Review of Borrowing Framework of Major Federal Government-Backed Entities : 1

## Archived - I  Executive Summary

Archived information

Archived information is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please contact us to request a format other than those available.

Today, seven different federal government-related entities each borrow billions of dollars annually in debt capital markets. These entities comprise the Government of Canada itself and six "Borrowers": four Crown corporations (Business Development Bank of Canada, Canada Mortgage and Housing Corporation, Export Development Canada and Farm Credit Canada) and two non-Crown corporations (Canada Housing Trust and Canadian Wheat Board). While each Borrower issues debt in its own name, all of their debt issues effectively provide recourse to the Government.

In the fall of 2004, KPMG was engaged to: (1) review the performance and governance of the current borrowing framework; and (2) consider alternative borrowing frameworks that could reduce risk or cost to Canada as a whole, while taking into account Borrower mandates and the need for good governance. To complete this engagement, KPMG conducted extensive research and held interviews with various stakeholders, including the Borrowers, the Government, the Bank of Canada, dealers and investors.

### Current Borrowing Framework

Each Borrower has established a well-respected and sophisticated treasury department. These departments have financial risk management policies in place, and are able to meet their primary goals to raise the funds needed by the Borrowers at a reasonable price. Furthermore, while each treasury department is directly accountable to its senior management, board of directors and Minister, it also has accountability to the Minister of Finance, which approves its annual borrowing plan. This accountability is required due to the Government's contingent liability regarding debt issued by the Borrowers, and to ensure that no individual Borrower takes actions that could be detrimental to Canada's AAA credit rating as a whole.

While the current system works well, Finance Canada and each of the Borrowers could make improvements to enhance the status quo. These improvements primarily involve changes to the annual "Borrowing Plans" submitted by the Borrowers, and changes to reporting on financial risk management and borrowing activity. Most of these changes are intended to either improve financial risk management or to support the Minister of Finance's accountability for the collective actual and planned borrowing activity of the seven entities sharing the Canada credit rating.

### Alternative Borrowing Frameworks

To consider alternative borrowing frameworks, we reviewed international and provincial borrowing frameworks, assessed various relevant attributes that must be considered before implementing any framework, and analyzed a number of alternatives. While we did reach conclusions in a number of areas, we were not engaged to rank the attributes, or to recommend any particular framework. As such, it is the ultimate responsibility of the Government to determine which framework would be optimal for Canada's taxpayers.

In considering other jurisdictions, we noted a gradual trend towards centralization of borrowing over the past 15 years. Furthermore, in analysis of possible frameworks, we rejected several options as impractical. Accordingly, our analysis focused on a comparison of the status quo with a full centralization, where one Centralized Funding Entity ("CFE") would borrow on behalf of both the Government and the Borrowers.

We examined the diversity and appropriateness of borrowing products. The Borrowers pay a risk premium of approximately 1 basis point (0.01%) over Government of Canada issues on money market borrowings and approximately 15 basis points (0.15%) on unstructured non-retail term borrowings. To reduce this gap, the Borrowers tend to issue debt opportunistically when market conditions appear favourable, and many Borrowers issue structured notes to lower their cost of funds. While both of these strategies are common for government agencies around the world, they are not common for sovereign issuers and they differ substantially from the debt management strategies of the Government in that they are less transparent to investors and involve more complexity and risk for borrowers. These strategies therefore raise concerns particularly given the absence of a risk-adjusted cost of capital calculation. Suitability issues also must be considered for retail investors in some of the more complex products.

We examined borrowing cost and we found that the Borrowers achieve widely divergent overall costs of funds, largely because they meet their borrowing needs using a variety of products issued to clients in a variety of Canadian and international markets. Our estimates indicate that the Borrowers' direct borrowing costs could be reduced under centralization by approximately $10 million in the first year, accumulating to$33 to $175 million over five years. If figures for CHT, a special purpose trust whose debt obligations are guaranteed by the Government, are included, then these figures increase to$33 million in the first year, accumulating to $330 to$620 million over five years. As stated in the previous paragraph, the Borrowers collectively issue a large amount of structured notes which are inherently higher risk. Therefore, on a risk-adjusted basis, the savings in direct borrowing costs would be substantially higher than the numbers presented in this paragraph.

We examined the quality and liquidity of Canadian capital markets. In an era of declining government borrowing requirements, market participants are becoming concerned about the liquidity of the Government's benchmark Canada Bonds, which are used as a reference to price most fixed income securities issued in Canada. Most market participants feel that the Borrowers' $11 billion in annual supply ($25 billion if CHT is included) could be added to the benchmark issues which would aid market liquidity and could be absorbed by the market without an increase in borrowing cost.

We examined Borrower operations. The Borrowers greatly value their treasury departments, their autonomy, plus the flexibility and market intelligence gained by accessing the capital markets on a daily basis. They are greatly concerned that any CFE could not be sufficiently responsive to their unique needs, and to their required service levels. Their concerns are valid, however many of their concerns may be able to be mitigated with a well-planned CFE.

Finally, while centralized borrowing would not change the collective outstanding market debt of the seven entities per se, the Borrowers' debt is not consolidated in the Public Accounts of Canada for financial reporting purposes and therefore centralization could increase the market debt reported in the Public Accounts of Canada by up to 11%. To reduce the possibility of misinterpretations by rating agencies or the media, a communication strategy would be critical.

Overall, some attributes support centralization, while others support the status quo. Others are more neutral.

If the Government decides to further pursue the possibility of centralization, then further preparation is required, including design and governance of the CFE, risk management at the CFE, required service levels from the CFE to the Borrowers, transfer pricing to the Borrowers, estimation of transition costs and changes to operating costs, and a review of the impact on financial reporting. These issues were not addressed in the scope of this report. Centralization should only proceed if the benefits exceed the costs, with both costs and benefits being considered on both a qualitative and quantitative basis.

## II   Introduction

### Glossary

A glossary of abbreviations and terms used herein is attached as Addendum 1.

### Scope of Work

In November 2004, Finance Canada ("Finance") and the Bank of Canada ("BOC") awarded KPMG LLP ("KPMG") a contract entitled Review of Borrowing Framework of Major Federal Government-Backed Entities (the "Review"). These government-backed entities included four Crown corporations (the "Crowns") - Business Development Bank of Canada ("BDC"), Canada Mortgage and Housing Corporation ("CMHC"), Export Development Canada ("EDC") and Farm Credit Canada ("FCC") - plus two other entities - the Canada Housing Trust ("CHT") and the Canadian Wheat Board ("CWB"). Throughout this report, these six entities are collectively referred to as the "Borrowers", and were included in the Review as each Borrower borrows billions of dollars per year from capital markets. KPMG was asked to perform an objective and factual review of: (1) the performance and governance of the current borrowing framework; and (2) alternative borrowing and governance frameworks, without judgment or recommendation as to the best alternative.

To complete this assignment, KPMG researched and reviewed a large number of proprietary and public documents (a listing is attached as Addendum 2), considered comparable international and provincial borrowers, and held interviews with various stakeholders - Finance, BOC, the Borrowers, major investors, Canadian and international dealers, and others (attached as Addendum 3). KPMG's interview guides are attached as Addendum 4, however, the interviews typically expanded upon such guides. Following the interview process, the analysis stage began, which involved extensive consultation and discussions with Finance, BOC and the Borrowers. KPMG would like to acknowledge and thank the Treasurers of the Borrowers for their high level of cooperation with this process.

This report summarizes the key results of the Review. For most topics, expanded details and analyses are presented in the appendices.

### Background

With the exception of CHT which was founded in 2001, the Borrowers have a long history - FCC was founded in 1927, CWB in 1935, BDC in 1944, CMHC in 1946, and EDC in 1969. While their structures and purposes have been modified somewhat over the years, their overall mandates have not significantly changed. Originally, the Borrowers financed their operations and growth through: (1) chartered bank financing; (2) Consolidated Revenue Fund ("CRF") loans provided by the Government of Canada (the "Government" or "GOC"); and (3) equity provided by GOC. By the 1970's, the GOC's own direct debt grew and, at the direction of GOC, some of the Borrowers began to raise debt on public markets either to replace more expensive bank debt, or to replace CRF borrowings. At that time, market spreads (i.e. the premium over interest rates on GOC's own debt issuance) were as high as 65 basis points (i.e., 0.65%). Finance was heavily involved, approving each issue upon commitment. To avoid competition for potential investors, the Borrowers typically raised financing in international markets. It was also felt that there would be governance benefits by having the Borrowers bear responsibility for issuing their own debt.

By the late 1980's and early 1990's, GOC's needs for debt financing peaked, and all Borrowers became responsible for directly financing their debt in the capital markets. The Borrowers matured, and the capabilities of their treasury departments deepened. Market spreads for the Borrowers (and direct borrowing costs) declined, and while the GOC maintained its governance authority, Finance's role was lessened to more of a gatekeeper, where annual standing approvals replaced individual borrowing authorizations. During this time, the Borrowers developed substantial corporate independence and self-sufficiency.

Over the last eight years, GOC surpluses have led to declining Canada Bond issuance, reducing the current and future liquidity of GOC bond and money markets. At the same time, the amount of capital market borrowings by the Borrowers has increased, particularly with the introduction of CHT's borrowing programs in 2001. The sophistication of the Borrowers' treasury departments has continued to improve as the types of available borrowing products and derivatives have increased, and as best practice risk management and governance procedures have evolved.

What is the situation today? BDC, CMHC, EDC and FCC - as Crowns - are Agents of Her Majesty in Right of Canada, and as such, their debt obligations are effectively obligations of GOC. CWB debt is irrevocably and unconditionally guaranteed by GOC. CHT issues Canada Mortgage Bonds ("CMBs"), which are guaranteed by CMHC, and therefore by GOC.

Effectively, seven entities - the six Borrowers and GOC - each borrow billions of dollars annually, all based on the Government's AAA credit rating (the "Canada Credit"). While spreads have continued to decline in recent years, the Borrowers still pay an interest premium over Canada Bonds with a similar structure.

The factors outlined in the preceding two paragraphs, plus the fact that no study of the framework had been undertaken in over a decade, led to a desire to undertake this Review, which would look at the current borrowing framework, and would consider alternatives.

In this Review, KPMG is providing preliminary independent analysis of the status quo and the various alternatives, to provide information to allow the reader to assess: "What is optimal for Canada?" An optimal framework for Canada would be a framework that optimizes various, sometimes competing, objectives such as minimizing borrowing cost, managing risk, supporting Borrower mandates, and ensuring good governance. Accordingly, KPMG is not examining what is optimal for each Borrower, nor what is optimal for GOC, Finance or BOC. Instead, KPMG is looking at the situation from an arm's length perspective; effectively through the eyes of the citizens and taxpayers of Canada.

## Part A Current Borrowing Framework

### III  Governance Practices[1]

Early in 2005, the Auditor General and Treasury Board each issued independent reports on Crown Corporation governance. The Treasury Board report emphasized: "Good governance requires transparency and accountability. Together, transparency and accountability build trust". These factors are therefore the critical building blocks for leading governance practices.

#### Borrower Governance Practices

The Borrowers have broad governance responsibilities to Finance and Treasury Board, but are primarily responsible to their own boards and Ministers. The Crowns, as per the Financial Administration Act ("FAA"), must submit annual Corporate Plans to Treasury Board and Borrowing Plans to Finance for approval. CWB has similar responsibilities under the Canadian Wheat Board Act. CHT is not subject to the FAA, and therefore its borrowing activities are covered under CMHC's Corporate Plan as a program activity. Other than this annual approval, there is little formal interaction between Finance and the Borrowers. Effectively, Finance operates as a gatekeeper of the Borrowers' borrowing programs, and not as a regulator. However, the Minister of Finance issues guidelines on treasury risk management which the Borrowers must follow.

The Borrowers have internal governance structures in place. All have implemented treasury risk management policies, as highlighted in Appendix B. Furthermore, the Borrowers have segregated duties between borrowing, accounting and risk management, have oversight provided by asset-liability management committees and have their Borrowing Plans and Corporate Plans approved by their boards. All are subject to audits and special examinations from time-to-time.

Comments about potential improvements to the current governance regime and borrowing framework are discussed in the next section.

### IV Enhanced Status Quo[2]

As part of KPMG's review of the performance and governance of the current borrowing framework, we were asked to identify suggested improvements which could enhance the status quo. To prepare these improvements, we considered leading practices, and also considered issues that became evident through our interviews and discussions.

While the current governance framework is not inherently flawed, there are things that could be done differently at Finance and at the Borrowers in order to improve the status quo. These changes would improve governance and management of the Canada Credit through an improved collective understanding of borrowing amounts, cost of funds, and risk. This information could then be used to better align Borrower and GOC activity, to improve consistency of pricing and to better manage risk. Some of these improvements could be made easily and immediately, others will first require further investigation. Many can be made now, even if further study on potential centralization is investigated, others may be inappropriate should a centralization be imminent.

Issue Observations Suggested Improvements Borrowing Plan . The Borrowing Plans are not uniform, and cannot readily be combined to determine the expected consolidated five-year borrowing requirements of GOC and the Borrowers. .  Commentary accompanying the Borrowing Plans is inconsistent. .  CHT's borrowing, which is currently over 50% of all Borrower term requirements, is part of CMHC's Corporate Plan as a program activity, but is not included in CMHC's Borrowing Plan. .  The Borrowing Plans contain limits on spreads and borrowing amounts which are significantly in excess of what the Borrowers require, total use of which would be inappropriate for a Borrower using the Canada Credit. .  A template format for Borrowing Plans should be established to facilitate consolidation. .  Guidelines for minimum commentary on Borrowing Plans should be established. . Due to the materiality of CHT's borrowings, CHT's projected borrowing should be included either: (1) as a formal element of CMHC's Borrowing Plan; or (2) as a CHT Borrowing Plan submitted to Finance for approval. .  The approval in the annual Borrowing Plans should include reduced but adequate and realistic spread and issuance levels. Reporting of Actual Results .  The Borrowers are inconsistent at reporting actual results, and in some cases, the Borrowers do not track achieved results from money market activity. .  The Borrowers do not consistently report their differential between the achieved cost of funds and the relevant benchmarks. .  Guidelines should be established for reporting of actual results, and for how to measure performance versus benchmarks. .  Such reporting should be submitted at least quarterly. Financial Risk Management . The Borrowers submit quarterly counterparty credit risk data, but are seldom given feedback from this report. .  The Borrowers have very different policies for management of market and financial risks. . The Borrowers do not consistently use Credit Support Annexes to the ISDA. .  Additional and standardized risk information should be collected, and the Borrowers provided with appropriate feedback. .  Guidelines and policies for financial risk management should be reviewed. .  With respect to credit risk, set guidelines as to counterparty credit limits and to clarify that all counterparties must be subject to a Credit Support Annex with collateral being called at fixed levels, consistent with GOC's practice. Asset-Liability Management Committees .  The format, depth and content of the ALCO reports varies widely among the Borrowers, and in one case, contains very little information about actual borrowings. . ALCO reports are not consistently provided to Finance. . Borrowers should ensure that ALCO reports fully discuss actual and planned borrowing. .  All ALCO reports should be distributed to Finance at least quarterly. Operating Cost .  Little commonality exists among the technology and infrastructure platforms used by the Borrowers. .  A review of Borrowers' infrastructures should be conducted with the objective of identifying benefits that could derive from some standardization and scale for negotiation with vendors. Board Governance . The Boards of the Borrowers do not appear to consistently have all of the required skills to independently evaluate their treasury operations. .  Each Borrower should require that at least one non-executive board member have skills in treasury, capital markets and financial risk management. Coordination of Borrowing .  There is no formal coordination of timing and pricing of borrowing among the Borrowers and with Finance. .  The Borrowers and Finance should improve coordination in order to achieve various benefits (including avoidance of competing significant debt issues).

The following potential enhancements are extracted from more detailed analysis provided in Sections VI and VII. Summaries of the observations and suggested improvements are reproduced here:

 Issue Observations Suggested Improvements Agency Spread .  The Borrowers pay an agency spread versus GOC, which is illogical for the Crowns as they are legally indistinct from GOC. .  If practical, Finance should work with the Borrowers to reduce, if possible, the agency spread. Borrower Liquidity .  Certain of the Borrowers maintain high levels of liquidity, and in certain cases, use such liquidity profits to support their mandates. .   The Borrowers typically arrange standby bank lines of credit for emergency liquidity requirements .  Liquidity policies and guidelines for Borrowers should be reviewed. .  The possibility of using GOC for an emergency line of credit should be considered. Borrowing Products .  Several of the Borrowers satisfy a significant portion of their term financing needs through structured notes, while GOC does not. .  Since the Borrowers share the Canada Credit, guidelines should be established for the issuance of structured notes.

## Part B Alternative Borrowing Frameworks

### V  International and Provincial Borrowing Frameworks[3]

In debt markets, issuers that are closely affiliated with the sovereign borrower, but who issue debt in their own name, are referred to as "agencies". Agency debt may or may not be guaranteed by the sovereign, but in either case, such agencies typically share the benefit of the sovereign credit ratings, and are either critical to government policy or are well-integrated with the government's operations.

In Appendix E, further detail is provided on various international and provincial government and agency borrowers.

As would be expected, there is a wide range of alternatives as to how sovereigns and their agencies borrow. The United States market is unique, due to its enormous size and due to the many government-sponsored entities ("GSEs"). Some of the mandates of Canada's Borrowers are similar to those of U.S. government departments, while other Borrower mandates are similar to those of some GSEs. Many significant GSEs do not have explicit government support, and in fact, have non-government shareholders - the most notable examples being Fannie Mae and Freddie Mac. There are five regional Farm Credit Banks ("FCBs") and twelve regional Federal Home Loan Banks ("FHLBs"). To achieve economies of scale and to reduce the regional risks, the FCBs have a central funder that borrows on behalf of all five FCBs. The FHLB system is slightly different in that each borrows independently, but they report and are rated as one entity.

Many sovereigns, including United Kingdom and New Zealand, have only one market-facing borrower. Other sovereigns, such as Germany and Japan, have several borrowers issuing debt. However, few countries of the size of Canada have as many borrowers in the marketplace - for example, the mandate of Germany's KfW encompasses that of several of the Borrowers. There has been a gradual trend to a single government issuer, with the UK, Ireland, Australia, and Portugal moving to the concept of centralized borrowing in the 1990s.

Similarly, the Canadian provinces have taken steps in recent years to reduce the number of market-facing borrowers and achieve better economies of scale. One example is the Ontario Strategic Infrastructure Financing Authority ("OSIFA"), which has issued debt on behalf of 170 municipality clients, and which is available for use by all 445 Ontario municipalities, and as of May 2005, for Ontario's universities. Since it is voluntary, OSIFA provides particular benefit to the smaller municipalities who otherwise would likely borrow at much higher rates.

Consolidation of borrowing entities seems to be taking place for one or more of four main reasons: (1) economies of scale (more borrowing, lower cost, attract and retain more qualified people, more market intelligence); (2) risk diversification (due to many agencies having mandates that often result in them having less diversification, lower mark-ups and lower quality portfolios than their for-profit counterparts); (3) political separation (where borrowing is being handled by a "debt management office" relatively isolated from the sovereign's political and monetary policy considerations); or (4) better performance management capability and monitoring.

### VI Attributes - Direct Borrowing Costs

Interest costs (referred to as direct borrowing costs) are an attribute that must be considered under various alternative borrowing frameworks. This attribute is identified in a separate Section not because it is more critical, but instead because of the level of financial analysis required in order to appropriately assess direct borrowing costs for various Borrowers and programs. Note that this section does not consider any other financial impact of a change to the status quo, such as transition costs or any change in ongoing costs.

One of the fundamental objectives of GOC's debt management strategy is to raise stable, low-cost funding for the Government. Any issuer benefiting from the Canada Credit has a responsibility to achieve the lowest possible risk-weighted cost of borrowing.

#### GOC and Borrower Debt - Today and Tomorrow[4]

GOC and the Borrowers collectively have about $545 billion in outstanding market debt, of which$440 billion (81%) has been issued by GOC, and $105 billion (19%) has been issued by the Borrowers. GOC expects its market debt to decline somewhat, while overall, Borrower market debt is expected to increase, implying that the percentage of outstanding debt issued by the Borrowers will increase. Furthermore, it should be noted that CHT represents just over 50% of current Borrower debt, and therefore Borrower debt would only be 10% of the total with CHT excluded. GOC's current plan is to increase its issuance of stock of treasury bills and other short-term instruments to$140 billion by March 2006, and potentially higher thereafter. The Borrowers' collective money market needs for the next five years will average about $17 billion. Accordingly, the Borrowers will collectively be issuing money market instruments at no more than 11% of GOC's levels each year. GOC expects that annual Canada Bond issuance, after taking into account maturities and bond buy-backs, will be between$30 and $35 billion annually. The Borrowers will issue an annual average of$25 billion (about 75% of GOC's borrowing) in term debt over the next five years, or $11 billion (33% of GOC's borrowing) excluding CHT's$14 billion issuance.

A table of planned annual debt issuance by the Borrowers follows:

2005 to 2009 Average Borrowing Requirement

 With CHT Without CHT CAD Millions Money Market Money Market CAD 12,300 12,300 Money Market USD 4,600 4,600 Subtotal 16,900 16,900 Term Term Floating CAD 4,300 2,100 Term Floating USD 6,400 6,400 Term Fixed CAD 14,700 2,900 Subtotal 25,400 11,400 Total Average Borrowing Requirements 42,300 28,300

#### The Agency Spread and the Power of a Basis Point[5]

Technically, the Crowns' debt obligations are legally indistinguishable from GOC. Despite the Borrowers' extensive explanations, most capital market participants consider all of the Borrowers as agencies. Worldwide markets assign an interest rate spread premium to agencies, even where the sovereign guarantees the indebtedness of these agencies. As a result, although the Borrowers also share the Canada Credit and Canada's AAA credit rating, the market expects and achieves a higher rate of return on like debt issued by the Borrowers versus that issued by GOC. This spread difference can be as low as 1 basis point for the Canadian money market to 15 basis points for conventional ("vanilla") term funding.

To illustrate the power of a basis point (i.e. 1/100th of 1%), a one basis point interest rate reduction in the Borrowers' $42 billion in annual debt issuance would result in reduced borrowing costs of$4.2 million in the first year, $38 million net present value over 5 years, and$175 million net present value over 25 years.

Many Borrowers utilize various structured products (described further in Section VII) to lower their weighted average cost of funds.

The Borrowers actively seek to minimize this spread every day. With the cooperation of GOC, various steps could be taken to reduce the agency spread being paid by the Borrowers. These measures could include items such as changing government collateral rules, clarifying default procedures, lobbying indices to remove distinctions between GOC and the Borrowers, and/or jointly marketing "Borrower = Canada" to potential investors.

Nonetheless, while these measures may result in a nominal reduction in the agency spread and should be investigated further, market participants indicated that no significant sustainable reduction in the spread could be achieved as long as the Borrowers issued debt in their own names.

#### Analysis of Change to Borrowing Cost under Centralization

Appendix H presents a detailed analysis of the direct borrowing costs under centralization with a Centralized Funding Entity ("CFE") versus the status quo. This analysis contains numerous assumptions, and readers are encouraged to review them carefully to ensure that the methodology and limitations of this financial analysis are clearly understood. Transfer pricing has been ignored as it is irrelevant to Canada, however, it could reduce the amount of savings in direct borrowing costs experienced by the Borrowers versus that calculated herein.

The "savings" presented herein represents a reduction in direct borrowing costs paid to the market, which may be passed on to stakeholders.

Note that Appendix H includes only the data from the Borrowers, and has not considered any other debt guaranteed by GOC or a Borrower, such as Canada Savings Bonds or the NHA Mortgage-Backed Securities guaranteed by CMHC.

The Borrowers meet their borrowing needs using a variety of products issued in a variety of Canadian and international markets. Accordingly, they achieve widely divergent overall cost of funds, particularly for each of the "Term" categories. Therefore, the impact of any centralization on direct borrowing costs for each particular Borrower varies markedly. The figures presented herein are blended on a weighted average basis for all Borrowers who issue debt in a particular category.

Based on the assumptions outlined in the appendix, we estimate that by centralizing all Borrower debt issuance, direct borrowing costs would be reduced as follows

 Reduced Borrowing Costs (First Year - CAD millions) Reduced Borrowing Costs Reduced Borrowing Costs With CHT Without CHT With CHT Without CHT (Five Years - CAD millions) (basis points) Money Market CAD $0 to$5 $0 to$5 $0 to$25 $0 to$25 0 to 4 0 to 4 Money Market USD $3 to$6 $3 to$6 $15 to$29 $15 to$29 7 to 12 7 to 12 Term Floating CAD $1 to$4 ($3) to ($1) $13 to$64 ($41) to ($15) 2 to 10 (13) to (5) Term Floating USD $1 to$5 $1 to$5 $19 to$80 $19 to$80 2 to 8 2 to 8 Term Fixed CAD $19 to$28 $3 to$4 $289 to$421 $40 to$56 13 to 19 9 to 13 Total $24 to$48 $4 to$19 $336 to$619 $33 to$175 6 to 11 1 to 7

Important notes regarding the figures on the "Without CHT" columns of the table follow:

• For Term Floating CAD, the Borrowers meet virtually all of their needs through issuance of structured products. Direct borrowing costs for this category could increase under centralization by 5 to 13 basis points because it is assumed that the CFE would not issue structure notes.

• For Term Floating USD, posted market swap data was used to swap Canadian borrowing to USD under centralization. It is probable that the CFE could achieve better than posted rates, which would increase the savings in this category.

• For Term Fixed CAD, the range of 9 to 13 basis points also includes some structured products. If the Borrowers solely issued vanilla fixed rate products, this range would be at least 5 basis points higher.

Under the "With CHT" columns, the only figures that differ are the Term Floating CAD and Term Fixed CAD categories. In both cases, high volumes of vanilla product issuance substantially increase the average basis point reduction in direct borrowing costs available with a CFE.

#### Summary

Excluding CHT, which is a guaranteed market housing support operation and not a direct funding cost to Canada, direct borrowing costs paid to the market under centralization could be reduced by approximately 3.5 basis points, representing approximately $10 million in the first year, and$33 to $175 million over five years. If CHT is included, these figures increase substantially to approximately 8 basis points, representing approximately$33 million of savings in direct borrowing costs in the first year, and $330 to$620 million of savings over five years. In summary, under centralization, Canada could expect to experience reductions in borrowing costs paid to the market. These savings could be redeployed to other Government priorities or to support the Borrower mandates.

It should be noted that this analysis excludes any transition costs or changes in annual operating costs under centralization. Furthermore, this analysis does not assign a risk premium to the structured products issued by the Borrowers, which inherently carry more risk than the vanilla products assumed to be issued by the CFE. If the Borrowers did not issue structured products, these savings figures would be much larger.

### VII  Attributes - Other[6]

In determining the feasibility and acceptability of any borrowing framework, there are various factors, referred to herein as "attributes", that must be considered in evaluating such a framework. Nine key attributes have been identified, as follows:

• Direct Borrowing Cost

• Management of the Canada Credit

• Diversity and Appropriateness of Borrowing Products

• Quality and Liquidity of Canadian Capital Markets

• Financial Reporting

• Borrower Liquidity

• Borrower Operations

• Financial Risk Management

• Governance and Accountability

Direct Borrowing Cost was covered in the previous Section. In this Section, we consider the reasons that the remaining attributes are relevant and consider issues relating to such attribute.

#### Management of the Canada Credit

This attribute is relevant as all Borrowers share the Canada Credit, and actions taken by any party sharing this rating could impact all of the other parties and their borrowing costs. As a sovereign, it is critical that the Canada Credit has a strong market and public perception. All of the Borrowers are aware of this shared rating and even share the rating agency costs with GOC. The Borrowers' continuously make efforts to maintain their own high reputations, but there is no formal coordination of the timing of debt issuance by the various Borrowers and GOC.

Leading practices for sovereign borrowers in managing their credit ratings involve high levels of transparency and fairness, typically involving published debt strategies, debt issuance via auction, and predictability as to timing and amounts of debt issuance.

The Borrowers, like many agencies around the world, are more opportunistic in their borrowing practices when compared to Canada. With the exception of CHT, they do not normally issue benchmark securities; instead they will issue debt when market conditions are favourable. Furthermore, most Borrowers issue structured notes, which are commonly issued by agencies but tend to be avoided by sovereigns (see next subsection).

While presenting only one credit to the capital markets would be more transparent and easier to understand for market participants, multiple borrowers can be an advantage in a time of high government need (e.g. early 1990's) due to their ability to issue different products and access different markets.

While GOC and the Borrowers use very different strategies for managing the Canada Credit, and subject to the suggested improvements described in Section IV, the Canada Credit can be well-managed under either the status quo or under a centralization.

#### Diversity and Appropriateness of Borrowing Products

This attribute is relevant as GOC and the Borrowers need to have sufficiently diverse sources of financing in order to ensure that they can cost-effectively meet their financing needs as market conditions change over time. However, the debt products issued by GOC and the Borrowers must be appropriate to maintain the high quality reputation of the Canada Credit.

The vast majority of GOC's borrowing needs are met through the issuance of Canada treasury bills and benchmark fixed-rate Canada bonds (at the 2-year, 5-year, 10-year and 30-year points on the yield curve).

The Borrowers raise their funds in very different ways. In the money market, they post daily commercial paper rates, raising about 75% of their needs in Canada, and meeting the rest of their needs in the United States and Europe. They tend to issue at maturity dates that differ from Canada treasury bills to get the most competitive rate.

In term markets, CHT has met its financing requirements to date through quarterly issuances of five-year fixed and floating rate debt. About 75% of the remaining $11 billion in annual term needs of the Borrowers are in floating rates due to the preference of their clients for floating rate loans. To meet this demand for floating rates, the Borrowers (other than CMHC) tend to issue fairly complex structured notes (also known as medium term notes or "MTNs", see examples in Appendix J), which are then swapped into floating rates at attractive financing rates. These structured notes are typically principal-guaranteed and option-embedded, with the return to investors dependent upon the performance of an index, currency or interest-rate benchmark. The structured note market is mature with notes commonly issued by AAA-rated agencies and then brokered and sold to institutions around the world. These structured MTNs allow the Borrowers to achieve attractive financing costs and are arranged on the basis that the issuer bears no direct risk relating to the structure of the MTN. The derivative transactions required to provide this hedging create varying degrees of counterparty exposure according to complexity. The Borrowers take risks into account in evaluating potential structured notes. However, the Borrowers do not calculate or report on risk-weighted cost of capital. If risk-weighted capital was considered, the cost of funds on structured notes would not appear as attractive. Some Borrowers have also been issuers of retail principal-protected structured note issues in the Canadian market incorporating instruments similar to those offered by hedge funds. The role of the Borrowers is to provide investors with confidence that the issues which guarantee the return of principal involve no repayment risk. The Borrowers are compensated with cost-effective financing for providing the underlying credit upon which the various structures are attached. The Borrowers' names figure prominently in the marketing of these issues, and the appropriateness of their role in transactions which may generate no return for Canadian retail investors should be reviewed. Unlike the Borrowers and other worldwide agencies, sovereign borrowers tend to find structured products unsuitable, due to their complexity; lack of transparency and understandability, and the reputational risk should such products under-perform. Furthermore, managing structured products adds complexity to debt management strategies, and the larger sovereign borrowing needs would often require hundreds of trades to make an impact. It is worth considering further the appropriateness of the Borrowers issuing such structured products when they share the Canada Credit. Under the status quo, guidelines should be established as to what types of structured notes may be appropriate, how to report the risk-adjusted cost of capital, and the minimum risk management strategies associated with permitted issuances. Overall, with centralization, Canada should be able to achieve cost savings without the need to issue these structured products. #### Quality and Liquidity of Canadian Capital Markets This attribute is relevant as Canada treasury bills and bonds are used as the benchmark against which all other borrowers in the Canadian marketplace are measured. GOC must therefore establish benchmark issues, and continually take steps to maintain a viable and liquid yield curve. GOC has reduced annual issuance of Canada Bonds from$54 billion in fiscal 1997 to a plan of $33 billion in fiscal 2006. During that time, GOC has dropped its 3-year benchmark issuance, reduced the number of auctions for the 30-year benchmark, and has instituted an aggressive buy-back program for non-benchmark bonds in order to achieve sufficient issuance of new benchmarks. This buy-back program has matured such that it can only be continued with gradually reducing volumes. The Debt Management Strategy 2005-2006 indicates that ".one of the key challenges for the Government in recent years has been to maintain a liquid, well-functioning Government securities market in the face of declining borrowing requirements and reduced issuance." The majority of market participants that were interviewed indicated that the liquidity of the Canadian bond market is generally sufficient but not optimal today, expressing concern about future market liquidity in an era of increasing investible assets and declining Government borrowing. Those who transacted at the long end of the yield curve tended to feel that liquidity is currently lacking at that point of the curve. Market participants indicated that demand for Canadian and international bonds is at an all-time high and is growing. This is partially evidenced by the strong acceptance of CHT's CMB program, which issued more 5-year bonds than GOC did in 2004. Most market participants felt that a centralization, which would add at least$11 billion in annual issuance (and $25 billion including CHT), could aid market liquidity and could be readily absorbed by the market. The$11 billion alone would allow a 33% increase in annual bond issuance (or a reduction in the buy-back program), which could significantly delay any future liquidity problems. While there is a risk that additional supply could impact the cost of capital, most market participants felt that a price impact was unlikely.

One important consideration is that the Borrowers' needs may not align sufficiently well with the Government's benchmark issuances at maturities of 2, 5, 10 and 30 years. However, there is no reason that a CFE could not use the additional benchmark funds raised to provide the Borrowers with funding at different points of the yield curve. As the Government does not need to match assets and liabilities in the same manner as the Borrowers, this should not be a significant constraint. However, centralization would require the Government to modify its Debt Management Strategy to account for the additional benchmark issuance and for the provision of funding to the Borrowers.

Other enhancements to liquidity outside of centralization should be considered. The GOC has a sizeable amount of non-marketable debt but will not likely bring this debt to market for several years. Furthermore, the Montreal futures market has not grown sufficiently to allow for alternate hedging tools. Therefore, no other obvious source of liquidity appears imminent in the medium-term.

Overall, centralization would allow GOC to address the strong market demand for additional supply of benchmark Canada Bonds.

#### Financial Reporting

This attribute is relevant as a change to the status quo could change the amount of debt reported in the "Public Accounts of Canada" - the GOC's financial statements - which could therefore change figures such as total marketable debt, debt to GDP ratio and interest expense as a percentage of spending.

Currently, all Borrowers except CHT are consolidated in the Public Accounts on a modified equity basis, such that none of their debt is consolidated in the Public Accounts.

Under current Public Sector Accounting Board ("PSAB") guidelines for financial reporting, it is probable that most debt issued under Centralization would be consolidated in the Public Accounts, thereby increasing the reported debt obligations of the Government. Since this guideline would be applied on a prospective and not retroactive basis, this increase could slow or reverse the improvement in the figures which has been reported in recent years. After one year, this could increase the gross marketable debt by up to 6%, and after a few years could increase the gross marketable debt up to 11%. If CHT was included, this increase would be substantially higher, and with current CHT issued debt, the increase would be 23%. Since the assets would also be consolidated, there would be no change in net debt.

It should be noted that centralization of borrowing does not mean that more debt is issued, and the obligations under the Canada Credit would not be increased versus the status quo. However, debt that is not currently reported as the Government's direct debt would become Government direct debt. On an economic basis, nothing would change.

A rating agency indicated that it primarily considers the amount of debt that is being serviced from taxpayer revenues. Since debt raised for Borrower purposes would be serviced from Borrower revenues, and since borrowings under the Canada Credit would not change, the rating agency indicated that it would not likely look at the Canada Credit any differently.

Nonetheless, centralization would require a communication strategy to explain the optics of possible reported increases in GOC's gross debt.

#### Borrower Liquidity

This attribute is relevant as the ability of Borrowers to manage their day-to-day short-term liquidity requirements is critical to the fulfillment of their mandates.

Centralized liquidity management could improve GOC's overall short-term cash management strategies and simplify market dealings. The Borrowers may also be able to use the CFE as their emergency line of credit, thereby allowing them to eliminate standby bank lines and save money.

While each Borrower has a liquidity policy in place, certain Borrowers are maintaining high levels of liquidity which is arbitraged to generate treasury profits, which in turn support their mandates. The appropriateness of such a strategy should be reviewed.

Assuming a well-functioning CFE, the amount of Borrower liquidity, no matter what the Borrower's policy is, does not need to change in any centralization. Day-to-day cash flow management should be able to be readily managed under either alternative. Should the CFE be less flexible or less functional, then the Borrowers would either need to be less responsive to their clients, or to arrange for additional liquidity, warehousing and hedging to compensate for that increased uncertainty.

#### Borrower Operations

This attribute is relevant as the Borrower's operations are critical to their mandate, and the Treasury Department is typically a critical piece of such operations. CHT has no operations per se, and its treasury requirements are managed by CMHC's treasury department. Accordingly this attribute is only relevant to CHT insofar as it is managed by CMHC.

The Borrowers are extremely concerned that centralization would negatively impact their mandates in three major areas, i.e. the loss of: (1) flexibility; (2) market intelligence; and (3) autonomy.

Regarding flexibility, the Borrowers feel that with funding handled by a CFE, treasury capabilities in the treasury department will become fragmented and inefficient. They also believe that the CFE will not be sufficiently responsive, thereby reducing the flexibility of the treasury department. However, Borrower treasury departments would maintain responsibility for cash and liquidity management, asset-liability management, hedging, pricing financial products, etc.

Regarding market intelligence, the Borrowers value the ability of such intelligence to help the lending and asset side of their businesses. Some Borrowers commented that having daily contacts with the capital markets - sometimes better than those enjoyed by the Bank of Canada as Borrowers are commission-paying clients - is a fundamental building block of the asset side of their business. It helps them to support the lending side of their business with economic reviews and opinions, to contribute to the new product and services development process, and to better position themselves versus the competition. Without borrowing activity, Borrowers are also concerned that pricing for derivatives hedging will not be as attractive.

Regarding autonomy, the key issue is that the Borrowers are well-established as relatively stand-alone entities with their own boards. They value treasury and capital markets operations as a core activity. Each views centralization as a forced outsourcing of a core activity, which reduces the ability to act as a relatively-independent stand-alone entity. They view centralization as splitting accountability and therefore increasing risk. Accordingly, they believe that centralization would significantly impair their ability to support their respective mandates.

We believe that all of these concerns are legitimate. Nonetheless, we believe that many of these issues can be mitigated with a well-planned CFE, with a tight service level agreement. Firstly, even under a centralized borrowing framework as envisioned herein, each Borrower will need to retain its Treasury Department expertise and practitioners to perform ALM and effect derivatives hedges needed for asset liability management purposes. Secondly, the high caliber capital markets expertise which already exists would likely be supplemented by additional resources at the CFE. If communications are set up properly, this knowledge pool could collectively benefit all Borrowers and GOC. Thirdly, some market intelligence gained is of limited value - international market intelligence gained by CMHC, FCC and BDC through borrowing on international markets would give very limited insight into their core Canadian businesses.

Due to the breadth and depth of the Borrowers' valid concerns, if centralization was desired, then more planning would need to be done on how centralization would be accomplished. Once that was completed, a better understanding of the ability to mitigate the concerns could be gained.

#### Financial Risk Management

This attribute is relevant as leading practices indicate that sovereigns and agencies should practice sound financial risk management, and have appropriate financial risk management practices in place. Such practices ensure that risks and results are communicated regularly to senior management and boards. In turn, such boards must have the ability to provide effective oversight in order to exercise their fiduciary obligations.

Currently each individual Borrower has appropriate financial risk management practices in place. Centralization would not change the need to maintain such practices, but would ease the capability to aggregate reporting and understanding of financial risks. However, this aggregation could also be accomplished through improvements to the status quo.

#### Governance and Accountability

This attribute is relevant as Finance is ultimately responsible for, among other things, the Canada Credit, GOC borrowing and investing, the smooth functioning of the Canadian debt capital markets and prudent debt management. The Borrowers are using the Canada Credit, and therefore must be partially accountable to the Ministry of Finance. On the other hand, the Borrowers operate relatively autonomously, and must be fully accountable to their own boards and Ministers.

No legal or policy constraint prevents the Crowns and CWB from sourcing their borrowing from GOC. If Centralization is pursued, further investigation will need to be undertaken regarding CHT as Finance may need to seek specific approval to borrow for CHT's specific, designated purpose. For CHT, modifications may also be required to CHT's trust instruments to allow a different funding source.

We believe that these factors can continue to be balanced under either alternative.

### VIII  Alternatives[7]

Various alternatives were considered throughout the course of this Review. Many of these alternatives proved to be inadvisable as they were explored through interviews and analysis, including the following:

• Fungibility - Under this alternative, the Borrowers would issue debt that would be "fungible", i.e. legally the same as, and interchangeable with, GOC debt. The Canadian Depository for Securities indicated that there would be significant logistical hurdles under its current tracking and accounting system. Even if those hurdles could be overcome, an extremely problematic market issue arises. Market participants reacted very unfavourably to the concept where a predictable auction schedule for issues of benchmark Canada Bonds could be supplemented opportunistically any time through an unpredictable commission-based Borrower issuance of a fungible bond. They expected that such a structure would create market confusion, damage the dealers' ability to take short or long positions and hurt Canada's credibility, all contributing to a decrease in market liquidity of the benchmarks.

• Create New Centralized Funding Entity ("CFE") for the Borrowers - Under this alternative, similar to the U.S. FCBs, a CFE would borrow on behalf of all of the Borrowers. This CFE could be a new Crown Corporation, or a special purpose borrowing and financing entity. This alternative would not be advisable, as it would involve significant transition and ongoing infrastructure costs, and yet would still be subject to the agency spread.

• One Borrower Provides Funding to All Other Borrowers - This is very similar to the previous point, except that it avoids some of the infrastructure and transition costs of a new entity. This alternative would not be advisable for three reasons: (1) the agency spread would not disappear; (2) the Borrowers have little interest in raising funds for other Borrowers, believing it would dilute their mandate and increase their risk; and (3) the Borrowers have little interest in having another Borrower raise funds on their behalf due to concerns regarding conflict from the need to prioritize between the Borrower's own needs and that of a third party, particularly in dynamic capital market conditions.

After rejecting the preceding alternatives, only two feasible alternatives remain. The first is an "Enhanced Status Quo", with certain changes made in order to optimize borrowing and governance. The second is "Centralization", where one entity borrows funds on behalf of both GOC and the Borrowers.

Under Centralization, while certain risk management and swap activities could be centralized, the concept of centralized asset-liability management ("ALM") was rejected, due to recognition that good knowledge of the asset side of their businesses is a key success factor for effective asset-liability management by the Borrowers. Furthermore, maintaining ALM would allow Borrowers to remain strongly accountable and independent.

Note that Centralization can have many variants, including: (1) centralization at the Bank of Canada or centralization in a new "debt management office"; (2) inclusion or exclusion of CWB and CHT as non-Crown Borrowers; (3) making borrowing from GOC optional where the Borrowers could use both GOC and their existing networks as a source of financing; or (4) partial centralization, where some debt products were centralized and others not, e.g. money market versus term, vanilla products versus structured products and Canadian currency versus other currencies.

Dealers presented mixed views on Centralization, although most investors and foreign dealers favoured Centralization due to the implied additional issuance and liquidity of Canada Bonds; and the feeling that the agency spread paid by the Borrowers was not in Canada's best interests.

As stated earlier, the Borrowers are extremely concerned about Centralization, expecting a loss of flexibility, market intelligence and autonomy. By the same token, all Borrowers are in favour of reducing their cost of funds in furtherance of their core mandates.

Should Finance wish to explore Centralization further, all of these issues, plus one-time transition costs and changes to ongoing administrative costs, would need to be considered in greater detail.

### IX   Arguments for Centralization and Arguments for Enhanced Status Quo

In this table, the results of the Review are summarized. With only two alternatives, the arguments in favour of one alternative tend to be the arguments against the other alternative. Therefore, this table considers the arguments for each alternative versus the nine attributes identified earlier.

   Arguments in Favour of Centralization Arguments in Favour of Enhanced Status Quo Borrowing Cost .  Due to the agency spread, GOC will always pay a lower cost of funds for a similar product .   Including CHT, reduced costs paid to the market of approximately 8 basis points, representing $33 million in the first year, and$330 to $620 million over 5 years. . Excluding CHT, reduced costs paid to the market of approximately 3.5 basis points, representing savings of approximately$10 million in the first year, and $33 million to$175 million over 5 years .  Opportunistic borrowing strategy and use of structured products by Borrowers can minimize cost differential .   The change to centralization would involve one-time transition costs and an uncertain change to annual operating costs which would be passed onto Borrowers .  Potential reduction in agency spread through campaign Management of the Canada Credit .   Simplicity of one issuer for Canada in debt financial markets .  Enhanced transparency .   Reduced risk of any Borrower inadvertently damaging the Canada Credit for all .   Borrowers are highly regarded in capital markets .  Multiple borrowers can be an advantage in a time of high government need (e.g., early 1990's) as they can access foreign currency and other markets Diversity and Appropriateness of Borrowing Products .   CFE can achieve lower borrowing costs without incurring the additional risk and complexity of structured products .   Issuance of structured products decreases transparency and understandability .  On a risk-adjusted basis, the yield of structured notes is not as attractive as stated .  Structured products allow Borrowers to achieve attractive funding costs and overcome the agency spread .   Issuance of structured products fills a market need .  Issuance of structured products enhances diversity of funding sources Quality and Liquidity of Canadian Capital Markets .   Market participants are concerned about current and future liquidity of Canadian bond markets, and indicated that significant increases in benchmark Canada Bond issues could be absorbed at no additional cost .   Market participants believe that CHT's current $14 billion in 5-year issuance could be readily redeployed to GOC's 5-year benchmark issuance, resulting in substantially reduced borrowing costs . The redeployment of$11 billion in other Borrower term funding, primarily from the structured market, would be a welcome addition to benchmark issues, increasing supply by about 33% .  Investors believe that elimination of the agency spread is in Canada's best interests .  Borrowers' needs may not align with the terms of GOC's benchmarks at 2, 5, 10 and 30 years .  It is possible that liquidity could be enhanced through other methods, such as converting non-marketable debt, or promoting futures market development Financial Reporting .  Increases transparency, by creating one consolidated debt position for Canada .   Facilitates comparisons with other developed sovereigns where borrowings are centralized .  Reduced debt complexity simplifies financial reporting .  Avoids complexity of accounting for transactions between GOC and the Borrowers .  Unlike today, centralization would probably lead to Borrower debt being included in the Public Accounts of Canada .  While no economic change, centralization would require a communication strategy Borrower Liquidity .  Simplifies borrowing, with only one borrower in money markets .  With proper planning and execution, CFE could meet the responsiveness requirements of the Borrowers .   CFE could replace the Borrowers' emergency standby bank lines of credit, and could improve GOC's overall liquidity management .  High degree of Borrower flexibility through full control over management of own liquidity .  Centralization may lead to Borrowers increasing their liquidity, warehousing and hedging activities .  CFE timeliness may be insufficient to allow Borrowers to maintain same level of client response .  Some Borrowers use high levels of liquidity to support mandate Borrower Operations .   Many of the arguments in favour of status quo could be mitigated with a well-planned CFE .  For all Borrowers except CWB and EDC, borrowing on international markets diverts focus from their core domestic lending business .  Borrower treasury departments are sophisticated and managed professionally to meet their mandates, and could become more inefficient under centralization .  Potential loss of market intelligence by Borrowers which is useful in managing the asset side of their businesses .  Potential loss of ability to negotiate cost-effective derivatives and hedging strategies Financial Risk Management .  Reduction of structured products in favour of vanilla products reduces required risk management .  Greater ability to create an aggregate understanding of financial risk .  Borrowers have financial risk management policies in place, and desirable improvements could be made without centralization .  Under centralization, if CFE borrows domestically to cover USD needs of Borrowers, it will increase required cross-currency swaps, and therefore counterparty credit risk Governance and Accountability .   Strong accountability maintained as CFE would be directly operationally accountable to its governing body, but would also be responsible for service levels to its clients, the Borrowers .   Better ability to build skills into "centre of excellence" with one market-facing borrower .  Inconsistent capability among Borrower boards to objectively evaluate treasury operations .  Borrowers have overall accountability to their own boards, and have accountability to the Minister of Finance regarding their Borrowing Plans .  Borrowers are subject to internal and external audits and to OAG special examinations .  Treasurers report to ALCOs .  Potential for overlap between responsibilities of the CFE and those of the Borrowers' Boards .  Autonomy of Borrowers as currently defined is maintained

Notes:

1  A more detailed discussion of these issues is attached as Appendix A. [Return]
2  A more detailed discussion of these issues is attached as Appendix C.[Return]
3  A more detailed discussion of these issues is attached as Appendix D.[Return]
4  A more detailed discussion of these issues is attached as Appendix F.[Return]
5  A more detailed discussion of these issues is attached as Appendix G.[Return]
6  A more detailed discussion of these issues is attached as Appendix I.[Return]
7  A more detailed discussion of these issues is attached as Appendix K.[Return]