## THE ISSUES

## Source details

**Canonical URL:** [THE ISSUES](https://www.imf.org/-/media/files/data/statistics/bpm6/fitt/f15-debt-concessionality.pdf)

## Other formats

- [Markdown version](/-/media/files/data/statistics/bpm6/fitt/f15-debt-concessionality.pdf.md)
- [Structured JSON version](/-/media/files/data/statistics/bpm6/fitt/f15-debt-concessionality.pdf.json)

---

### Background and scope
- BPM6 and 2008 SNA identify debt concessionality as a research topic (BPM6, paragraph 1.43 (h); 2008 SNA, paragraph A4.44).
- Concessional loans: lending intentionally provided at a contractual interest rate below market interest rates, for similar grace and repayment periods, with the purpose to convey a benefit, occurring in a non-commercial context (BPM6, paragraphs 3.79 and 12.51; 2008 SNA, paragraphs 3.131, 3.134, 22.123, and 22.124; GFSM 2014, paragraphs 3.123 and A3.39-41).
- Concessional loans may include other beneficial characteristics (e.g., more favorable grace and maturity periods) (2008 SNA paragraph 22.123; BPM6 A2.68; EDS Guide paragraph 6.22).
- Scope of Guidance Note (GN): low interest loans provided in a non-commercial context, including entities acting on behalf of others, with an intention to convey a benefit; excludes lenders with very low average funding costs as de facto non-concessional in the GN’s meaning.

### Link to valuation guidance
- GN relates to GN F.9 “Valuation of Loans (Fair Value)”: GN F.9 focuses on re-estimates in loan value; the present GN deals with valuation at inception.

### Core issues for discussion
- Manuals acknowledge a transfer element in concessional loans but generally do not prescribe recording expense (creditor) or revenue (debtor) in core accounts; typically reported as memorandum items.
- Current practice: recording concessional loans at inception at face/nominal value rather than discounted/fair value.
- Mismatch: recognition of a transfer/gift in principle versus absence of uniform core-account recording recommendation.
- GN considers three recording options and discusses appropriate discount rates to measure the transfer component.

---

### OPTIONS FOR STATISTICAL TREATMENT OF CONCESSIONAL LENDING (ISSUE 1)

### Options overview
- Option A: No change — record concessional loans at face/nominal value at inception with no imputation; transfer element remains memorandum/supplementary information.
  - Sub-option A1: No change in core accounts but provide more detailed memorandum/supplementary information on the transfer element, notably the fair value of the loan and the entailed accrued interest.
- Option B: Record loan at face/nominal value at inception, but recognize the transfer element in core accounts by increasing the stream of interest earned (D.41) on the loan using a suitable non-concessional rate and recording a matching transfer expense so that the net effect is deficit-neutral in every accounting period.
- Option C: Partition concessional loans at inception between a “genuine” loan element (F.4) and an explicit “transfer element”; impute interest receivable (D.41) on the loan that capitalizes on the new nominal value over time so that the loan’s nominal principal reaches face value before maturity; deficit-neutral across life of loan.

### Numerical example (summarized; full details in Annex II)
- Creditor grants on January 1 a bullet-loan (5 years, redeemed on December 31) of 100 Units with contractual zero interest.
- Creditor’s average funding interest rate used as discount rate: 5 percent.
- Net present value at inception of the payment stream {0, 0, 0, 0, 100} at a 5 percent discount rate is 78.4 U.
- Transfer recorded at inception equals 21.6 (=100-78.4).
- Under all Options, end Year 5, just before redemption, the instrument is valued at face value 100.

### Additional considerations on implementation and discount rate
- GN recommends extending same rules to debt restructurings where changes render loans concessional (see Annex VII).
- Determination of the appropriate nonfinancial transaction to record the transfer element requires case-by-case analysis (Annex VI).
- Candidate discount rates: creditor’s observed funding cost, OECD Commercial Interest Reference Rates (CIRR), or borrower’s funding cost.
- GN conservative recommendation: use average cost of the grantor/creditor as the discount rate; recognizes CIRR as aligned with current manual orientations and easily observable.

---

### GRANT ELEMENT WHEN CONCESSIONAL LOANS SUBSTITUTE CONTRIBUTIONS (ISSUE 2)

### Problem framing
- Governments sometimes provide official contributions to agencies as long-term zero-interest loans, with beneficiaries calculating an implied “grant element.”
- Many compilers treat such concessional loans as substitutes for regular contributions and recognize the grant/transfer element in core accounts at disbursement to uphold substance over form.
- Consistent recording is required to avoid incentives to substitute transfers with low-interest loans.
- Where contributions are recorded as equity (e.g., paid-in capital), concessional loans to such beneficiaries should be partitioned at inception between a loan and an equity element.

### Options for Issue 2 (if Option C for Issue 1 is rejected)
- Option A: No specific rule for cases where concessional loans substitute for contributions/transfers.
- Option B: Require explicit recording in core accounts at inception of the grant/transfer element when concessional loans substitute for regular contributions to ensure substance over form.

### GN recommendation and policy stance
- GN recommendation: transfer element of concessional loans granted as substitutes for regular or other transfers/grants should be recorded in core accounts consistent with regular contributions or other transfers/grants.
- Recording at face value low/zero-interest rate loans extended by donors in place of outright grants should not be permitted because it defers or changes deficit impact and is an "outright fiscal illusion maneuver."
- Recommended option for Issue 2: Option B (majority support in FITT consultation), with some members advocating consistency with Issue 1 treatment or separate guidance for beneficiary institutions.

---

### COMPARATIVE ASSESSMENT OF OPTIONS A, B, C

### Core analytical point
- Option C recognizes at inception a partition between a loan element and a grant/transfer element; Options A and B do not.
- Main objection to Option C: it can delay the deficit impact of the grantor relative to Options A and B.

### Key arguments summarized
- For Option A:
  - Simplicity.
  - Creditor’s “cost” proxied by difference between low interest earned and higher funding costs; net assets deteriorate over time.
- For Option B:
  - Corrects misreporting of expenditure by recognizing a transfer from creditor to debtor in core accounts spread over life of instrument.
  - Builds on 2008 SNA paragraph 7.126 and extends central-bank special-case approach to government below-market lending.
  - Limitation: overestimates subsidy component because it does not use present value; cumulated recorded subsidy can increase without limit as loan maturity increases.
- For Option C:
  - Recognizes transfer at inception using present value, consistent with international accounting standards.
  - Cumulated transfer/subsidy is bounded (e.g., up to the loan value for zero interest loans).
  - Provides transfer recognition at the correct period (inception), avoiding delay of deficit impact.

---

### NUMERICAL EXAMPLES ILLUSTRATING EFFECTS ON ACCOUNTS

### Bullet loan example (5-year, zero contractual interest, market/discount rate 5 percent)
- Face value: 100 U.
- Option C partition at inception:
  - Loan element: 78.4 U.
  - Grant/transfer element: 21.6 U.
- Under Option B imputed interest each period on constant nominal 100 U: 5 U in year 1 (5 percent of 100).
- Interest accrues and capitalizes on nominal principal under Option C (accrued on 78.4 U initially).
- All three options: same cumulated deficit impact; stock of loan before redemption equals 100 U for all options.
- Imputed interest differs: transfer received/paid is higher in Option B than in Option C because Option B accrues interest on 100 U, Option C on smaller nominal value that grows with capitalization.

### Constant-annuity loan example (5-year, zero contractual interest, market/discount rate 5 percent)
- Face value: 100 U; annual constant annuity repayments: 20 U.
- Option C partition at inception:
  - Loan element: 86.6 U.
  - Grant/transfer element: 13.4 U.
- Under Option B interest is imputed over the decreasing stock and neutralized by imputed transfer.
- All three options: same cumulated deficit impact.
- Imputed interest higher under Option B because it accrues on initial nominal 100 U versus 86.6 U under Option C.

### Example including creditor own financing costs
- Creditor funds with a 5-year funding loan with contractual interest at market rate 5 percent, generating annuities {23.1, 23.1, 23.1, 23.1, 23.1}.
- Creditor interest funding costs cumulate to 15.5 U over the years.
- For the creditor:
  - Options A and B: same deficit profile equal to interest accrued and paid on creditor’s financing (cumulated 15.5 U).
  - Option C: one-off inception impact (13.4 U) and thereafter smoother deficit impact that accounts for both interest on creditor’s liability and imputed interest receivable on concessional loan nominal 86.6 U.
- All three options: same cumulated deficit impact for the creditor: 15.5 U.

### Exact numeric values preserved from examples
- Face value loan: 100 U.
- Market/discount rate: 5 percent.
- Bullet loan partition under Option C: 78.4 U loan element; 21.6 U grant/transfer element.
- Constant-annuity loan partition under Option C: 86.6 U loan; 13.4 U grant/transfer.
- Imputed interest figure in examples: 5 U in year 1 on 100 U (5 percent).
- Creditor funding cumulated interest cost: 15.5 U.
- Constant annuity repayment each year: 20 U.
- Creditor funding annuities (own financing): {23.1, 23.1, 23.1, 23.1, 23.1}.

---

### PRACTICAL IMPLEMENTATION, CAPACITY, AND FITT CONSULTATION OUTCOMES

### Practical challenges
- Implementing Options B or C is challenging for countries with low statistical capacity; Option C requires statistical infrastructure and can be resource intensive where many concessional loans exist.
- Compilers using IPSAS-compliant financial statements could automate implementation using accounting data.

### FITT consultation outcomes
- Issue 1: majority support for Option C; Options A/A1 rejected because macroeconomic statistics should explicitly capture transfers in core accounts.
- Preference for Option C over Option B: recognizes transfer at inception and provides present value consistent with international accounting standards.
- Minority views: two members favored Option A1 for more detailed information; two members preferred either Option B or C.
- Issue 2: majority support for Option B; some members favored consistency across Issues 1 and 2; others argued beneficiary MDBs warrant separate guidance or that guidance belongs in compilation manuals rather than SNA/BPM.

---

### ANNEX III — DEFINING AND MEASURING CONCESSIONALITY (SUMMARY)

### Conceptual framing
- Concessional lending defined as lending with an interest rate "below market interest rates"; requires a discount rate to measure transfer (grant) component.
- Degree of concessionality can be assessed as difference between loan terms and one of:
  - (i) the debtor’s actual funding terms,
  - (ii) the creditor’s own funding terms, or
  - (iii) a typical market rate.
- Consistent choice of reference discount rate is important to avoid cross-border asymmetries.
- In Paris Club / BPM6 context, concessionality for debt reorganization usually calculated against a common discount rate, typically the OECD’s CIRR.

### Options for discount rate for new loans
- Option A: debtor’s typical financing cost.
- Option B: a market-rate.
- Option C: creditor’s observed financing cost.
- Option D: commonly agreed CIRR.

### Trade-offs and GN proposal for new loans
- Option A measures benefit from debtor viewpoint but varies by debtor credit risk and can be circular.
- Option B avoids per-debtor variability but multiple candidate market rates complicate implementation.
- Option C is easily assessable for governments and directly measures creditor-conveyed benefit; raises measurement questions for MDBs (inclusion of grants/concessional contributor funds).
- Option D (CIRR) aligns with current macroeconomic orientations and is easily observable.
- GN proposal: refer to both Options C and D as viable for new concessional lending; for loans by international agencies, measure concessionality against their financing cost (Option C); for contributions to MDBs, Option A may be more relevant.

### Discount rate options for debt restructuring
- Option A: same rule as for new loans.
- Option B: use the CIRR as in BPM6.
- Option C: use the original interest rate/present value of the loan (original contract rate).
- GN proposal for restructuring: follow Option C — use original contract interest rate as reference.

### Materiality thresholds
- Options:
  - Option A: 25 percent threshold (OECD proposal).
  - Option B: lower threshold of 5 or 10 percent.
  - Option C: no threshold, with implementation flexibility.
- GN stance:
  - Rejects Option A (25 percent).
  - If creditor borrowing rate reference is used, no threshold is necessary (Option C).
  - If market-rate reference is used, a 5 or 10 percent threshold could be useful (Option B).

### Measurement practicalities and special cases
- Preferable creditor funding-cost measure: marginal funding costs for same maturity and time.
- Plausible alternative: average funding cost (reduces scope of concessionality).
- MDBs: funding-cost measurement complicated by grant funding of MDBs.
- Distinction: new loans lack historical contract rate; restructurings have original contract rate available.

### Interaction with IPSAS (summary)
- IPSAS prescribes recording fair (discounted) value of concessionary loans on balance sheets of both debtor and creditor, recognizing at inception revenue (or liability) for debtor and expense for creditor; subsequent recognition involves interest measured above contractual facial interest.
- IPSAS uses a 10 percent threshold to decide derecognition upon restructuring (if change in present value <10 percent, loan not derecognized and original rate used).

### GN practical recommendations (summary)
- For new concessional lending:
  - Prefer referencing creditor financing cost (Option C) or CIRR (Option D).
  - Use marginal funding cost where feasible; average funding cost acceptable if marginal impractical.
  - For MDB-originated loans, measure concessionality against MDB financing cost; for MDB contributions, debtor financing cost (Option A) may be relevant.
- For debt restructurings:
  - Use original contract interest rate (Option C).
- On thresholds:
  - Avoid high fixed thresholds (e.g., 25 percent). If creditor-cost reference is used, no threshold is necessary. If market-rate reference is used, consider 5 or 10 percent.

---

### ANNEX VI — RECORDING THE TRANSFER/GRANT ELEMENT AND ANNEX VII — RESTRUCTURINGS, ANNEX VIII — SALE IMPACT

### Recording transfer element (Annex VI)
- Recording choices framed as transaction-type selection:
  - Option A: Current transfers.
  - Option B: Capital transfers.
- Observations:
  - Gift element from government to an international organization could be recorded as a capital transfer (paralleling outright transfers to MDBs) (2008 SNA, paragraph 8.38).
  - Grant element may be seen as current transfer when recurrent and purpose-aligned (2008 SNA, paragraph 22.99 (c); BPM6, paragraph 12.51).
  - Case-by-case analysis recommended to determine appropriate non-financial transaction.
- GN proposes to address recording choice after progress on Issues 1 and 2.

### Treatment in debt reorganization (Annex VII)
- Options for reorganizations that render existing loans concessional:
  - Option A: Apply same treatment as for new concessional lending.
  - Option B: Apply a different treatment (e.g., spreading the transfer).
- Considerations:
  - BPM6 paragraphs A2.67–A2.70 do not differentiate loans concessional at inception from those made concessional by restructuring.
  - Arguments exist both for differential and consistent treatment.
  - Possible compromise: restrict any exemption for restructurings to government-to-government loans.
- GN recommendation: Option A (apply same treatment), while offering alternative voting sub-options.

### Impact of sale of zero interest rate loans (Annex VIII)
- Purpose: illustrate accounting impact on grantor of resale of zero interest loans; GN paragraph 36 argues Options A and B distort net lending/net borrowing in case of secondary sale, Option C does not.
- Numerical assumptions:
  - 5-year bullet loan of 100 with discount rate 5 percent (creditor funding rate).
  - Creditor funds loan by issuing a 5-year zero coupon bond issued at 100 with face value 127.63.
- Key quantitative findings:
  - "Lend and hold": Options A, B, C yield identical cumulated B.9 over time: -27.63 (total interest costs of creditor’s financing).
  - Option C front-loads net assets: starts at -21.65 while Options A and B start with 0 net assets.
  - "Lend and sell" with immediate resale at market value/NPV at inception (78.35):
    - Option A cumulated B.9 = -5.98 instead of -27.63 in lend-and-hold; whole gift element of 21.65 can evade the deficit.
    - Option C: lend-and-hold and lend-and-sell yield identical cumulated B.9 and annual B.9; Option C reflects in the deficit the effective change in net worth regardless of resale timing.
    - Option B: yields same outcome as Option A because subsidy and matching increased interest revenue are imputed flows that cannot be separated upon sale.
- Resale price components:
  - First component: initial discount amortized to date using inception discount rate (accrued under Option C).
  - Second component: difference due to change in market rates (recorded as revaluation).
- ESA 2010 approach: where initial discount not recorded in deficit (Options A/B), a capital transfer at time of sale may be appropriate in some circumstances.

---

### QUESTIONS FOR DISCUSSION (AS POSED BY THE GN)
- Which option do you favor for the statistical treatment of concessional lending?
- Which option do you favor for the statistical treatment of the grant element of concessional loans provided as substitutes of contributions to agencies?
- Do you support changing terminology "concessional loans" to "concessionary loans" in updates to SNA/BPM?
- Do you support clarifying in SNA/BPM that scope of concessional loans is limited to loans granted by creditors that are nonmarket or that conduct their loans on behalf of another nonmarket unit?
- Which discount rate(s) do you favor to define and measure concessionality for new concessional loans and cases of debt reorganization (Annex III)?
- Do you support that the option recommended for new concessional loans also apply to restructured loans (Annex VII)? If not, what alternatives do you support?
- Any other views on statistical treatment of concessional lending and the grant element of concessional loans provided as substitutes of contributions to agencies?

*Source: Guidance Note (GN) and FITT consultations.*

### SECTION I: THE ISSUES

### SECTION I: THE ISSUES

### Background
- BPM6 and 2008 SNA identify debt concessionality as a topic for research (BPM6, paragraph 1.43 (h); 2008 SNA, paragraph A4.44).
- Concessional loans are described as lending intentionally provided at a contractual interest rate below market interest rates, for similar grace and repayment periods, with the purpose to convey a benefit, occurring in a non-commercial context (BPM6, paragraphs 3.79 and 12.51; 2008 SNA, paragraphs 3.131, 3.134, 22.123, and 22.124; GFSM 2014, paragraphs 3.123 and A3.39-41).
- Concessional loans may also be designed with other characteristics aimed at conveying a benefit, such as more favorable grace and maturity periods (2008 SNA paragraph 22.123; BPM6 A2.68; EDS Guide paragraph 6.22).
- Concessional lending is frequently observed in government accounts and international organizations; it can be provided at inception or arise later in debt restructurings. Concessional lending also appears at the national level (e.g., student loans, housing loans) and in the private sector, but private-sector low-interest arrangements typically do not qualify as concessional loans when the difference between fair value and redemption value is not intended as a transfer.
- The Guidance Note (GN) restricts its scope to low interest loans provided in a non-commercial context, including entities acting on behalf of others, with an intention to convey a benefit, and addresses valuation and the treatment of the grant element. It recommends that lenders with very low average funding costs (e.g., some development or international agencies, including IMF examples cited) are de facto not extending concessional loans in the GN’s meaning.

### Link to valuation guidance
- The recording of concessional loans relates to GN F.9 “Valuation of Loans (Fair Value)” insofar as GN F.9 elaborates on loan valuation in general; GN F.9 focuses on re-estimates in loans value, whereas the present GN deals with valuation at inception.

### Issues for discussion
- Manuals recognize a transfer element in concessional loans but generally do not prescribe recording any expense (from creditor) or revenue (for debtor) in the core accounts to capture the transfer; instead the transfer element is typically reported as a memorandum item in supplementary tables.
- Current practice records concessional loans in balance sheets of debtor and creditor at inception at face/nominal value rather than discounted/fair value, which contrasts with treatment of low-interest commercial loans or employer-employee low-interest loans.
- There is a mismatch between acknowledgment of a gift/transfer and absence of an explicit/uniform recommendation for recording the transfer in core accounts.
- The GN considers three alternative recording options for concessional loans and discusses the appropriate discount rate to measure the transfer component.

### Issue 1 — Options for statistical treatment of concessional lending
- Option A: No change. Continue recording concessional loans at face/nominal value at inception with no imputation; transfer element remains memorandum/supplementary information.
  - Sub-option A1: No change in core accounts but provide more detailed memorandum/supplementary information on the transfer element, notably the fair value of the loan and the entailed accrued interest.
- Option B: Record the loan at face/nominal value at inception, but recognize the transfer element in the core accounts by increasing the stream of interest earned (D.41) on the loan using a suitable non-concessional rate and recording a matching transfer expense so that the net effect is deficit-neutral in every accounting period.
- Option C: Partition concessional loans at inception between a “genuine” loan element (F.4) and an explicit “transfer element”; subsequently impute interest receivable (D.41) on the loan that capitalizes on the new nominal value/principal over time so that the loan’s nominal principal reaches face value before maturity; deficit-neutral across the life of the loan.

- Numerical example summarized (full details in Annex II):
  - Creditor grants on January 1 a bullet-loan (5 years, redeemed on December 31) of 100 Units with contractual zero interest.
  - Creditor’s own average funding interest rate used as discount rate: 5 percent.
  - Net present value at inception of the payment stream {0, 0, 0, 0, 100} at a 5 percent discount rate is 78.4 U.
  - Transfer recorded at inception equals 21.6 (=100-78.4).
  - Under all Options, end Year 5, just before redemption, the instrument is valued at face value 100.

- Additional considerations:
  - The GN recommends extending the same rules to debt restructurings where changes (e.g., grace period) render loans concessional (see Annex VII).
  - Determining the most suitable nonfinancial transaction to record the transfer element requires case-by-case analysis of the underlying operation (Annex VI).
  - Appropriate discount rate candidates include: creditor’s observed funding cost, the OECD’s Commercial Interest Reference Rates (CIRR), or the borrower’s funding cost.
  - The GN takes a conservative approach and recommends using the average cost of the grantor/creditor as the discount rate; it recognizes CIRR as aligned with current manual orientations and as easily observable.

### Issue 2 — Grant element when concessional loans substitute contributions to agencies
- Situation: Governments sometimes provide official contributions to agencies as long-term zero-interest loans, with beneficiaries calculating an implied “grant element.”
- Many compilers treat such concessional loans as substitutes for regular contributions and consider the grant/transfer element of the loan to be recognized in the core accounts at disbursement to uphold substance over form.
- Consistent recording of transfers is required to avoid incentives to substitute transfers with low-interest loans.
- Where contributions are recorded as equity (e.g., paid-in capital), concessional loans provided to such beneficiaries should be partitioned at inception between a loan and an equity element.
- If Option C for Issue 1 is rejected, two options for Issue 2 are proposed:
  - Option A: No specific rule for cases where concessional loans substitute for contributions/transfers.
  - Option B: Require explicit recording in core accounts at inception of the grant/transfer element when concessional loans are substitutes for regular contributions to ensure substance over form.

### Outcomes and recommendation
- Recommendation for Issue 1: Option C is recommended.
  - Rationale:
    - Consistent with the economic principle of time value of money.
    - Partitioning at inception follows substance over form and better represents net assets of creditor and debtor over time (a view shared by IPSAS).
    - Memorandum items in manuals reference recording the transfer element at inception, suggesting preference for Option C.
    - Option C does not imply continuous market-price present-value re-estimates during the loan’s lifetime; it is consistent with normal loan valuation where the contract rate is an appropriate discount rate.
  - Practical advantage:
    - Prevents the transfer from escaping deficit (net lending/net borrowing) if the creditor later sells the concessional loan at a steep discount; recording the transfer at inception addresses counterintuitive outcomes where the deficit impact would otherwise be deferred until resale.
  - Counterargument addressed:
    - BPM6 paragraph A2.69 notes potential revisions if a concessional loan is retired before maturity; while this could support Option B, early redemption replacing concessional debt with more onerous debt is unlikely, and early reimbursement typically reflects market-rate declines, which should be treated like any early loan redemption.

*Prepared by Philippe de Rougemont and Martim Assunção (Eurostat), Paula Menezes (Banco de Portugal) and Nabila Akhazzan and Rita Mesias (International Monetary Fund, Statistics Department (STA)).*

### 22. Similarly, it is argued that Option C is unwarranted as it could allow debtors or creditors to

### f15-debt-concessionality - 22. Similarly, it is argued that Option C is unwarranted as it could allow debtors or creditors to

### Comparative assessment of Options A, B, and C (treatment of concessional loans)
- Core analytical point:
  - Option C recognizes at inception a partition between a loan element and a grant/transfer element; Options A and B do not. The transfer is "definitely extended by the creditor and acquired by the debtor at inception."
  - Main objection to Option C is that it can delay the deficit impact of the grantor relative to Options A and B.
- Key arguments for Option A (status quo):
  - Merit of simplicity.
  - The creditor’s “cost” is proxied by difference between low interest earned and higher funding costs; net assets deteriorate automatically over time.
  - Nominal valuation requires recording the loan in the balance sheet at inception for amounts lent.
- Key arguments for Option B:
  - Corrects misreporting of expenditure by recognizing a transfer from creditor to debtor in the core accounts spread over the life of the instrument (including COFOG allocation).
  - Builds on 2008 SNA paragraph 7.126 (special case of interest rates set by the central bank) and extends that approach to government lending at below market rates (including lending on behalf of government).
  - Recognizes transfers spread over time consistent with interest accrual.
  - Limitation: overestimates the subsidy component because it does not use a present value approach; cumulated transfer/subsidy recorded under Option B can increase without limit as loan maturity increases.
- Key arguments for Option C:
  - Recognizes the transfer at inception using present value, yielding the present value of the loans consistent with international accounting standards.
  - Cumulated transfer/subsidy is bounded (e.g., up to the loan value for zero interest loans) under Option C.
  - Provides transfer recognition at the correct period (inception), avoiding delaying deficit impact for the grantor.

### Numerical examples illustrating effects on accounts (Bullet loan and Constant-annuity loan)
- Bullet loan example (5-year, zero contractual interest, market/discount rate 5 percent):
  - Face value: 100 U.
  - Option C partitions at inception into:
    - Loan element: 78.4 U.
    - Grant/transfer element: 21.6 U.
  - Under Option B interest imputed each period on the constant nominal 100 U: 5 U in year 1 (5 percent of 100).
  - Interest accrues and capitalizes on the nominal principal under Option C (accrued on 78.4 U initially).
  - All three options have the same cumulated deficit impact; stock of loan at (before) redemption equals 100 U for all options.
  - Imputed interest differs: amount of transfer received/paid is higher in Option B than in Option C because Option B accrues interest on 100 U, Option C on a smaller nominal value that grows with capitalization.
- Loan with constant annuity example (5-year, zero contractual interest, market/discount rate 5 percent):
  - Face value: 100 U; annual constant annuity repayments: 20 U.
  - Option C partitions at inception into:
    - Loan element: 86.6 U.
    - Grant/transfer element: 13.4 U.
  - Under Option B interest is imputed over the decreasing stock (reducing by 20 U each year) and neutralized by imputed transfer.
  - Options A and B show same deficit profile (Option B shows more expense/revenue due to imputations); Option C shows one-off impact at inception (13.4 U) and reverse deficit impact thereafter via imputed interest.
  - All three options have the same cumulated deficit impact.
  - Imputed interest is higher under Option B because it accrues on initial nominal 100 U versus 86.6 U under Option C.
- Example including creditor own financing costs:
  - Creditor takes a 5-year funding loan with contractual interest at market rate 5 percent, generating annuities {23.1, 23.1, 23.1, 23.1, 23.1}.
  - For the creditor the interest funding costs cumulate to 15.5 U over the years.
  - For the creditor:
    - Options A and B exhibit the same deficit profile equal to the interest accrued and paid on the creditor’s loan liability (cumulated 15.5 U).
    - Option C exhibits a one-off inception impact (13.4 U) and a thereafter smoother deficit impact that accounts for both interest on the creditor’s liability and the imputed interest receivable on the concessional loan with nominal value 86.6 U.
  - All three options have the same cumulated deficit impact for the creditor: 15.5 U.
- Exact numeric values preserved from examples:
  - Face value loan: 100 U.
  - Market/discount rate used: 5 percent.
  - Bullet loan partition under Option C: 78.4 U loan element; 21.6 U grant/transfer element.
  - Constant-annuity loan partition under Option C: 86.6 U loan element; 13.4 U grant/transfer element.
  - Imputed interest figure in examples: 5 U in year 1 on 100 U (5 percent).
  - Creditor funding cumulated interest cost: 15.5 U.
  - Constant annuity repayment each year: 20 U.
  - Creditor funding annuities (own financing): {23.1, 23.1, 23.1, 23.1, 23.1}.

### Practical implementation, capacity, and consultation outcomes
- Practical challenges:
  - Implementing Options B or C is challenging for countries with low statistical capacity; Option C requires statistical infrastructure and could be resource intensive where many concessional loans exist.
  - Compilers using IPSAS-compliant financial statements (or equivalent) could automate implementation using accounting data.
- FITT consultation outcomes:
  - Issue 1: majority support for Option C. Options A/A1 rejected because macroeconomic statistics should explicitly capture all transfers extended by government within the core accounts.
  - Preference for Option C over Option B because Option C recognizes transfer at inception and provides present value consistent with international accounting standards.
  - Minority views: two members favored Option A1 for more detailed information; two members preferred either Option B or C.
  - Issue 2: majority support for Option B. Some members favored consistent treatment across Issues 1 and 2; others argued beneficiaries (often MDBs) are unique and warrant separate guidance or that guidance should be in compilation manuals rather than SNA/BPM.

### Recommendation for Issue 2 and policy stance on recording
- This Guidance Note (GN) recommendation:
  - The transfer element of concessional loans granted as substitutes for regular or other transfers/grants should be recorded in the core accounts consistently with regular contributions or other transfers/grants.
  - Macroeconomic statistics manuals should explicitly clarify this recording.
  - Recording at face value low/zero-interest rate loans extended by donors in place of outright grants should not be permitted because it defers or changes the deficit impact by engaging in a financial operation—described as an "outright fiscal illusion maneuver."
  - Principle: substance over form; paying contributions owed through direct payment or through a larger payment to be then refunded should impact the deficit for the same amount and at the same time.
- Recommended option for Issue 2: Option B (majority support in FITT consultation), with some members advocating consistency with Issue 1 treatment or separate guidance for beneficiary institutions.

### Other recommendations and terminology
- Proposal to switch terminology in updated SNA/BPM from "concessional" to "concessionary" loans to enhance convergence with IPSAS.
- Proposal to clarify that the scope of concessionary loans is limited to loans granted by creditors that are nonmarket or that conduct their loans on behalf of another nonmarket unit.

### Questions for discussion (as posed)
- Which option do you favor for the statistical treatment of concessional lending?
- Which option do you favor for the statistical treatment of the grant element of concessional loans provided as substitutes of contributions to agencies?
- Do you support the proposal to change the terminology "concessional loans" to "concessionary loans" in the update to the SNA/BPM?
- Do you support the proposal to clarify in the SNA/BPM that the scope of concessional loans is limited to loans granted by creditors that are nonmarket or that conduct their loans on behalf of another nonmarket unit?
- Which discount rate(s) do you favor to define and measure concessionality for new concessional loans and cases of debt reorganization (Annex III)?
- Do you support the proposal that the option recommended for new concessional loans should also be applicable to cases of restructured loans (Annex VII)? If not, what alternative option(s) do you support for cases of restructuring?
- Any other views on the statistical treatment of concessional lending and the grant element of concessional loans provided as substitutes of contributions to agencies?

*Source: This Guidance Note (GN) and FITT consultations.*

### Annex III. Defining and Measuring Concessionality

### Annex III. Defining and Measuring Concessionality

### Definition and conceptual framing
- Concessional lending is defined as lending with an interest rate "below market interest rates", but the precise meaning of "below market interest rates" requires a discount rate to measure the transfer (grant) component.
- The degree of concessionality can be assessed as the difference between the loan terms provided to the debtor and one of:
  - (i) the debtor’s actual funding terms,
  - (ii) the creditor’s own funding terms, or
  - (iii) a typical market rate.
- Consistent choice of reference discount rate is important to avoid cross-border asymmetries.
- In the Paris Club / BPM6 context, concessionality for debt reorganization is usually calculated against a common discount rate, typically the OECD’s CIRR.

### Options for the discount rate (new loans)
- Issue: In the context of new loans, which discount rate should define and measure concessionality?
- Options presented:
  - Option A: the “typical” financing cost of the debtor.
  - Option B: a market-rate.
  - Option C: the observed financing cost of the creditor.
  - Option D: the commonly agreed CIRR.
- Practical and conceptual trade-offs highlighted:
  - Option A (debtor financing cost) measures the benefit from the debtor’s viewpoint but:
    - Imparts varying transfer elements by debtor credit risk.
    - May inflate the scope of concessional loans.
    - Is circular where debtor financing costs are influenced by the existence of concessional lending.
  - Option B (market rate) captures average borrower conditions and avoids per-debtor variability, but:
    - Multiple candidate market rates complicate implementation.
    - Measures benefit relative to market conditions rather than the creditor’s actual concession.
    - Could be operationally addressed by specifying a frequently published reference rate in manuals.
  - Option C (creditor financing cost) is:
    - Easily assessable (especially for government creditors).
    - Directly measures the benefit conveyed by the creditor.
    - May be problematic where debtors have extremely limited market access, but a below-creditor-rate loan indicates an undisputed benefit.
    - Can be measured using marginal funding costs for same maturity and timing or by average funding cost (the latter likely reduces scope of concessionality).
    - Raises specific measurement questions for MDBs because MDBs often finance themselves with grants or concessional contributor funds; whether grants on the liability side should be included in MDB funding cost is an issue.
  - Option D (CIRR) aligns with current macroeconomic statistics orientations and is:
    - Easily observable by NSIs.
    - Generally low, reflecting very low credit risks.

- Guidance Note proposal for new loans:
  - Refer to both Options C and D as viable for new concessional lending because both are conceptually sound and imply generally low discount rates, limiting concessionality to undisputed subsidized funding.
  - For loans by international agencies (e.g., World Bank, IMF), measure concessionality against their financing cost (Option C).
  - For contributions to MDBs or similar, Option A may be more relevant because each MDB calculates grant element using its own funding cost.

### Options for the discount rate (debt restructuring)
- Issue: For debt reorganization, which discount rate should be used?
- Options presented:
  - Option A: apply the same rule as for new loans.
  - Option B: use the CIRR as in the BPM6.
  - Option C: use the original interest rate/present value of the loan (i.e., the original contract rate).
- Observations and trade-offs:
  - Debt restructuring has the original contract rate available, simplifying measurement of the grant component.
  - Using a market rate could convert holding gains into transactions (income or capital transfers), which is generally not recommended.
  - IPSAS guidance recommends using the original (market) rate of the contract for measuring concessionality in renegotiations when the loan is not derecognized (i.e., restructuring changes present value by less than 10 percent). If the loan is derecognized, use the market rate observed at restructuring.
  - 2008 SNA refers to the CIRR for debt restructuring, but this may understate transfer if debtor risk is worse than average.
- Guidance Note proposal for debt restructuring:
  - Follow Option C: use the original interest rate of the contract as the reference to assess whether a benefit is conveyed and to measure the benefit in restructurings.

### Materiality thresholds (use of a grant-element threshold)
- Issue: Should a materiality threshold be used to decide concessionality?
- Options presented:
  - Option A: 25 percent threshold between face value and present value as proposed by OECD.
  - Option B: lower threshold of 5 or 10 percent.
  - Option C: no threshold, with implementation flexibility.
- Guidance Note stance:
  - Rejects Option A (25 percent) because thresholds can create incentives to circumvent accounting impacts and 25 percent is large.
  - If compilers use creditor borrowing rate (Issue a, Option C), a threshold is unnecessary: material lending below cost should establish concessionality without a threshold (Issue c, Option C).
  - If compilers use a market rate (Issue a, Option B), then a 5 or 10 percent threshold could be useful (Issue c, Option B).

### Measurement practicalities and special cases
- Measuring creditor funding cost:
  - Preferable method: marginal funding costs for the same maturity and at the same time.
  - Plausible alternative: average funding cost (reduces scope of concessionality).
- MDBs and funding-cost measurement:
  - MDBs often fund themselves via grants or concessional contributor funds; whether such grants are included in MDB funding cost affects assessment of concessionality.
- Distinction between new loans and restructurings:
  - New lending lacks an explicit historical contract rate; reference-rate choice is more ambiguous.
  - Restructuring has the original contract rate available, making the grant component straightforward to establish.

### Interaction with IPSAS and accounting treatments (summary from Annex IV)
- IPSAS defines concessionary loans as loans granted/received at below market terms with intent at the outset to provide/receive resources at below market terms, distinct from waiver of debt.
- IPSAS prescribes:
  - Recording the fair (discounted) value of concessionary loans on the balance sheets of both debtor and creditor.
  - Recognizing at inception: revenue (or liability) for the debtor and an expense for the creditor.
  - Subsequent recognition: interest expense for the debtor and interest revenue for the creditor higher than the contractual facial interest.
- IPSAS treatment examples and thresholds:
  - IPSAS uses a 10 percent threshold to decide derecognition upon restructuring (irrespective of concessionality): if restructuring changes present value by less than 10 percent, the loan is not derecognized and the original rate is used for measuring the effect.
  - Example (IPSAS 23.IG54): a currency unit (CU) 5 million 5-year 5 percent loan, with a comparable market interest rate of 10 percent, produces a non-exchange revenue for the borrower at inception equal to CU 784,550; the loan is recorded at CU 4,215,450 instead of CU 5 million.
- IPSAS distinguishes concessional loans (ex-ante) from waivers of debt (ex-post) and prescribes separate recognition for exchange and non-exchange components.

### Practical implications and recommendations (guidance)
- For new concessional lending:
  - Prefer referencing creditor financing cost (Option C) or CIRR (Option D) to limit scope to undisputed subsidized funding and simplify implementation.
  - Use marginal funding cost of creditor where feasible; average funding cost is acceptable if marginal cost is impractical.
  - For MDB-originated loans, consider measuring concessionality against MDB financing cost; for MDB contributions, debtor financing cost (Option A) may be more relevant.
- For debt restructurings:
  - Use the original contract interest rate (Option C) to determine and measure the benefit.
  - Avoid using market rates that could reclassify holding gains as transactions.
- On thresholds:
  - Avoid high fixed thresholds (e.g., 25 percent). If creditor-cost reference is used, no threshold is necessary. If market-rate reference is used, consider a 5 or 10 percent threshold.

*International Monetary Fund — Annex III. Defining and Measuring Concessionality*

### 9.      The correct recording is to consider that the P.11 is the present value of the cash to be collected

### f15-debt-concessionality - 9. The correct recording is to consider that the P.11 is the present value of the cash to be collected

### Recording of commercial loans at low interest rates
- The GN records P.11 as the present value of the cash to be collected from the client: 7.500 U (approximately), which yields B.2.n = 500 U at time of sale/production.
- In subsequent years:
  - B.1 = B.2n = 0
  - B.101 = 0
  - The bank earns interest of 500 U, which generates no FISIM if the interbank rate is presumed to be 5 percent.
- The GN presents a comparative table (distorted recording versus correct recording) illustrating entries such as:
  - P.1 entries: 110,000-5000 7,500 7,5000 7,500
  - D.17 entries: 7,0000 7,000 7,000 7,000
  - F.2 and F.3 entries showing 7,500 and 500 adjustments in production and subsequent recordings.

### Distortions from a narrow SNA interpretation (loans recorded always at face value)
- A narrow reading of SNA on nominal value of loans would distort:
  - The B.1, B.2n, B.10.1, and B.9 of the carmaker/group, frontloading future profits.
  - The NACE classification of value added, producing too much car-making activities and too little/negative FISIM.
- Because value added is distorted over time under the narrow SNA interpretation, the difference enters the GDP deflator:
  - The GN states that the (erroneous) approach increases the GDP deflator when low interest rate loans are granted.
  - Intuition: low interest rate loans act as a discount on cars that should reduce household consumption and GDP deflators; failing to recognize this inflates the two deflators.

### Annex VI — Recording of the transfer/grant element in concessional loans
- The GN frames the recording question as choosing among transaction types (or combinations):
  - Option A: Current transfers.
  - Option B: Capital transfers.
- Observations and guidance:
  - The gift element of a concessional loan from a government to an international organization could be recorded as another capital transfer, paralleling outright transfers from governments to MDBs (2008 SNA, paragraph 8.38).
  - Capital transfers in cash can require the second party (the MDB) to use funds for acquisition of assets (ESA 2010, paragraph 4.146).
  - The grant element is also seen by some as a current transfer (2008 SNA, paragraph 22.99 (c); BPM6, paragraph 12.51) because of recurrent nature and typical purposes (e.g., disaster aid).
  - Typical distinctions: capital transfers are usually large and infrequent; current transfers normally smaller and regular (2008 SNA, paragraph 8.38), though largeness/irregularity is typical rather than a strict condition (2008 SNA, paragraph 10.200).
  - If a concessional loan supports acquisition of assets by the recipient, recording as a capital transfer is straightforwardly appropriate.
  - Grants from government to non-financial corporations could be recorded as subsidies on production if aimed at covering production costs.
  - In commercial transactions, the one-off element is most straightforwardly recorded as reduction to sales (P.11).
- Recommendation on approach:
  - The GN stresses case-by-case analysis of the exact nature of the operation underlying the loan to choose the appropriate non-financial transaction.
  - The GN proposes to address the recording issue later, after progress on Issues 1 and 2.

### Annex VII — Treatment of concessional lending in debt reorganization
- The GN poses whether new concessional-loan rules should extend to refinancing, restructuring, or rescheduling that render existing loans concessional.
- Options:
  - Option A: Apply the same treatment as for new concessional lending.
  - Option B: Apply a different treatment (for example, spreading the transfer).
- Key considerations:
  - BPM6 paragraphs A2.67 to A2.70 do not differentiate loans concessional at inception from those made concessional by restructuring.
  - Arguments for differential treatment:
    - In restructurings, the creditor undisputedly intends to convey a benefit to the debtor; for new concessional loans, intent may be less clear.
  - Arguments for consistent treatment:
    - A reduction in promised cash flows is substantively a transfer/grant at time of reduction, regardless of timing or whether it reduces interest or principal.
    - Different treatment in restructurings could incentivize fiscal illusion (e.g., governments avoiding deficit impact by restructuring later).
  - Possible compromise:
    - Restrict any exemption for restructurings to government-to-government loans to limit gaming while avoiding undue exemptions for government operations with public corporations.
- GN recommendation:
  - Recommends Option A while offering alternative options for vote via Sub-options B.b and C.c in Issue 1.

### Annex VIII — Impact of sale of zero interest rate loans (numerical example and option comparisons)
- Purpose:
  - Illustrate accounting impact on grantor (government) of resale of zero interest loans.
  - GN paragraph 36 argues Option A and B distort net lending/net borrowing in case of secondary sale, while Option C does not.
  - ESA 2010 (following Option A together with 2008 SNA and BPM6) recommends recording a capital transfer at time of sale to correct this in some circumstances.
- Numerical assumptions and setup:
  - 5-year bullet loan of 100 with a discount rate of 5 percent, which is also the funding rate of the creditor.
  - Creditor funds loan by issuing a 5-year zero coupon bond issued at 100 with face value 127.63.
  - Scenarios compared:
    - "Lend and hold" (creditor holds asset to maturity).
    - "Lend and sell" (loan sold immediately after issuance).
- Key quantitative findings:
  - In the "lend and hold" scenario, Options A, B, C yield identical cumulated B.9 over time: -27.63 (total interest costs of creditor’s financing).
  - Option C differs in time profile (front loading): net assets under Option C start at -21.65 (difference between face value and nominal/NPV at inception), while Options A and B start with 0 net assets.
  - In "lend and sell" with immediate resale at market value/NPV at inception (78.35):
    - Creditor uses 78.35 to redeem its financing; creditor retains a loan net liability of 21.65 after the sale.
    - Option A results: cumulated B.9 = -5.98 instead of -27.63 in the lend-and-hold scenario.
      - The -5.98 corresponds to the sole carrying cost of the initial gift of 21.65 (26.63 = 21.65 + 5.98).
      - With Option A, immediate resale causes the whole gift element of 21.65 to evade the deficit.
      - If resale occurs mid-maturity, only part (around half) of the gift would evade the deficit.
    - Option C results:
      - "Lend and hold" and "lend and sell" lead to identical cumulated B.9 and identical annual B.9.
      - In "lend and sell", D.41R is netted from D.41P—no claim remains and only carrying cost of the gift is borne.
      - Option C always reflects in the deficit the effective change/decrease in net worth, regardless of resale timing.
    - Option B results:
      - Option B yields the same outcome as Option A because its "subsidy" and matching "increased interest revenue" approach is an imputation of two matching flows that cannot be separated upon sale; cumulated subsidy would otherwise be too large.
- Table-derived asset/liability numeric entries illustrate time paths under each option, including lines such as:
  - Asset/Liability entries showing sequences like 100 100.00; 100 105.00 Year 1; 100 110.25 Year 2; 100 115.76 Year 3; 100 121.55 Year 4; 0 27.63 Year 5; Sum 27.63.
  - Option C and sell: asset nominal/face and liability sequences showing 78.35 100 100.00; subsequent year values 82.27 100 105.00 Year 1 etc., and sum rows showing 27.63Sum 21.65 27.63 21.65-27.63 27.63Sum 0.00 5.98 21.65-27.63 (reflecting decomposition of holding loss vs. carrying cost).
- Consideration of resale price components:
  - Resale price typically reflects:
    - First component: initial discount amortized to date using the inception discount rate.
    - Second component: any difference in discount rate (market rate changes).
  - Under Option C:
    - First component is already accrued in net lending/net borrowing.
    - Second component is generally recorded in revaluation, potentially decomposed into:
      - Part 1: genuine change in market rates.
      - Part 2: difference because discount rate at inception was not a true market rate.
  - Under Options A and B:
    - Because the initial discount is not recorded in the deficit, it may be appropriate to record a capital transfer at time of sale (as ESA 2010 prescribes in some circumstances, excluding second component part 1).

---


_Source: https://www.imf.org/-/media/files/data/statistics/bpm6/fitt/f15-debt-concessionality.pdf_