Archived - Department of Finance

Evaluation of the Debt Auction Process Nominal Bonds, Real Return Bonds, Treasury Bills

Summary Report

OCTOBER 21, 2010

Submitted by

Marlene Puffer
Alexandra MacKay
David Goldreich

Twist Financial Corp.
www.Twistfinancial.com
416-460-3945

Table of Contents

0. Executive Summary
1. Background and Scope of the Evaluation
2. Methodology
3. Limitations of the Approaches
4. Overview of Governance for Debt Management Activities
5. Description of the Auction Process
6. Sovereign Comparison
7. Literature Review
8. Market Data Analysis
9. Review and Recommendations: Process
10. Risk Measurement and Management
11. Conclusions
Appendix A - Sovereign Comparison Summary
Appendix B - Literature Review Summary
Appendix C - Data and Market Analysis Summary
Appendix D - Primary Dealers’ Role
Appendix E - Endnotes
Appendix F - Management Response and Action Plan

PDF Version - [581 KB]

0. Executive Summary 

This report summarizes the results of the evaluation of the Government of Canada debt auction process for nominal bonds, Real Return Bonds (RRBs) and treasury bills (excluding cash management bills) that was performed by Twist Financial Corp from November 2008 through November 2009. The purpose of the evaluation is to assess the design and approach to managing the auction process, and the controls around it, including the determination of the terms of the call for tenders, the system for accepting bids, the release of results, and settlement.

The scope of the evaluation includes an assessment of whether the existing procedures and practices support the well-functioning of the auctions and whether they provide sufficient information, on a timely basis, to the participants and the market to meet the debt strategy objectives and principles. This study looked at governance of the current auction process in Canada, examined how the process works, and compared the Canadian approach to other sovereigns.

Our research has led us to make the following overall conclusions:

• The current structure works well, and supports the transparency, effectiveness and efficiency of debt auctions.
• Overall, the auction process has been successful in its immediate, intermediate, and ultimate goals of raising necessary funding at a low cost. Moreover, the auction process has helped sustain a liquid and efficient secondary market for Government of Canada debt.

While we suggest making several marginal changes to the auction process, we expect that the process will continue to achieve its objectives in the longer term in its current form or with the suggested changes in process that we cite.

We make four recommendations primarily focused on internal and external process. For example, we suggest specific ways to improve process and communication in the pre-auction phase between the Department of Finance and the Bank of Canada, as well as with the various players in the process (such as more detailed reports or new and improved channels of communication including more frequent conference calls and consultation with market participants). We also suggest ways to encourage more auction participation in certain market conditions by relaxing the upper limit on bids and counteracting the potential impact on short squeezes by reserving the right to re-open issues at any time. We recommend encouraging greater participation of customers already holding a bidder number, and identifying and actively soliciting potential new customers. We recommend motivating government securities distributor to strive to become primary dealers (who are subject to more stringent minimum bidding requirements) by attaching more visibility and prestige to the primary dealer status. Finally, we make recommendations about dissemination of auction results and staffing.

Summary of Recommendations

Recommendation 1: Improve communication and reporting between Department of Finance and the Bank of Canada

• For potential improvements in efficiency, consider changing the interaction between the Department of Finance and the Bank of Canada occurring ahead of bond auctions to a reporting function rather than an approval process.
• Replace authorization letters for debt auction operations with more informative, timely reporting and accountability to senior management.

Recommendation 2: Improve communications with market participants

• Increase the frequency of consultations with a larger sample of current and potential customers utilizing a mix of communication formats (including more frequent telephone conference calls and multilateral discussions in addition to the traditional bilateral visits with individual dealers).
• Obtain detailed feedback from a larger number of market participants, including large and small customers, prior to and following auctions of Real Return bonds and 30-year nominal bonds.
• Release auction results simultaneously on the Bank of Canada pages on Reuters and Bloomberg and through CanDeal and other relevant trading systems used by Canadian market participants.

Recommendation 3: Improve participation at auction

• Encourage all market participants to play a more active role in the design of the debt strategy, such as through an annual workshop in addition to the standard consultation process.
• Promote greater participation of customers already holding a bidder identification number, and identify and actively solicit potential new customers. Large customers should be offered direct access to the Communication, Auction and Reporting System (CARS).
• Relax the upper limit on bids and emphasize the government’s right to re-open a security at any time through something like the United Kingdom mini-tender facility, or similar to the Australian approach of re-opening at its discretion, which will both support participation and counteract the possibility of a participant attempting to exert undue influence on the price of a security.
• Encourage Government Securities Distributors to strive to become Primary Dealers (who are subject to more stringent minimum bidding requirement) by attaching more visibility and prestige to the primary dealer status (e.g., as in France).

Recommendation 4: Support the transfer of corporate knowledge

• Review and update annually detailed guidelines and process documentation at the senior management level at both the Bank of Canada and Department of Finance.

1. Background and Scope of the Evaluation 

This report summarizes the results of the evaluation of the Government of Canada debt auction process that was performed by Twist Financial Corp. Twist Financial was engaged by the Department of Finance to perform this evaluation. The intended audience and key stakeholders related to this report include the Department of Finance, the Bank of Canada, government securities distributors, investors and other capital market participants. The purpose of the evaluation is to assess the design and approach to managing the auction process, and the controls around it, including the determination of the terms of the call for tenders, the system for accepting bids, the release of results, and settlement. This third-party independent evaluation is intended to ensure that policies and practices in the domain of federal treasury operations are appropriate and achieving their objectives.

The Government of Canada’s primary debt strategy objective is to raise stable, low-cost funding. An associated objective is to sustain a liquid and efficient market for Government of Canada securities. A well-functioning auction framework and process is central to the achievement of these objectives.

The scope of the evaluation includes an assessment of whether the existing procedures and practices support the well-functioning of the auctions and whether they provide sufficient information, on a timely basis, to the participants and the market to meet the debt strategy objectives and principles. The scope of the evaluation does not include an evaluation of the auction format nor does it include an assessment of the IT system which embodies the processes and controls.

2. Methodology 

Time period and structure: The period of evaluation is from April 1, 2003, to March 31, 2008. The sovereign comparison focused on the current status of auction processes in each country as at August, 2009. Interviews were conducted with officials from the Bank of Canada and the Department of Finance in November and December 2008. The structure of the evaluation consists of sovereign comparison case studies, a literature review, data analysis, and interviews with market participants. This structure was determined by the Department of Finance and the Bank of Canada and was detailed in the publicly issued Request for Proposal. A list of evaluation questions corresponding to the criteria outlined in the request for proposal was developed by Twist Financial and approved by the Department of Finance and the Bank of Canada.

Sovereign Comparison: Case studies of Canada, Australia, the United Kingdom (UK), the United States (US), and France were conducted, and auction officials in each country (except France) were interviewed from June through September 2009. The sovereign comparison case studies focused on the current status of auction processes in each country as at August 2009. These G7 sovereigns were chosen in consultation with the Bank of Canada and the Department of Finance. The choices ensure inclusion of the largest and most liquid sovereign markets (US and UK) as well as a small less liquid market (Australia) and a representative market that lies between these extremes (France). To structure the case studies and to conduct interviews with each sovereign, a list of questions was designed and agreed upon with the Department of Finance and the Bank of Canada. The list was comprehensive and addressed all aspects of the effectiveness and efficiency of debt operations, but did not address implementation systems in detail, which is outside the scope of this study.

Literature Review: A literature review was conducted, comprised of a study of relevant auction literature and interviews with two leading treasury auction academics. Most academic literature focuses on the auction format, and there is relatively little academic literature that directly addresses the issue of auction effectiveness or efficiency.

Data Analysis: Data analysis was performed using auction data provided by the Bank of Canada for all auctions from spring 2003 to spring 2008 (May 7, 2003 to February 27, 2008 for nominal and RRBs and April 8, 2003 to March 18, 2008 for treasury bills). The analysis examined each security type separately and, for each, we measured the effect of Bank of Canada participation and financial market conditions on auction effectiveness. This report summarizes the results regression analysis aimed at evaluating auction effectiveness. We conducted a detailed analysis using multiple and simple regressions, but conclusions were generally similar. Any material differences are noted in this report. We also examined dealer and customer participation in auctions and their effects on auction results using detailed data provided by the Bank of Canada. The dataset represents all auctions during the evaluation period.

Interviews: Interviews with the Bank of Canada, the Department of Finance, and market participants were conducted in November and December 2008. Interview questions were developed to address the relevant evaluation questions. Six of the largest primary dealers and one smaller primary dealer were interviewed. At each, the head Government of Canada bond trader, at least one other bond trader, the head of money market trading, and the senior relationship manager for the Government of Canada were interviewed, usually simultaneously. In addition, several customers were interviewed, selected in consultation with the Bank of Canada based mainly on auction participation. Some customers participated primarily in nominal and/or real return bond auctions, others primarily in treasury bill auctions, and some in all auctions.

3. Limitations of the Approaches 

Time period and structure: It is important to note that the period of evaluation was prior to the financial crisis which began in 2008 when market volatility and debt issuance increased. However, interviews with market participants and sovereigns occurred after the crisis had begun, and some input was gathered to reflect the more volatile environment. We make comments about the impact of the crisis on conclusions throughout this document where relevant, but the full data analysis and interviews related to the auction process during the crisis period was not conducted.^[1]

Sovereign Comparison: A complete review of all G7 markets was deemed to be cost prohibitive and unnecessary. There are many common elements, but also many differences in the auction process across countries. Many aspects of the process are interrelated and each case study must be analysed in detail to draw relevant comparisons to Canada. General conclusions about specific best practices are not possible. Given the limited nature of the recommended improvements to the Canadian auction process based on the comparison to the selected sovereigns, it seems unlikely that additional analysis would have changed the conclusions or recommendations significantly.

Literature Review: We do not believe there are limitations to the literature review that would impact the conclusions of this study. The literature search was thorough, and we believe we captured a very broad sample of relevant literature. Most literature on bond auctions focuses on the auction structure (which is not considered in this report) rather than the auction process or the influence of market conditions on auction effectiveness or efficiency.

Data Analysis: Limitations of the data analysis are mainly due to the relatively short sample period which includes only a limited number of auctions conducted in periods of high volatility, credit market distress, or illiquidity. However, a longer historical sample period may not be indicative of future market conditions. We believe that the evaluation period represents a reasonable balance. The variability of different market conditions (liquidity, credit conditions, and volatility), within the sample period is limited, which may drive the generally weak data evidence. Although they were observed to be effective and efficient during the credit crisis, and we do not believe the conclusions of this study would change, the auctions during the credit crisis in 2008 and 2009 may be worthy of further data analysis. The data analysis is the primary line of evidence regarding some of the specific evaluation areas, such as the impact of Bank of Canada participation and the bond buyback program on auction effectiveness and the impact of different market conditions on auction effectiveness. Interviews with market participants and the sovereign comparison offer additional support for some of the conclusions, but the literature review provides no evidence in these areas. The market condition variables and measurement approaches are common in academic literature. However, the volatility measure could have been represented by implied option volatility. We did not use this more computationally complex measure because Canadian bond options are not highly liquid, and implied volatility may not be a better measure of overall bond market volatility compared to the rolling average we use. Historical and implied volatility of financial securities are generally highly correlated.

Interviews: A larger sample of dealers and customers would likely be of limited added value since the participants interviewed represented the largest most active dealers and customer participants, as well as two smaller dealers. The opinions of individual interviewees required some subjective interpretation since they were divergent and often appeared to reflect individual self interest rather than the effectiveness and efficiency of the auctions from the government perspective.

4. Overview of Governance for Debt Management Activities 

The legislative basis for the government’s borrowing program is Part 4 (Public Debt) of the Financial Administration Act (FAA).^[2] The FAA empowers the Governor in Council to authorize the Minister of Finance to borrow money. The Act provides the Minister of Finance with legislative authority to establish rules governing the auction of debt. In addition, the Act provides the Minister with powers over the management of the Government’s assets and liabilities. The powers of the Minister can be delegated to officials of the Department of Finance.

Ultimate decision-making authority for the Government’s funds management (activities related to market debt and liquid assets, including the foreign exchange reserves) rests with the Minister of Finance. The Minister of Finance approves policies for funds and risk management activities, and is responsible for seeking Governor in Council approval to borrow money on behalf of the Government for each fiscal year.^[3] Prior to the start of each fiscal year, the Minister must report to Parliament on the Government of Canada’s debt management strategy for the coming fiscal year (the Debt Management Strategy).^[4] Within 30 sitting days of the tabling of the Public Accounts of Canada in Parliament, the Minister must submit a report on federal debt operations for the previous fiscal year (the Debt Management Report).^[5]

Section 24 of the Bank of Canada Act^[6] provides statutory authority for the Bank of Canada to act as the government’s fiscal agent in the payment of interest and principal and generally in respect of the management of the public debt of Canada. Under a funds management governance framework,^[7] the design of key strategies and policies, the oversight of operations and the coordination of activities are jointly borne by officials at the Department of Finance and the Bank of Canada. Regular governance committee meetings facilitate work planning, coordination and communication between the institutions.

The Funds Management Committee, which consists of senior officials from the Department of Finance and the Bank of Canada, heads the committees under the governance framework. The mandate of the Funds Management Committee is to advise the Minister, through the Deputy Minister, on policy and strategy, to oversee the implementation of approved policies and plans, and to review performance outcome reports.

The Funds Management Committee is supported by the Risk Committee, whose mandate is to oversee and advise on the risk management policy and to report on financial risk positions. The Financial Risk Office at the Bank of Canada provides analytical support to the Risk Committee in this role and is responsible for monitoring and regularly reporting on the financial performance and position of certain financial assets and foreign-currency-denominated derivatives, including market, credit, operational, liquidity and legal risks.

Funds management programs and activities are subject to independent advice obtained through ongoing program evaluations. The Minister of Finance tables reports on the findings of these evaluations, including Department of Finance comments, with the Public Accounts Committee of the House of Commons. A copy is also sent to the Auditor General of Canada.^[8]

5. Description of the Auction Process 

a. Domestic Debt Operations

Domestic borrowing is conducted on a regular, transparent basis to maximize investor interest and participation. Nominal bonds, RRBs and treasury bills are sold via auction, with the Bank of Canada operating as the fiscal agent, to Government of Canada securities distributors and customers. Tenders are submitted to the Bank of Canada via the electronic auction system CARS (Communication, Auction and Reporting System).

Bonds are auctioned on a quarterly basis for nominal 2-, 5-, and 10-year maturities and for the RRB 30-year maturity, and on a semi-annual basis for the 30-year nominal maturity. Bonds may be either new maturities or reopenings of previously auctioned bonds. New bond maturities are generally reopened several times in order to achieve the target benchmark bond size to enhance liquidity.

The Quarterly Bond Schedule (QBS), setting out details of the planned quarterly issuance of marketable bonds, is published by the Bank of Canada prior to the start of each quarter. Final details, including the amounts to be auctioned, the maturity date, and the amount outstanding are released the week prior to the auction.

Bond sales take place via multiple-price auctions, with the exception of RRBs, which are sold via single-price auctions. Government securities distributors and customers may submit competitive tenders or non-competitive tenders. For multiple-price auctions, competitive bids are accepted, up to the bidding deadline, in rising order of yield (declining order of price) until the full amount of the issue being auctioned is reached, and all non-competitive bids are allotted first at the average of the accepted competitive bids. For single-price auctions of RRBs, bonds are allotted at the price equivalent of the highest real yield of accepted competitive tenders, plus accrued interest and inflation adjustment.

Treasury bills are sold via auction on a discount basis. Those with terms to maturity of approximately 3-, 6-, and 12-months are currently auctioned on a biweekly basis, generally on a Tuesday for delivery Thursday. Under the biweekly issuance pattern, 3-month treasury bills are reopenings of previous 6- and 12-month bills; while new issues of 6- and 12-month treasury bills are offered in the same week and then reopened once at the next regular auction two weeks later.

When conducted, bond buyback operations on a cash basis are held 20 minutes after nominal bond auctions. The QBS includes the target amount of bonds the government intends to repurchase during the quarter. Final details of each operation, including the maximum amount to be repurchased and the basket of eligible bonds, are released the week prior to the operation along with the release of bond auction announcement.

Buyback operations on a switch basis offer an opportunity for participants to exchange less liquid securities for new more liquid benchmark securities. Switch buybacks are announced in the QBS. Final details of switch buyback operations, including eligible bonds for repurchase, the replacement bond and the maximum replacement amount for the replacement bond, are published the week prior to the operation.

Cash management bond buyback operations target large bonds with less than 18 months to maturity. These are held on an irregular basis to meet government cash management needs. They are held on most Tuesday mornings after treasury bill auctions. Details of the operations, including the maximum amount to be repurchased and the basket of eligible bonds, are announced one week in advance.

Regular bond buyback and cash management bond buyback operations are settled on a cash basis and take place via multiple-yield reverse auctions. Switch buyback operations are settled on a replacement bond and cash basis. In all bond buyback operations, competitive offers are accepted in decreasing order of yield (increasing order of price) until the maximum amount to repurchase or the maximum replacement amount is met. The amount repurchased may be less than the maximum amount.

b. Domestic Distribution System

The participation of government securities distributors and customers at Government of Canada debt auctions is governed by a set of standard terms and terms of participation introduced in October 1998.

There are 21 government securities distributors that participate in the primary distribution of bonds and treasury bills. All must be either members or affiliate members of the Investment Industry Regulatory Organization of Canada (IIROC) and have their core trading and sales operation for Government of Canada securities in Canada.

Under the terms of participation, there are specific bidding limits that apply to government securities distributors and customers at treasury bill and bond auctions. The limits vary by government securities distributor based on the firms' relative market activity in the primary and secondary markets as well as its participation at buyback operations (bonds) and non-fungible cash management bills (treasury bills). However, for bond buyback operations, and non-fungible cash management bills, all participants have a 100 per cent bidding limit.

Primary dealers in the bond market have bidding limits tiered from 10 to 25 per cent of the auctioned amount for bids on their own account while other government securities distributors have bidding limits tiered from 0 to 9 per cent. Primary dealers in the treasury bill market have bidding limits of 25 per cent of the auctioned amount for bids on their own account while other government securities distributors have bidding limits of 10 per cent. Primary dealers can submit up to 25 per cent of the auctioned amount on behalf of customers whereas other government securities distributors can submit up to 10 per cent of the auctioned amount on behalf of customers.

In addition, the aggregate limit for the sum of the bids submitted by a primary dealer on its own behalf and on behalf of its customers is 40 per cent of the tender less the dealer’s excess net long position (up to the dealer’s bidding limit). A customer may bid for up to 25 per cent of the amount auctioned.

All government securities distributors also have ongoing reporting responsibilities to provide weekly statistical reports on their domestic fixed-income trading activities to IIROC and the Bank of Canada. In addition, government securities distributors and customers must report their aggregate net positions in the auctioned security to the Bank of Canada for all treasury bills, bonds, RRBs and switch operations when submitting their own bids or bids on behalf of their customers. Their net positions must be reported, whether they are long or short positions.

Government securities distributors that maintain a certain threshold of activity in the primary and secondary market for Government of Canada securities may become primary dealers, and form part of the core group of distributors of Government of Canada securities. The primary dealer classification can be attained in either treasury bills or marketable bonds, or both. Primary dealers assume a number of responsibilities with respect to Government of Canada securities. They must comply with minimum bidding requirements for every securities auction (except for non-fungible cash management bill auctions) and they must consistently make two-sided markets to a broad customer base. At every auction, a primary dealer’s submitted bids must total a minimum of 50 per cent of its auction bidding limit or 50 per cent of its formula calculation, whichever is less.

Prior to the start of each quarter, the Bank of Canada, on behalf of the Minister of Finance, publishes the QBS,^[9] which provides details on the Government of Canada marketable bonds to be auctioned. The calls for tender, auction results, terms of participation and other related information can also be found on the Bank of Canada website.^[10]

Regular and on-going consultations with government securities distributors, institutional investors and other interested parties are considered to be an integral part of the debt management process. Official consultations are held typically once a year (or as required) in order to obtain views on the liquidity and efficiency of the Government of Canada securities market, and the design and operation of the Government debt program. Consultation notices, questions and summaries are posted on the Bank of Canada website.

6. Sovereign Comparison 

Case studies of Canada, Australia, the United Kingdom (UK), the United States (US), and France, as well as interviews with auction officials in each country (except France) were conducted. Refer to Appendix A for a discussion of the results as well as a detailed table comparing the features and approaches in each country. Where relevant, a comparison to the experience of other sovereigns is referenced throughout this document.

In the context of increasingly integrated global markets, the major global issuers of debt tend to use broadly similar issuance procedures and debt management policies that facilitate or encourage liquid markets. This is due to a common desire to achieve low cost funding. This is facilitated by broad and deep primary and secondary markets and a high degree of transparency and predictability. For example, auction calendars and electronic auction systems are commonly used to achieve transparency in primary markets.^[11]

The many similarities between Canadian auctions and those of other sovereigns suggest limited ways in which other sovereigns’ practices might be used to improve Canada’s auctions. Transparency and communication with participants is at least as effective in Canada as it is in other sovereigns. Canada succeeds at rapidly releasing relevant information to participants after the auction, and the information released is similar to that of other sovereigns.

7. Literature Review 

See Appendix B of this report for a summary of the literature review and references.

8. Market Data Analysis 

The focus of the data and market analysis was on measuring the effect of Bank of Canada participation and financial market conditions on auction effectiveness. We also examined dealer and customer participation in auctions and their effects on auction results. Methodology and results are summarized in Appendix C of this report. Market effectiveness is measured by the auction Tail (the difference between the highest accepted yield and the average auction yield  – usually less than one basis point in normal market conditions) and the Bid-Cover ratio (ratio of total bids to auction size – usually over 2 times).

Auction sizes vary depending on the product and term-to-maturity of the instrument being auctioned. The chart below shows the average auction sizes over this period for each instrument. Real Return Bonds (RRBs) are only auctioned for 30-year terms and have a relatively small auction size, averaging $405 million compared to $1.4 billion for 30-year nominal bonds, $3 billion for 2-year bonds, and $5.1 billion for 3-month treasury bills. Auction sizes for 3-month treasury bills and 2-year bonds are significantly larger than for other terms, reflecting demand and contributing to their greater liquidity.

We measure market conditions by credit spreads, yield volatility, and trading volume. There is generally little statistical evidence of the impact of market conditions on bond auction characteristics. However, for treasury bills, higher volatility, wider credit spreads, and lower liquidity are related to wider auction tails, i.e. less effective auctions.

The Bank of Canada participates up to 15 per cent in nominal bond auctions and up to 25 per cent in treasury bill auctions. The general evidence is that these different levels of Bank of Canada participation do not impact auction effectiveness for any of the types of auction securities.

For nominal bonds, customer participation, which ranges from zero to 8 per cent, has little impact on auction effectiveness. Likely due to significantly larger customer participation in RRB auctions (representing 10 to 40 per cent of total bids), Bid-Cover and Tail are both favourably affected by increased customer participation. The relationship is stronger for auctions that occur near coupon payment dates. We find no relationship between the willingness of dealers to bid at auction and market conditions, but we did not measure this effect during the crisis.

The data suggest that the buyback program does not impact auction effectiveness for 2- and 5-year auctions. For 10-year auctions, larger buybacks are associated with wider tails for switch buybacks. There is weak evidence that as buyback amounts increase, the tail increases (effectiveness decreases). Note that the buybacks are intended to increase auction size and liquidity, which might be expected to contribute to more effective auctions.