Residential Mortgage Securitization
in Canada: A Review
Adi Mordel and Nigel Stephens
Residential mortgage securitization (together with
mortgage insurance) plays an important role in the
Canadian system of housing nance, especially given
the rising share of government-supported (i.e., public)
securitization over the past 15 years.
The main social benet for Canadians of public
securitization is the support it provides for both
diversity of choice and access to mortgage nancing
through a stable, cost-effective supply of funding to
mortgage lenders. Public securitization also supports
competition in the mortgage market by providing
funding to small lenders,
1
which have fewer alterna-
tive funding sources. Financial institutions also benet
from public securitization by using these highly rated
assets to meet regulatory requirements.
The recent increase in public securitization has also
led to public discussions about the government’s
exposure to the housing market, the balance between
investment in residential real estate and other forms
of investment, and the potential effects on household
borrowing and the housing market. One approach to
reducing the government’s involvement in the housing
market would be to consider adopting measures
to reinvigorate private mortgage securitization in
Canada.
Introduction
Mortgage securitization, the process of converting
illiquid mortgage loans into tradable securities, plays
an important role in the Canadian nancial system.
Over the past 15 years, the share of mortgage credit in
Canada that has been securitized has grown from about
1 In this report, the term “small lenders” refers to all nancial institutions that
access public securitization programs in Canada, excluding the Big Six
banks.
10 per cent to 33 per cent. Of the amount securitized
during that period, the share executed through public
securitization increased from 50 per cent to almost
100 per cent.
In this report, we analyze the evolution of both public
and private mortgage securitization in Canada to better
understand the underlying public policy and economic
determinants.
In particular, we consider the uses of mortgage securi-
tization by nancial institutions (FIs) to meet their funding
needs and regulatory liquidity requirements. As well,
we estimate that signicant benets accrue to the
nancial system as a whole from public securitization.
Aggregate mortgage funding costs are reduced by
about $870 million annually. In addition, Canadian FIs
save at least $120 million per year for every $100 billion
of National Housing Act Mortgage-Backed Securities
(NHA MBS) held for regulatory liquidity adherence.
2
We then review potential implications of the extent of
public securitization, noting that the Canadian govern-
ment has taken steps to adjust its framework for housing
nance to restrain the growth of public securitization.
We conclude with a discussion of policies that could
be considered to reinvigorate private securitization in
Canada.
2 On 11 December 2015, the government announced changes to its public
securitization programs. The estimates in this report are based on the
guarantee fees that existed before the 11 December announcement.
See the Canada Mortgage and Housing Corporation press release at
http://www.cmhc-schl.gc.ca/en/corp/nero/nere/2015/2015-12-11-0900.cfm.
RESIDENTIAL MORTGAGE SECURITIZATION IN CANADA: A REVIEW
39
BANK OF CANADA • FINANCIAL SYSTEM REVIEW • DECEMBER 2015
Mortgage Securitization in Canada:
The Context
Institutional background
The federal government supports housing nance
in Canada through mortgage insurance and public
securitization programs.
3
Federally regulated lenders
are required to obtain mortgage insurance on loans in
which the homebuyer has made a down payment of less
than 20 per cent of the purchase price.
4
Mortgage insur-
ance is provided by the Canada Mortgage and Housing
Corporation (CMHC) and private insurers; insurance
from both sources is guaranteed by the government,
although not to the same degree.
5
In Table 1, we illustrate the interaction between mort-
gage insurance and securitization in Canada. Public
securitization is provided through the NHA MBS and
Canada Mortgage Bond (CMB) programs, both admin-
istered by CMHC (Box 1 provides further detail on these
types of securities). Both programs use only insured
mortgages; public securitization of uninsured mortgages
does not exist in Canada.
Private-label securitization has existed in Canada since
1985. To reduce taxpayer exposure and encourage
development of private mortgage markets, the govern-
ment announced its intention to prohibit the use of
insured mortgages as collateral in non-CMHC securi-
tization vehicles.
6
To date, private-label securitization of
3 The government also supports housing nance through other means,
including tax credits and RRSP withdrawals for rst-time home buyers.
4 Low-ratio mortgages (with down payments greater than 20 per cent) can
also be insured by CMHC and private insurers with portfolio or trans-
actional insurance. See Crawford, Meh and Zhou (2013) for a detailed
discussion of the Canadian mortgage market.
5 CMHC mortgage insurance has a 100 per cent public guarantee, while for
private insurers it is only 90 per cent. The government guarantee is activated
when the insurer fails to honour its commitment to the lender.
6 See Government of Canada (2015).
uninsured mortgages primarily consists of short-term
asset-backed commercial paper (ABCP)
7
and some
longer-term residential mortgage-backed securities
(RMBS). New issuance of RMBS has been close to non-
existent in Canada in recent years.
Covered bonds are another important source of funding
that used to be backed by insured mortgages. However,
in April 2012, the federal government announced a regis-
tered covered bond framework to be administered by
CMHC. Under the framework, the bonds are only backed
by uninsured mortgages. Subject to the prudential limit
established by the Ofce of the Superintendent of Financial
Institutions (OSFI), an FI can have outstanding covered
bonds of no more than 4 per cent of its total assets. To
date, under the framework, total issuance stands at over
$70 billion. Covered bonds are an alternative to public and
private securitization as a source of funding for FIs.
8
Table 2 compares the level of government involvement in
the mortgage market across Canada, the United States,
the United Kingdom and Australia. Compared with the
United States, Canada exhibits a higher level of govern-
ment involvement in mortgage insurance but a lower
level of involvement in mortgage securitization. Australia
has an active private mortgage insurance system with no
public support, while the United Kingdom’s private mort-
gage insurance system is limited. The United Kingdom
also has temporary public mortgage insurance programs
created by the government during the nancial crisis. The
table also indicates that the four countries have broadly
similar rates of home ownership.
7 About 20 per cent of the underlying residential mortgages backing ABCP are
uninsured at present.
8 See the 2015 report by the C.D. Howe institute, “How to Make the World
Safe for (and from) Covered Bonds” (Poschmann 2015).
Table 2: Cross-country comparison (per cent)
Canada
United
States
United
Kingdom Australia
Share of public
mortgage
insurance
58.0 14.0 0.4 0
Share of public
securitization
34.0 55.0 0.0 0
Home-ownership
rate
67.6 65.1 64.6 67
Note: Public insurance in Canada is the insurance-in-force, i.e., the total amount
of outstanding loan balances covered by mortgage loan insurance policies by
CMHC and private insurers. For the United States, public insurance relates to
Federal Housing Administration and Veterans Affairs loans, which are insured
by the federal government. For the United Kingdom, it is the NewBuy Guaran-
tee and the Help to Buy programs. Share of public insurance is to outstanding
mortgage debt as of 2013 (for the United Kingdom, as of 2014). Shares of public
securitization to outstanding mortgage debt for Canada and the United States
are as of 2014. U.S. data are from the Securities Industry and Financial Markets
Association and the Federal Reserve Board and are based on the ratio of agency
MBS outstanding to total mortgage credit. See Chart 1-A for the Canadian data.
Home-ownership rates are as of 2013 except for Australia (2011).
Source: Bank of Canada
Table 1:
Types of residential mortgage securitization
inCanada
Underlying mortgage type
Securitization
Insured Uninsured
Public National Housing Act
Mortgage-Backed
Securities and Canada
Mortgage Bonds
Does not exist
Private-label Intention to disallow
announced
Asset-backed
commercial paper and
residential mortgage-
backed securities
Note: Covered bonds are a direct obligation of  nancial institutions issuing the
bonds and, hence, are not considered securitizations. Investors have recourse to
the covered pool in the event of issuer default.
Source: Bank of Canada
40
RESIDENTIAL MORTGAGE SECURITIZATION IN CANADA: A REVIEW
BANK OF CANADA • FINANCIAL SYSTEM REVIEW • DECEMBER 2015
Box 1
National Housing Act Mortgage-Backed Securities and Canada Mortgage Bonds
The NHA MBS Program, introduced in 1987, allows fi nan-
cial institutions (FIs) to issue mortgage-backed securities
(MBS) that are backed by pools of residential mortgages
insured under the National Housing Act. NHA MBS investors
are not subject to payment risk or the underlying mortgage
credit risk, owing to the Canada Mortgage and Housing
Corporation’s (CMHC) timely payment guarantee of interest
and principal, as well as the insurance on the underlying
mortgages. Before 2015, the annual cost of guaranteeing
the timely payment on a typical 5-year NHA MBS was four
basis points.
1
Although investors face no credit risk, they are
exposed to prepayment risk on the underlying mortgages
that o er amortizing monthly cash fl ows. The majority of
NHA MBS are fi xed rate and are issued for a 5-year term,
refl ecting the popularity of the 5-year fi xed-rate mortgage.
Since 2001, NHA MBS could be sold to the Canada Housing
Trust (CHT), which funds these purchases by issuing Canada
Mortgage Bonds (CMB). Similar to NHA MBS, CMB o er
investors a timely payment guarantee; the guarantee fee
is paid up front by the participating fi nancial institution.
Approximately half of newly issued CMB are fi xed rate for
5-year terms. Unlike NHA MBS, the CMB Program converts
monthly amortizing cash fl ows into typical bond-like payments
(i.e., semi-annual or quarterly coupon payments and a fi nal
full principal payment). Thus, CMB appeal to a much broader
investor base, and funding can be achieved at a relatively lower
cost than for NHA MBS.
The public policy objectives of the NHA MBS and CMB
programs are to “contribute to the e cient functioning,
competitiveness, and stability of the housing fi nance
system by helping ensure lenders and, in turn, borrowers
have access to a reliable source of funding for residential
mortgages regardless of economic cycles and market con-
ditions” (CMHC 2014). These objectives address the goal of
providing a reliable funding source throughout the economic
cycle and supporting competition in mortgage lending by
supplying cost-e cient funding to small lenders that have
limited access to alternative sources. Figure 1-A provides
breakdowns of the total amount of outstanding NHA MBS of
approximately $425billion by usage and issuer.
1 The cost of the annualized guarantee fee is higher than four basis points when
the average life of the 5-year NHA MBS is less than fi ve years. If, for example, the
average life were three years, the cost would be roughly seven basis points.
File information
(for internal use only):
Figure 1 -- EN.indd
Last output: 03:21:35 PM; Mar 14, 2012
Note: FRFIs are federally regulated  nancial institutions. Retained NHA MBS
are reported as pooled but unsold by FRFIs. Syndicated NHA MBS are not sold
only by the bank that created them, but rather by a syndicate of dealers. “Other
captures all remaining NHA MBS.
Sources: Canada Mortgage and Housing Corporation and Of ce of the
Superintendent of Financial Institutions
Sold to the CMB Program
Retained by FRFIs (NHA MBS pooled and unsold)
Syndicated NHA MBS
Other
Big Six banks
Non-Big Six banks
Figure 1-A:
Composition of outstanding National Housing
Act Mortgage-Backed Securities as of June 2015
By usage, Can$ billions
By issuer, Can$ billions
$213
$172
$24
$15
$305
$119
RESIDENTIAL MORTGAGE SECURITIZATION IN CANADA: A REVIEW
41
BANK OF CANADA • FINANCIAL SYSTEM REVIEW • DECEMBER 2015
Market developments
Chart 1-A shows the substantial rise in the share of
outstanding securitized mortgage debt. In 2000, only
about 10 per cent of the outstanding mortgage debt was
securitized, and half of that was through private pro-
grams. By 2015, about a third of the outstanding mort-
gage debt was securitized, almost all through public
programs. Not surprisingly, mortgage credit in Canada
has tended to move directionally with public securitiza-
tion, as is evident in Chart 1-B, which compares the
annual growth rates for the two series.
The rapid expansion of public securitization is espe-
cially evident in the period between 2008 and 2010, in
response to the Insured Mortgage Purchase Program,
which allowed mortgage lenders to pool insured mort-
gages into NHA MBS and sell them to CMHC to obtain
additional liquidity during the nancial crisis.
9
Currently,
the stock of public securitization continues to increase,
although at a slower pace, in part because of limits
imposed by the government on NHA MBS and CMB
issuance (Chart 2).
10
Before the nancial crisis, there was also an active
market for ABCP and, in 2006, approximately $20billion
of the underlying assets were residential mortgages
(some of which were insured). The non-bank-sponsored
ABCP market, which mainly invested in complex credit
derivatives known as collateral debt obligations that
were backed by U.S. subprime mortgages, experienced
severe disruptions in the summer of 2007, since issuers
were unable to roll over their short-term debt.
11
Since
then, the ABCP market has contracted substantially
and, as of June 2015, only about $10 billion of the
outstanding securities were backed by residential mort-
gages (Chart 3).
Several factors explain the rising share of public securi-
tization in Canada from both the demand and supply
perspectives. For FIs, CMB are a cost-effective funding
9 A description of the program is available at http://www.parl.gc.ca/content/
lop/researchpublications/prb0856-e.htm.
10 While annual issuance of CMB since 2013 has been held to $40 billion a
year, the annual issuance of NHA MBS was lowered to $80 billion a year for
2014 and 2015 (from $85 billion in 2013).
11 Kamhi and Tuer (2007) discuss the collapse of the non-bank ABCP market
in Canada.
File information
(for internal use only):
1-A.indd
Last output: 03:21:35 PM; Mar 14, 2012
Sources: Canada Mortgage and
Housing Corporation and Statistics Canada Last observation: June 2015
Private Public
0
5
10
15
20
25
30
35
40
2000 2003 2006 2009 2012 2015
%
Chart 1-A:
Ratio of outstanding securitization to residential
mortgagedebt
File information
(for internal use only):
1-B.indd
Last output: 03:21:35 PM; Mar 14, 2012
Sources: Canada Mortgage and Housing
Corporation and the Bank of Canada Last observation: December 2014
Public securitization (left scale) Mortgage debt (right scale)
0
2
4
6
8
10
12
14
0
10
20
30
40
50
60
2001 2003 2005 2007 2009 2011 2013
%
%
Chart 1-B:
Growth rates of outstanding public securitization
and residential mortgage debt
File information
(for internal use only):
2.indd
Last output: 03:21:35 PM; Mar 14, 2012
Source: Canada Mortgage
and Housing Corporation Last observation: June 2015
Canada Mortgage Bonds NHA Mortgage-Backed Securities
100
200
300
400
500
1987 1991 1995 1999 2003 2007 2011 2015
0
Can$ billions
Chart 2:
Outstanding public mortgage securitization
42
RESIDENTIAL MORTGAGE SECURITIZATION IN CANADA: A REVIEW
BANK OF CANADA • FINANCIAL SYSTEM REVIEW • DECEMBER 2015
tool, especially for smaller institutions that do not have
a branch network of deposits and lack alternative
funding sources. As well, from a regulatory perspec-
tive, NHA MBS qualify (as do CMB) as high-quality
liquid assets (HQLA) under the terms of the Basel III
Liquidity Coverage Ratio (LCR).
12
As of June 2015, about
40 per cent of the outstanding stock of NHA MBS was
retained by federally regulated FIs, which could help
them meet the LCR requirement.
For NHA MBS and CMB, investors also benet from a
timely payment guarantee (offered by the government
through CMHC for a fee, called the guarantee fee) on the
securities’ interest and principal. This enhances demand
for the securities, since investors do not face credit risk
or uncertainty as to the timing of cash ows from the
securities. In addition, the timely payment guarantee
allows NHA MBS and CMB to be government securities
from a credit perspective, which enhances their attract-
iveness to investors.
Quantifying the Impact of
Government-Supported Securitization
In this section, we examine the potential impacts
of public securitization in Canada, specically, the
benets that accrue to the nancial system and
12 Under Basel III, a bank needs to have an adequate stock of unencumbered
HQLA that can be converted easily and immediately in private markets
into cash to meet their liquidity needs for a 30-calendar-day liquidity
stress scenario. The LCR is the ratio of the stock of HQLA to total net cash
outows. The standard requires that, absent a situation of nancial stress,
the value of the ratio should be no lower than 100 per cent (i.e., the stock
of HQLA should at least equal total net cash outows). During a period of
nancial stress, however, institutions may use their stock of HQLA, thereby
causing the ratio to fall below 100 per cent.
FIs, and attempt to quantify two of them: the cost-
effectiveness of funding and the regulatory benet of
meeting the LCR.
13
Canadian mortgage lenders and borrowers benet
from the certainty and availability of funding provided
by CMHC securitization, especially through the CMB
Program. The regular schedule of CMB issuance and
relatively steady issuance volumes on a quarterly basis
provide lenders with certainty of cost-effective funding,
which is valuable for business planning purposes.
That value was highlighted in 2008 during the nancial
crisis, when access to market funding for FIs world-
wide became severely restricted. During that time, the
CMB Program continued to issue bonds on its regular
schedule, in increased volumes, albeit at wider spreads.
This is shown in Chart 4, which reports indicative
(expected) spreads for new issuances of NHA MBS and
CMB over 5-year Government of Canada bonds.
Another important benet of government-backed securi-
tization programs is that they limit severe procyclical
contractions in the extension of mortgage credit during
a crisis, when access to funding may be impaired. For
example, between 2008 and 2014, the average annual
growth rate in outstanding mortgage credit in Canada
was 6 per cent, whereas in the United States, mortgage
13 It is challenging to disentangle the benets of mortgage insurance from
those of securitization. For that, we would need a type of mortgage
securitization that does not exist in Canada, one in which the government
provides a timely payment guarantee on MBS that are backed by uninsured
mortgages (Table 1). Evidence from the United States suggests that in the
1990s and 2000s, the difference in interest rates for borrowers between
mortgages that were more easily securitizable and those that were not was
up to 24 basis points (Adelino, Schoar and Severino 2012).
File information
(for internal use only):
3.indd
Last output: 03:21:35 PM; Mar 14, 2012
Source: Dominion Bond Rating Service Last observation: June 2015
Total residential mortgages
0
5
10
15
20
25
2006
2007 2008 2009 2010 2011 2012 2013 2014 2015
Can$ billions
Chart 3:
Outstanding residential mortgages funded through
asset-backed commercial paper
File information
(for internal use only):
4.indd
Last output: 03:21:35 PM; Mar 14, 2012
Source: Royal Bank of Canada Last observation: September 2015
Canada Mortgage Bonds National Housing Act
Mortgage-Backed Securities
0
50
100
150
200
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Basis points
Chart 4:
Spreads of NHA MBS and CMB over 5-year
Government of Canada bonds
RESIDENTIAL MORTGAGE SECURITIZATION IN CANADA: A REVIEW
43
BANK OF CANADA • FINANCIAL SYSTEM REVIEW • DECEMBER 2015
credit contracted by approximately 2 per cent per year,
even after accounting for the increased issuance of MBS
by government-sponsored enterprises. While there were
clearly other important factors at play, public securitiza-
tion in Canada helped support growth in mortgage
credit during this period. Finally, Canadian banks also
use NHA MBS and CMB as collateral in repo trans-
actions. Gravelle, Grieder and Lavoie (2013) document
that these securities account for about 20 per cent of
repo collateral (classied as obligations of Crown cor-
porations). Further, FIs can pledge NHA MBS and CMB
in the Large Value Transfer System, which allows them
to use other securities for other purposes.
The cost-eectiveness of public securitization
Canadian institutions use a variety of sources to meet their
funding needs, with the mix depending on the cost-effect-
iveness of the options. Funding sources for the Big Six
banks include wholesale instruments such as short-term
debt and senior unsecured bonds, covered bonds backed
by pools of mortgages, securitized issuances (including
the NHA MBS and CMB programs and vehicles backed
by credit card receivables), and retail and corporate
deposits. Funding by the large Canadian banks may also
take place in a variety of currencies, in particular for senior
unsecured bonds and covered bonds, with the foreign
currency proceeds typically swapped back to Canadian
dollars. Small lenders are more limited in their funding
options and rely to a greater extent on the NHA MBS and
CMB programs for funding, as indicated in Chart 5.
We estimate the cost-effectiveness of funding from the
NHA MBS and CMB programs by comparing their cost
of funds with the cost of the next-cheapest source of
long-term wholesale funding. We measure how much
funding costs for lenders would rise if the NHA MBS and
CMB programs did not exist.
This approach follows the methodology employed by
CMHC’s evaluation of the CMB Program, which was
prepared by KPMG and released in 2008.
14
Although the
approach allows us to compare the cost-effectiveness
of the NHA MBS and CMB programs, its drawback
is that the methodology requires some simplifying
assumptions; namely, that the funding cost of the next-
cheapest alternative would not increase if the programs
ceased, that funding in sufcient size would be avail-
able from the alternative, and that CMB and NHA MBS
funding is raised only at the 5-year term.
Since funding costs on the cheapest alternative change
over time, as indicated in Chart 6, we report a range for
the funding advantage of CMB and NHA MBS in Table 3,
which is based on the chart.
15
The table indicates that,
over the sample period, the average cost advantage for
a Big Six bank from the CMB Program relative to the
next-best alternative was about 40 basis points, and the
relative benet of NHA MBS was about 11 basis points.
14 Canada Mortgage Bonds Program Evaluation (KPMG 2008).
15 Funding costs are based on biweekly dealer quotes between January 2013
and September 2015 and include guarantee and syndication fees. Guarantee
fees on NHA MBS and CMB are based on the fee level before 1 April 2015.
Funding costs are swapped back to Canadian dollars and expressed in terms
of a spread to the 3-month Canadian-Dollar Offered Rate.
File information
(for internal use only):
5.indd
Last output: 03:21:35 PM; Mar 14, 2012
Note: Ratios are as of June of each year.
Sources: Of ce of the Superintendent
of Financial Institutions and Canada Mortgage
and Housing Corporation Last observation: June 2015
Big Six banks Small lenders
0
5
10
15
20
25
30
2013 2014 2015
%
Chart 5:
Ratio of outstanding National Housing Act
Mortgage-Backed Securities to total liabilities
File information
(for internal use only):
6.indd
Last output: 01:45:37 PM; Nov 13, 2015
Note: CDOR = Canadian-Dollar Offered Rate
Source: Dealer quotes Last observation: September 2015
Canada Mortgage Bonds
National Housing Act
Mortgage-Backed Securities
US$ covered bonds
Can$ covered bonds
US$ senior unsecured debt
Can$ senior unsecured debt
2013 2014 2015
-10-10
10
30
50
70
90
110
Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul
Basis points
Chart 6:
Indicative 5-year all-in funding costs
(as spread to3-month CDOR)
44
RESIDENTIAL MORTGAGE SECURITIZATION IN CANADA: A REVIEW
BANK OF CANADA • FINANCIAL SYSTEM REVIEW • DECEMBER 2015
For comparison, the KPMG report, which evaluated only
the CMB Program over the 2001–06 period, concluded
that the average cost advantage of that program over
the next-best alternative was about 18 basis points.
Estimating the cost advantage of NHA MBS and CMB as
funding sources for small mortgage lenders is more dif-
cult. On the one hand, the cost of CMB funding is higher
for lenders who require third-party assistance in the
CMB swap and sourcing of replacement assets backing
CMB issues, reducing the relative cost advantage of
CMB funding.
16
On the other hand, the cost of alternative
sources of wholesale funding for small lenders is gener-
ally higher than that of the Big Six banks, increasing the
relative cost advantage of CMB funding. In addition, to
the extent that small lenders meet a higher proportion
of their total funding needs through the NHA MBS and
CMB programs than do the Big Six banks (as indicated
in Chart 5), the programs provide a greater relative
advantage to small lenders.
Given the overall supply constraint on CMB and NHA
MBS, small lenders also benet from the allocation
methodology used by CMHC to distribute NHA MBS
issuance and CMB funding among FIs. Available funding
is allocated equally to all FIs, regardless of their size
or requests for funding. As a result, small lenders are
able to access the public securitization programs for a
greater relative share of their funding needs, providing
more-stable funding sources and helping them to com-
pete against other mortgage lenders.
One can roughly estimate the alternative private funding
costs for small lenders by considering an RMBS issuance
completed in 2014. The weighted average spread of all
the tranches issued in the market was about 40basis
points over NHA MBS. Assuming that the averages from
Table 3 are representative, the issuer paid its RMBS
investors about 70 basis points more than what it would
have paid for CMB funding.
16 Lenders participating in the CMB Program must substitute maturing NHA
MBS sold to the Canada Housing Trust with replacement assets and must
engage in a swap with CHT where they exchange the interest ows on CMB
issues with those on the securities backing the CMB.
Based on the current outstanding stock of NHA MBS
and CMB, and assuming that their relative funding
cost advantage is the same across all institutions, we
estimate the aggregate annual funding benet of these
programs to be about $870 million for all FIs that access
the programs.
17
The use of public securitization to meet the
Liquidity Coverage Ratio
The Basel Committee on Banking Supervision requires
banks to have (at a minimum) sufcient HQLA to cover
stressed cash outows over a 30-day period (BCBS
2013). The total amount of HQLA distinguishes between
the highest-quality liquid assets (Level 1 HQLA) and
those that are somewhat less liquid (Level 2 HQLA).
While there is a cap on the amount of Level 2 assets
(they can comprise no more than 40 per cent of total
HQLA), there is no cap and no haircut on Level 1 assets.
In this sense, they can be held in unlimited amounts for
LCR purposes (i.e., total HQLA requirements can be met
entirely by any specic Level 1 asset).
18
Both NHA MBS and CMB qualify as Level 1 assets. They
have the added advantage of carrying a zero risk-weight
capital requirement because they are government guar-
anteed.
19
However, NHA MBS are an attractive instrument
for FIs to hold for the LCR because they are readily con-
vertible from mortgages on their books and have a higher
yield than Government of Canada bonds and CMB.
The advantage of using NHA MBS to meet the LCR
requirement can be estimated by considering the cost
of holding the next-cheapest alternative, provincial
bondswhich also qualify as Level 1 assets. We com-
pare the cost of converting insured mortgages to a
5-year NHA MBS held for the LCR versus buying provin-
cial bonds for the LCR by funding the purchase through
the cheapest wholesale funding instrument, covered
bonds, on the assumption that the NHA MBS and CMB
programs did not exist.
17 As of June 2015, the stock of outstanding CMB totalled $213 billion,
whereas the estimated outstanding stock of syndicated NHA MBS stood at
about $15 billion. Multiplying the outstanding amounts of these instruments
by their respective average cost advantages (40 basis points and 11 basis
points, respectively) yields a total benet of about $870million.
18 For further discussion on the Basel III liquidity standards, see Gomez and
Wilkins (2013).
19 According to OSFI, because NHA MBS are guaranteed by CMHC, they receive
a zero per cent risk weight in recognition of the fact that obligations incurred
by CMHC are legal obligations of the Government of Canada. See http://www.
os-bsif.gc.ca/eng/-if/rg-ro/gdn-ort/gl-ld/Pages/CAR_chpt3.aspx.
Table 3: All-in funding cost advantage of Canada Mortgage
Bonds and NHA Mortgage-Backed Securities versus the
next-cheapest private alternative (basis points)
Minimum Average Maximum
CMB 28 40 51
Syndicated
NHA MBS
0 11 28
Source: Bank of Canada
RESIDENTIAL MORTGAGE SECURITIZATION IN CANADA: A REVIEW
45
BANK OF CANADA • FINANCIAL SYSTEM REVIEW • DECEMBER 2015
Table 4 indicates that the benet of NHA MBS for LCR
purposes differs, based on the amount issued.
20
For the
rst $6 billion, FIs would save, on average, 22 basis points,
and for any amount issued above $6billion, FIs would save
about 12 basis points. This means that, for FIs in aggre-
gate, the benet for each $100 billion of NHA MBS held for
LCR purposes amounts to at least $120 million annually.
21
Potential Implications of
Government-Supported Programs
Although benets accrue to mortgage lenders from
accessing CMHC securitization programs, there are also
risks associated with these programs. There is the risk
that CMHC will be called upon under the timely payment
guarantee to meet interest and/or principal payments on
NHA MBS or CMB issues. CMHC reserves for this risk
by charging lenders guarantee fees, and it holds capital
against its securitization exposures of about $1.6 billion
(year-end 2014). There are, however, other potential
vulnerabilities and risks associated with public securiti-
zation from a nancial stability perspective that may not
be fully incorporated in the level of guarantee fees. We
review those below.
Impact on the supply of mortgage credit
Since lenders can securitize mortgages under the public
securitization programs in a cost-effective manner,
they may overextend mortgage credit and underinvest
in other productive assets (such as small business
loans). The latter may occur because mortgage-backed
20 The cost of 5-year NHA MBS is based on the guarantee fee schedule as
of 1 April 2015. For the rst $6billion, the upfront guarantee fee was set at
0.30 per cent, or 10 basis points annually, assuming that the average life of
NHA MBS is three years. Similarly, for any amount above $6 billion, the fee
was set at 0.60 per cent, or 20 basis points annually. We exclude the cost
of insuring the mortgages. Spread levels are relative to the 3-month CDOR
and are based on average biweekly dealer quotes between January 2013
and September 2015. We use 5-year Ontario bonds as the provincial proxy.
21 Holding CMB for collateral purposes is more expensive than holding prov-
incial bonds, given the relatively lower yield on CMB. However, the liquidity
of CMB may make them an attractive security for LCR purposes.
funding for FIs through public securitization is more
cost-effective and stable than non-mortgage-backed
funding, creating an incentive to extend more mortgage
credit than would occur without public securitization. An
increase in mortgage credit could lead to more lever-
aged households and elevated house prices.
While public securitization programs may support
competition, they may also increase vulnerabilities in
the nancial system by inuencing the business models
of mortgage lenders. For example, mortgage nance
companies (MFCs) are important participants in the
residential mortgage market. MFCs typically underwrite
and service insured mortgages sourced from brokers.
They tend to sell a large proportion of their mortgage
loans to federally regulated nancial institutions (FRFIs),
which may use them in CMHC securitization programs
for funding or regulatory purposes, or into CMHC
securitization programs. In this way, MFCs rely to a
considerable extent on funding from public securitiza-
tion programs. Without these programs, it is not clear if
MFCs’ other sources of funding, which are less stable
than deposits (e.g., syndicated lines of credit from
banks), would be reliable and large enough to support
their mortgage activities.
MFCs are less-regulated lenders (i.e., they are not dir-
ectly regulated by OSFI), although they must abide by
residential mortgage underwriting guidelines for FRFIs.
22
Limited available data also suggest that MFCs are highly
leveraged, leaving them less able to manage liquidity
and maintain income following an increase in mortgage
defaults (although mortgage insurance limits the eventual
losses). The participation of MFCs (supported by public
securitization programs) in the residential mortgage
market increases competition, but more transparency
and analysis are needed to better understand their
business models and their potential impact on nancial
system risk (see the June 2015 Financial System Review).
Use of securities for regulatory requirements
As noted earlier, since NHA MBS (and CMB) qualify as
Level 1 assets, FIs can use them in unlimited amounts to
meet the LCR requirement. As of year-end 2014, about
$184 billion in NHA MBS were retained on-balance-sheet,
mainly by the Big Six banks, and NHA MBS represent the
most effective asset for FIs to use for LCR purposes.
From a public policy perspective, when the government
was restricting the use of portfolio mortgage insurance
to limit public exposure to housing nance, it noted that
“[T]hese measures will restore taxpayer-backed portfolio
22 MFC-originated mortgages purchased by FRFIs must conform to OSFI
Guideline B-20, and MFCs are motivated to follow the principles set out for
mortgage insurers in OSFI Guideline B-21 so that mortgages can qualify for
CMHC securitization programs.
Table 4:
The cost advantage of adhering to the Liquidity
Coverage Ratio with NHA Mortgage-Backed Securities vs.
provincial debt
Annual cost of creating
NHA MBS
For the  rst $6 billion: 10 basis points
Any amount above $6 billion: 20 bps
Cost to hold provincial bond
for the LCR
Issue covered bond
Return on provincial bond
Total holding cost
Canadian-Dollar Offered Rate + 44 bps
CDOR + 12 bps
32 bps
Cost differential for the  rst $6B: 22 bps
for any amount above $6B: 12bps
Source: Bank of Canada
46
RESIDENTIAL MORTGAGE SECURITIZATION IN CANADA: A REVIEW
BANK OF CANADA • FINANCIAL SYSTEM REVIEW • DECEMBER 2015
insurance to its original purpose of allowing access to
funding for mortgage assets” (Government of Canada
2013).
Effective 1 April 2015, CMHC increased the guarantee
fees applied to NHA MBS for each FI and, in particular,
doubled the fees on issuances above $6 billion.
23
In addi-
tion to encouraging the development of alternative funding
options in the private market, the differential guarantee fee
structure may reect the variety of ways in which FIs use
NHA MBS. It is also consistent with the program’s stated
objective of promoting competition, since smaller lenders,
who are more likely to use NHA MBS for funding and
demand less than the $6 billion cut-off, will be paying lower
fees than FIs that demand larger amounts.
Eect on alternative funding models
Alternative funding vehicles, such as private-label securi-
tization markets, can be used to fund mortgages and
transfer and diversify risk in a way that would benet
the real economy (BoE and ECB 2014). In Canada, the
availability of low-cost publicly guaranteed funding may
reduce the incentive for FIs to explore the development
of alternative mortgage funding vehicles, namely private-
label mortgage securitization. For example, during its
review of the CMB Program, KPMG interviewed repre-
sentatives of the big ve banks, which indicated that “in
the absence of the CMB program, private securitization
vehicles would have been issued, probably by the big
ve banks as single issuers and possibly as multi-seller
vehicles for smaller players” (KPMG 2008, p. 31).
It is not certain, however, that FIs would develop alterna-
tive funding models if access to public securitization
programs were reduced. FIs could choose to utilize
existing funding sources to a greater extent in situations
where the benets of the alternative models are uncer-
tain, set-up challenges are high and their additional
funding needs may not be large.
24
Private securitizations
may also be limited, since they cannot be backed by
insured mortgages. As such, the development of private
vehicles depends in part on the growth rate of uninsured
mortgage credit and the extent to which it outstrips FIs’
existing funding sources.
23 See http://www.cmhc-schl.gc.ca/en/hoclincl/mobase/upload/
MBS_Advice_Guarantee_Fee_Increase-Dec-1-2014.pdf. For the issuance
of 5-year NHA MBS of up to $6 billion, the upfront guarantee fees increased
from 0.20per cent to 0.30 per cent. For any amount above $6 billion, the
fee was set at 0.60 per cent. For the issuance of a 5-year CMB, the guar-
antee fee was raised from 0.20 per cent to 0.40 per cent.
24 One alternative model is covered bonds, which are limited to four per cent
of the total applicable assets of the deposit-taking institution.
Policy Options to Promote Private
Securitization
The government could continue to reduce public
involvement in the housing market by adopting policy
measures to promote a private-label securitization
market.
25
In addition, the government could consider
changes to public securitization, which could take the
form of some or all of the following: further increases in
the cost to access CMHC programs, additional reduc-
tions in the issuance caps under CMHC programs or
restrictions on the eligibility of lenders able to participate
in the programs.
Fostering a private-label mortgage securitization market
in Canada could help to achieve a rebalancing of
private and public securitization. Such a market could
benet the economy by helping lenders fund assets
and diversify risks (Schembri 2014; BoE and ECB 2014).
In that respect, the Bank of Canada announced that,
as of April 2015, term asset-backed securities of high
quality, including residential mortgage-backed secur-
ities, would be considered as eligible collateral for the
Standing Liquidity Facility (SLF).
26
Other measures that
could promote an appropriate framework for private-
label mortgage securitization include principles for
eligible collateral, reporting requirements and structure
standardization.
Some steps to reduce public securitization have already
been put in place. As part of its 2014 budget, the federal
government announced that it would implement meas-
ures to reduce taxpayer exposure to the housing sector
and increase market discipline in residential lending.
For example, while the annual issuance of CMB since
2013 has been kept at $40 billion a year, the annual
issuance of NHA MBS was lowered to $80 billion a
year for 2014 and 2015 (from $85 billion in 2013) (GoC
2014). And, as mentioned earlier, the government also
raised the guarantee fees on NHA MBS and CMB as
of 1 April 2015, and announced further changes on
11 December 2015 (effective July 2016), to encourage
the development of alternative funding options in the
private sector.
Going forward, the government has other options in
addition to a further increase in guarantee fees or a
reduction in issuance caps on these securities. It could
also consider an auction-based mechanism whereby
the right to issue NHA MBS and the allocation of funding
under the CMB Program could be distributed based on
bidding by nancial institutions. As such, an institution
in need of funds would be willing to offer a higher price.
25 Another alternative could be to expand the use of covered bonds as a
source of mortgage funding. See Poschmann (2015).
26 For a detailed description of this change, see http://www.bankofcanada.
ca/2015/01/planned-changes-assets-eligible-collateral/.
RESIDENTIAL MORTGAGE SECURITIZATION IN CANADA: A REVIEW
47
BANK OF CANADA • FINANCIAL SYSTEM REVIEW • DECEMBER 2015
This approach might be appropriate, since prices would
more accurately reect demand and the riskiness of
the lenders.
27
However, this mechanism could impair
the competitive position of smaller lenders on a relative
basis.
Another approach could be to lower the size of the CMB
Program and dedicate it to small lenders, recognizing
that small lenders do not enjoy the same access to
funding as large lenders. An alternative that is less
distortionary than quantity constraints is for the govern-
ment to consider setting higher fees for large lenders
that participate in the programs. In general, both options
would be consistent with the policy objectives of the
CMHC securitization programs and with the philosophy
that government intervention in the market should take
place only in cases of market failure.
27 An institution that needs funding will offer a higher yield than an institution
with less-pressing needs at the time it submits its auction offer to CMHC.
Conclusion
The public “footprint” in the Canadian mortgage securi-
tization market has increased in recent years. The public
role provides stable mortgage funding for FIs and pro-
motes competition from small lenders in that market. It
also has consequences for the allocation of savings, the
business models chosen by small lenders and the cost
of regulatory compliance by banks.
This increase in the public footprint has led to a discus-
sion about the government’s role in housing nance
from a range of perspectives, including that of nancial
stability. The government has implemented a number of
measures in recent years to reduce the public’s involve-
ment. Further discussion and analysis of potential policy
options, including those to promote private mortgage
securitization, would be useful.
References
Adelino, M., A. Schoar and F. Severino. 2012. “Credit
Supply and House Prices: Evidence from Mortgage
Market Segmentation.” NBER Working Paper
No.17832.
Bank of England and the European Central Bank (BoE
and ECB). 2014. “The Case for a Better Functioning
Securitisation Market in the European Union.” Joint
Discussion Paper, May.
Basel Committee on Banking Supervision (BCBS). 2013.
“Basel III: The Liquidity Coverage Ratio and Liquidity
Risk Monitoring Tools.
Canada Mortgage and Housing Corporation (CMHC).
2014. Canadian Housing Observer 2014.
Crawford, A., C. Meh and J. Zhou. 2013. “The
Residential Mortgage Market in Canada: A
Primer.” Bank of Canada Financial System Review
(December): 5363.
Gomes, T. and C. Wilkins. 2013. “The Basel III Liquidity
Standards: An Update.” Bank of Canada Financial
System Review (June): 3743.
Government of Canada (GoC). 2013. Economic Action
Plan 2013. March.
Government of Canada (GoC), 2014. Economic
Action Plan 2014. February.
. 2015. Economic Action Plan 2015. April.
Gravelle, T., T. Grieder, and S. Lavoie. 2013. “Monitoring
and Assessing Risks in Canadas Shadow Banking
Sector.” Bank of Canada Financial System Review
(June): 5563.
Kamhi, N. and E. Tuer. 2007. “The Market for Canadian
Asset-Backed Commercial Paper, Revisited.” Bank
of Canada Financial System Review (June): 13–16.
KPMG LLP. 2008. Canada Mortgage Bonds Program
Evaluation. Available at www.cmhc.ca/en/hoclincl/
in/camobo/upload/CMB-Evaluation-Jun08.pdf.
Poschmann, F. 2015. “How to Make the World Safe for
(and from) Covered Bonds.” C. D. Howe Institute
E-Brief. Available at http://www.cdhowe.org/
how-make-world-safe-and-covered-bonds.
Schembri, L. L. 2014. “Housing Finance in Canada:
Looking Back to Move Forward.National Institute
Economic Review 230 (1): R45R57.
48
RESIDENTIAL MORTGAGE SECURITIZATION IN CANADA: A REVIEW
BANK OF CANADA • FINANCIAL SYSTEM REVIEW • DECEMBER 2015