliate of Deloitte NSE LLP, a member firm of Deloitte Touche Tohmatsu Limited, a UK private company limited by guarantee (“DTTL”). DTTL and each of its member firms are legally separate and independent entities. DTTL and Deloitte NSE LLP do not provide services to clients. Please see www.deloitte.com/about to learn more about our global network of member firms.

© 2026 Deloitte LLP. All rights reserved.

 

These procedures were established with the Issuer and the Seller, and the appropriateness of the procedures is solely the responsibility of the Issuer, the Seller, the Joint Lead Managers, the Joint Arrangers and the Managers as discussed above. The Issuer and the Seller shall be solely responsible for providing accurate and complete information requested by Deloitte necessary to perform the procedures. Deloitte has no responsibility for the accuracy or completeness of the information provided by or on behalf of the Issuer and the Seller, even if Deloitte had reason to know or should have known of such inaccuracy or incompleteness.

The procedures that we performed and our findings are as follows:

1.              Scope of our work and factual findings

The procedures performed were as defined in Appendix 3 of the Engagement Letter (the “Agreed Upon Procedures”).

The Seller provided us with a data file ‘Project Kronos sub account list 311225.xlsx’ (the “First Pool Run”) containing an account number for each of the 43,144 loans and further advances in the Loan Pool as at 31 December 2025 (the “Cut-off Date”).

A random sample of 457 loans and further advances (the “Sample”) was selected from the First Pool Run using the sampling approach below.

The Seller then provided the data file ‘Deloitte Sample - Project Kronos 311225 - MB UPDATED.xlsx’ (the “Sample Pool”) containing information for each loan in the Sample as at the Cut-off Date.

We have carried out the Agreed Upon Procedures on the Sample Pool during the period 19 to 26 January 2026.

The sampling approach

Sampling confidence is the probability that the actual errors, within a total population, are contained within the range of an estimate. Precision is the range of that estimate. The precision limit is the estimated maximum predicted number of errors within the total population. Sampling confidence and precision are stated in percentages.

Attribute sampling is a method of assessing the rate of occurrences of a specified attribute in a population and requires agreed upon procedures in relation to certain characteristics of a random sample of individual loans. In this case the Agreed Upon Procedures related to the documentation and procedures that support the Sample Pool contained in the First Pool Run.

Our method of calculating attribute sample sizes is based on the binomial probability distribution.

Statistical tests can only provide estimates of the error.

The procedures we have undertaken, set out in paragraphs 2.1 to 2.22 below, have been limited to confirming that the selected attribute from the Sample Pool information relating to the First Pool Run agreed to the original loan documentation or copies thereof provided to us in the loan file. We are entitled to assume that the loan documentation is correct and we have not sought to independently verify this information. The loan documentation used in these Agreed Upon Procedures was the Offer, product change documentation, valuation report, Mortgage Archive Retrieval System (“MARS”), Borrower’s origination printout (“AHD”), Information Lifecycle Management System (“ILM”), Mortgage Deed held on the land registry website, Land Registry Charge data file ‘Lloyds Kronos Jan 2026 Ownership Spreadheet Results.xls’ and Land Registry prints (together, the “Land Registry Charge”), Certificate of Title (“CoT”) and the primary system of record (the “UFSS System”).

Objectives

On the assumption that the information in the First Pool Run is a complete and accurate representation of the Loan Pool, the sample sizes chosen were designed with the objective of us being able to state that there is a 99% confidence that not more than 1% of the population of the First Pool Run contained an error in the relevant attribute. Where errors were found in the Sample Pool, we have re-calculated the percentage errors that there might be in the First Pool Run and noted the revised percentage below. It is for you to ensure you understand this basis of reporting and to determine whether the errors are acceptable to you and for the purposes of the Issue.

 

2.              Statistical Agreed Upon Procedures

For each loan in the Sample Pool, we carried out the following Agreed Upon Procedures and have given a confidence and precision percentage for each procedure.

For the purposes of this report, the failure of a single attribute is termed an error. We report our findings, which are the factual results of the agreed-upon procedures performed:

In cases where the surname of one borrower changed to the surname of a joint borrower, we were informed by the Seller that such changes are as a result of changes in the marital status of the borrowers. For the avoidance of doubt, we did not check that any change in the surname of the borrower was due to a change in their marital status, as no evidence of changes to the marital status was available in the mortgage file.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

As a result of the procedure performed there is a 99% confidence that not more than 3% of the First Pool Run contained errors.

We were instructed by the Seller that this procedure should not include agreeing the postcode if other segments of the Property address were sufficient to substantially agree the property address. Additionally, for new build properties, we were instructed by the Seller to assume that differences between the property address in the Sample Pool and the loan documentation should not be considered to be errors if the property address was not substantially different as it can be expected that the property address can change as a mortgage progresses from application to completion. For the avoidance of doubt, if the postcode was missing, incomplete or incorrect, but the other segments of the property address substantially agreed then we did not treat this as an error. For the avoidance of doubt if the property address and the Sample Pool were not substantially different then we did not treat this as an error.

 

As a result of the procedure performed there is a 99% confidence that not more than 3% of the First Pool Run contained errors.

As a result of the procedure performed there is a 99% confidence that not more than 6% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed whether the property tenure shown in the Sample Pool agreed with the Land Registry Charge, ILM, MARS or AHD. We found that the property tenure agreed to the Land Registry Charge, ILM, MARS or AHD, except for 2 cases.

As a result of the procedure performed there is a 99% confidence that not more than 2% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed whether the property type shown in the Sample Pool agreed with ILM, MARS or AHD. We found that the property type agreed to ILM, MARS or AHD, except for 12 cases.

 

As a result of the procedure performed there is a 99% confidence that not more than 5% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed whether the loan purpose shown in the Sample Pool agreed with ILM, offer, MARS or AHD. We found that the loan purpose agreed to ILM, offer, MARS or AHD, except for 18 cases.

As a result of the procedure performed there is a 99% confidence that not more than 7% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed whether the purchase price shown in the Sample Pool (where remortgage = ND and purchase contains a purchase price) agreed with ILM, MARS or AHD. We found that the purchase price agreed to ILM, MARS or AHD, except for 2 cases.

As a result of the procedure performed there is a 99% confidence that not more than 2% of the First Pool Run contained errors.

 

For each mortgage loan in the Sample Pool, we confirmed whether the amount advanced shown in the Sample Pool agreed to within £50 of that shown on the Offer, ILM, MARS or AHD. We found that the amount advanced agreed to within £50 of that shown on the Offer, ILM, MARS or AHD, with no exception.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

For each mortgage loan in the Sample Pool, we confirmed whether the interest rate type shown in the Sample Pool agreed to that shown on the Offer, subsequent product switch documentation or the UFSS System. We found that the interest rate type agreed to that shown on the Offer, subsequent product switch documentation or the UFSS System, with no exception.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

For each mortgage loan in the Sample Pool, we confirmed whether the current interest rate shown in the Sample Pool agreed to that shown on the Offer, subsequent product switch documentation or the UFSS System. We found that the current interest rate agreed to that shown on the Offer, subsequent product switch documentation or the UFSS System, with no exception.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed whether the date the amount was advanced in the Sample Pool agreed with that shown on the UFSS System. We found that the date the amount was advanced in the Sample Pool agreed with that shown on the UFSS System, with no exception.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed whether the securitisation balance amount shown in the Sample Pool agreed to within £50 of the securitisation balance amount shown on the UFSS System as at the Cut-off Date. We found that the securitisation balance amount shown in the Sample Pool agreed to within £50 of the securitisation balance amount shown on the UFSS System as at the Cut-off Date, with no exception.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

For each loan in the Sample Pool, using the term start date and the term shown on the System we recalculation the maturity date (the “Recalculated Maturity Date”). We confirmed whether the Recalculated Maturity Date agreed to the maturity date shown in the Sample Pool, to within one month. We found that the Recalculated Maturity Date agreed to the maturity date shown in the Sample Pool, to within one month, with no exception.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

 

For each loan in the Sample Pool, we confirmed whether the valuation date shown in the Sample Pool agreed to the valuation report, ILM, MARS or AHD. We found that the valuation date shown in the Sample Pool agreed to the valuation report, ILM, MARS or AHD, with no exception.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed whether the valuation amount shown in the Sample Pool agreed to the valuation report, ILM, MARS or AHD. We found that the valuation amount shown in the Sample Pool agreed to the valuation report, ILM, MARS or AHD, except for 1 case.

As a result of the procedure performed there is a 99% confidence that not more than 2% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed whether the approved valuer shown in the Sample Pool agreed to the list of approved valuers on the UFSS System. We found that the approved valuer on the Sample Pool agreed to the list of approved valuers on the UFSS System, with no exception.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed whether Bank of Scotland plc or Halifax plc held first lien status on the Land Registry Charge. We found that Bank of Scotland plc or Halifax plc held first lien status on the Land Registry Charge, except for 8 cases.

As a result of the procedure performed there is a 99% confidence that not more than 4% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed the account number shown on the Sample Pool agreed with that shown on the UFSS System. We found that the account number agreed with that shown on the UFSS System, with no exception.

 

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed whether the repayment type shown in the Sample Pool agreed to that shown on the Offer or the UFSS System. We found that the repayment type agreed to that shown on the Offer or the UFSS System, with no exception.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

For each loan in the Sample Pool, we confirmed whether the payment due shown in the Sample Pool agreed to that shown on the UFSS System. We found that the payment due agreed to that shown on the UFSS System, with no exception.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

 

For each loan in the Sample Pool, we confirmed whether the interest reset date shown in the Sample Pool agreed to that shown on the UFSS System. We found that the interest reset date agreed to that shown on the UFSS System, with no exception.

As a result of the procedure performed there is a 99% confidence that not more than 1% of the First Pool Run contained errors.

3.              Scope of this Asset Agreed Upon Procedures Report

The scope of our work in preparing this Asset Agreed Upon Procedures Report was limited solely to those procedures set out above. We were not engaged to and did not conduct an audit the objective of which would be the expression of an opinion on the matters subject to the procedures described above. Accordingly, we do not express any opinion or overall conclusion on the procedures we have performed. You are responsible for determining whether the scope of our work specified is sufficient for your purposes and we make no representation regarding the appropriateness of these procedures for your purposes. If we were to perform additional procedures, other matters might come to our attention that would be reported to you.

This Asset Agreed Upon Procedures Report should not be taken to supplant any other enquiries and procedures that may be necessary to satisfy the requirements of the recipients of this Asset Agreed Upon Procedures Report.

This Asset Agreed Upon Procedures Report has been prepared for use in connection with the offering of securities inside of the United States and in accordance with United States attestation standards established by the American Institute of Certified Public Accountants and not in accordance with United Kingdom or other professional standards. This Asset Agreed Upon Procedures Report is only intended to be relied on in connection with any obligations or responsibilities in connection with the Issue that you may have under any legislation, regulations and/or rule of law under United States federal or state securities law. We accept no responsibility to, and deny any liability to, any person or in any way arising from or in connection with the use of this Asset Agreed Upon Procedures Report in connection with any offering of securities outside the United States.

4.              Use of this Asset Agreed Upon Procedures Report

Subject to the terms specified in in our Engagement Letter dated 30 January 2026, this Asset Agreed Upon Procedures Report is provided solely for the private information and use of the Issuer, the Seller, the Joint Lead Managers, the Joint Arrangers and the Managers and is not intended to be and should not be used by anyone other than the Issuer, the Seller, the Joint Lead Managers, the Joint Arrangers and the Managers. This Agreed Upon Procedures Report should not be included in any publicly filed or publicly available document nor should this Agreed Upon Procedures Report, the services performed hereunder, or Deloitte’s engagement hereunder be referred to in any publicly filed or publicly available document, in each case, except under the terms specified in our Engagement Letter. For the avoidance of doubt, the Joint Lead Managers, the Joint Arrangers and the Managers are being provided with the Asset Agreed Upon Procedures Report in their capacity as addressees and not the engaging party.

If you have any questions or would like to discuss the content of this Asset Agreed Upon Procedures Report please do not hesitate to contact Louise Gill on +44 113 292 1299.

Yours truly,

/s/ Deloitte LLP

Deloitte LLP

9