Attracting large institutional investors for affordable housing developments

What is the policy issue?

Australia's affordable housing industry, where community housing providers (CHPs) supply affordable housing to households on low incomes, has, up to now, been modest and small scale. In order to expand so as to supply the large numbers of lower income households currently on waiting lists for public and community housing, the industry will need access to a large scale, dependable and sustainable source of finance that is available at an appropriate interest rate.

While Australian governments are unwilling to take on direct debt themselves to fund the affordable housing industry, there is interest in encouraging alternative sources of funding from the capital reserves held by large institutional investors, such as Australian superannuation funds. Indeed, the Australian government created the Affordable Housing Working Group (AHWG) to look at 'innovative, transformative and implementable financing models. These models will focus on the social housing sector, and the private rental market, and be targeted at low-income households.'

In November 2016 the AHWG recommended that Government investigate a Bond aggregation model along the lines outlined in AHURI research. In December 2016, the Australian Government’s Council on Federal Financial Relations established a bond aggregator taskforce to design a bond aggregator model and report back to Heads of Treasuries by mid-2017.

Adara Apartments, Success, Western Australia

Key fact

When governments are able to provide bond guarantees or to maintain bond aggregation entities they can help the CHPs borrow money from the financial markets at very competitive rates. Between 1993 and 2012 Austrian Housing Construction Convertible Bonds (HCCBs) raised Euro 15 billion (AUS $23.52 billion in 2016$) and were able to loan to social housing associations at interest rates that were just 0–0.3 per cent above the interbank lending rate.

HCCB finance helps the Austrian social housing sector supply between 14,000 and 16,000 dwellings per year, which is around 30 per cent of all new housing. In 2014 housing associations and co-operatives in Austria built 15,770 new dwellings, which was 31 per cent of all the 50,738 new dwellings built that year.

Austria and Australia new social housing compared 2014
Austria and Australia: new social housing compared (2014).

Sources: GBV, 2015; Statistik Austria, 2015; DHA 2015; ABS 2015; Australian Government Productivity Commission.

See also: Estimating numbers of social housing dwellings constructed in Australia.

Policy development options

There are a number of different policy options that could realistically encourage investment by large-scale institutional investors. These options include:

  • Housing supply bonds
  • Bond aggregation entity
  • Public Private Partnerships
  • Social impact bonds.
1. Explanatory statement

A combined suite of low-risk, low-yield 10-year housing supply bonds (HSBs) would allow CHPs to buy, develop and maintain affordable and social housing at interest rates below those available from the market. The bonds would be issued by a financial intermediary, with the assets and income streams associated with each bond issue isolated so that investors know the risks (if any) for their invested money.

Different sorts of investors need different sorts of bond types, with different tax incentives and government guarantee structures. As a result, three types of bonds would be required.

  • The largest bond type, raising 70 per cent of each bond's finance suite, would be the AAA housing supply bond that would raise money from superannuation funds (who pay lower tax rates). It would have a 15 per cent tax rate, a fixed interest, be long-term (up to 10 years); and have a level of a government guarantee.
  • The second private finance bond would be the tax smart bond for retail investors. It would raise 20 per cent of each bond's finance suite and would feature varying tax rates, a fixed term, fixed or indexed interest, lower yield, long-term bond (up to 10 years) with an appropriate tax incentive to generate a competitive after tax yield.
  • The third part (raising 10% of the HSB suite) could come from a government public growth bond that paid no interest, in effect a long-term revolving loan.

As a result of enhancements that include a guarantee on corporate bonds, which lifts their rating and reduces yield required; a tax exemption on income earned from retail housing bonds to reduce the before tax yield required; and a zero interest government loan, the cost of funds will be reduced to below the cost of current loans for affordable housing.

2. A real world example

The HSB model is based on the successful Housing Construction Convertible Bonds (HCCBs) scheme that has operated in Austria since 1993.

HCCBs provide lower-cost funds for commercial loans with 20–30 years maturity at 0.3 per cent (0–30 basis points) above the Euribor rate, at either a fixed or a variable interest rate. [Euribor stands for Euro Interbank Offered Rate. Its rates are based on the average interest rates that a large number of European banks borrow funds from one another at.]

Being able to borrow at a rate so close to the rates banks use is of great benefit. If Australian CHPs were able to source money at similar rates to HCCBs they might be paying around 2.6% at a time housing investors are paying 4.79 per cent or more. Such a scenario is based on the Australian 3 month interbank lending rate being around 2.28 per cent in February 2016, while the rate banks loaned to housing investors for one year was 4.79 per cent, rising to 7.59 per cent for seven years.

For investors, HCCBs provide a yield that is 1 per cent lower than capital market bonds. When combined with a tax advantage, however, the bond offers an attractive long-term low-risk, ethical form of investment that is widely held in Austria. The HCCB has no government guarantee but is backed by public loans and grants. Additionally, the sound financial position and robust regulation of the limited profit sector gives comfort to investors.

One of the lessons of the Austrian model is that bonds must be structured to not only generate low cost private finance but also to ensure that this produces developments that serve intended target groups and meet affordability requirements or any other specified policy objectives. Government regulations are required to impose strict conditions on how the funding is to be used, monitored and accounted for.

3. Scope of the practice

The 2012 HCCBs had generated around Euro 15 billion (AUD $20.5 billion) of investment, and contributed around 40% of project finance for new affordable and social rental housing and the renovation of older social housing in Austria.

4. Effectiveness/impact

In Austria, the limited-profit housing sector comprises 23 per cent of the housing stock and 48 per cent of the rental stock. A competitively allocated public loans and grants system drives efficiency and quality between providers, contributing a stable supply of between 14,000 and 16,000 dwellings per year (typically around 30% of all new supply).

Although HCCBs pay a lower yield than capital market bonds, combined with a tax advantage, they are popular with risk adverse investors and are held by an estimated 300,000 bond holders, 95 per cent of whom hold bonds to maturity.

5. Guide to evidence

Evidence on the operation of housing supply bonds is found in AHURI Final Report No. 188 Housing Supply Bonds: a suitable instrument to channel investment towards affordable housing in Australia?

1. Explanatory statement

A bond aggregation entity would source large amounts of capital ($50–200 million and upwards) through low-yield, (initially starting at around 1.5 per cent above interbank lending rates) long-term (up to 10 years) bonds from the finance market and then loan that money to CHPs at lower costs than would be possible through expensive one-off transactions such as borrowing from a bank (who charge interest rates up to 2.5 per cent above the interbank lending rate).

One bond aggregation option is for goverment to create an expert, independent, not-for-profit entity, such as the Affordable Housing Finance Corporation (AHFC).

Through its management team, the AHFC would:

  • assess the risks and benefits of applications for borrowing money from individual CHPs
  • combine the approved borrowing applications from many different CHPs
  • raise large volumes of money ($50–200 million and upwards [(p.94)]) from long-term, low-yield bonds issued by specific banks to institutional investors
  • distribute the money to the applying CHPs
  • monitor that the CHPs were using the money properly and effectively
  • collect the repayments (of both interest and principal) from the CHPs
  • repay the banks who would return money to investors.

AHFC bonds would come with a carefully structured guarantee, such as that interest and loan payments due to investors would be paid by the Australian Government should a CHP borrower be unable to make a repayment. This lowers the perceived riskiness of the bond resulting in a lower interest rate paid to the investor.

Standard and Poor’s noted that, compared to Australian Government balance sheets, the size of potential risks to be guaranteed by the AHFC was minute, and therefore was unlikely to impact the Government's credit ratings (currently at the highest AAA level).

The AHFC would initially restrict borrowers to CHPs that have Tier 1 status under the National Regulatory System for Community Housing. Tier 1 CHPs have demonstrated they have performed well while managing a high level of risk based on operating at large scale (e.g. managing more than 350 properties) and have ongoing substantial development activities.

For the Australian scheme to remain economically feasible, new affordable housing projects built by finance sourced through the AHFC would be administered by CHPs and the properties rented in the private rental market (with rents charged at 80% or less of those charged in the local area) or sold under some form of shared equity model. In both cases, the dwellings would be targeted at people working in lower paid jobs or those who otherwise can generate an adequate income (such as self-funded retirees). As a consequence, the properties are unlikely to be available for households whose only income is a government welfare payment.

The AHFC's benefits are that it is relatively simple and transparent; minimises the impact of debt on government budgets; draws on the successful experience and expertise of other countries; and provides lower cost finance to CHPs which is therefore likely to maximise the sustainable expansion of affordable housing stock.

2. A real world example

The AHFC model has been adapted to Australian financial market conditions from established Swiss and UK schemes that have been the tried and proven for the past two decades. Financial intermediaries such as the Swiss Bond Issuing Cooperative Emissionzentrale für Gemeinnützige Wohnbauträger (EGW) and the UK’s The Housing Finance Corporation (THFC) have played a market-forming role by substantially reducing the cost of financial intermediation and have ensured a stable flow of lower cost longer term investment towards the affordable rental housing market.

In December 2016, the Australian Government’s Council on Federal Financial Relations established a bond aggregator taskforce to design a bond aggregator model and report back to Heads of Treasuries by mid-2017.

3. Scope of the practice

The EGW bonds have raised money for affordable housing at rates that have consistently matched the very low rates for Swiss Government Bonds[(p.31)]. In April 2015 the Swiss Government sold 10 year bonds that had a 1.5 per cent interest rate and an effective (very low!) yield of minus 0.055 per cent.

The first EGW bond issue in 1991 raised 85.1 million Swiss francs (CHF) for a 10 year period at considerably lower rates than commercial lenders were offering (with a discount of 3–5%). As of March 2013, the EGW had issued CHF 4.49 billion in a series of 63 bonds and placements. [1 CHF = AUS $1.34 in December 2016]

Demand for EGW bonds are highly sought after, particularly amongst pension and insurance funds (buyers of 99% of bonds issued), and the bonds are sold within moments of issue. On average, four EGW bond issues are made each year, with bonds offered in 5000 CHF lots. The bonds are rated AAA and since 2003 have recorded no repayment defaults.

In the UK, the THFC, which started in 1987, has been licensed to issue guarantees on the government’s behalf. Since the beginning of 2012, more than £1 billion of funding has been raised by some CHPs via the bond issues to major investors such as pension funds and insurance companies.

4. Effectiveness/impact

The existing international bond schemes are successful when there is a diverse range of revenue streams. The UK and Swiss schemes have found they need some direct government subsidies to facilitate borrowing for developing new housing.

In June 2013, the UK government granted THFC (through its subsidiary Affordable Housing Finance–AHF) the right to issue government backed guarantees for loans to the value of a £3.5 Billion, although this right finishes in 2016.

Although evaluating the effectiveness of the guarantee scheme by simply counting the number of houses built is not possible (as money raised is also used for refurbishing dwellings), the financial effectiveness of the scheme can be seen in a late example (in January 2016) whereby AHF arranged a bond priced at an all-in rate of 2.81 per cent, which was just 0.32 per cent above the government bond rate and was amongst the lowest cost long term fixed rate funding ever seen in the sector.

In Australia, developing decent affordable housing close to employment opportunities will require a level of government subsidy, which may include Government guarantees, grants (in cash or land), planning gains through rezoning land or increasing densities, or a revolving low-or-no-interest public loan, as well as government funded Commonwealth Rent Assistance for those tenants who are eligible.

5. Guide to evidence

Evidence on the effectiveness of raising institutional funding through an intermediary is presented in AHURI Final Report No. 220, (2014) Enhancing affordable rental housing investment via an intermediary and guarantee, and AHURI Positioning Paper No.156 (2013) The use of guarantees in affordable housing investment—a selective international review.

Evidence on securitisation is found in AHURI Final Report No. 202 Financing rental housing through institutional investment—Volume 1: outcomes from an Investigative Panel, and AHURI Final Report No. 202 Financing rental housing through institutional investment—Volume 2: supplementary papers.

1. Explanatory statement

In Public Private Partnerships (PPPs) governments contract providers to build and/or manage social housing infrastructure.

Typically, initial funds for capital investment are raised by a private partner rather than the government. This investment is repaid over the concession period specified in the contract, through ongoing payments by the government as well as revenue from user fees. This is an advantage for the government in that such a funding model does not necessarily count in its financial books as a debt and can enable governments to initiate development of new projects without a substantial negative effect on its credit

2. A real world example

By the mid-2000s, an estimated 10–15 per cent of all Australian infrastructure investment came through PPPs, with 29 social infrastructure and transport projects contracted from 2005 to September 2009, the majority of these in NSW (10) and Victoria (9). In NSW, PPPs have procured social infrastructure and non-core services in a wide range of sectors, including transport (motorways, rail), housing, health, sports (Sydney Olympic Park) and other public services such as prisons, energy, waste and water treatment.

In Australia, the first social housing PPP redevelopment was to have occurred in the south-west Sydney suburb of Bonnyrigg. At the time of the signing of the initial contracts, the NSW government estimated the costs of facilities and services would be 6.3 per cent lower than if redevelopment was undertaken by government itself. However the Bonnyrigg PPP ran into trouble, with the developer Becton going into receivership in 2013, which meant the NSW Government has stepped in to complete the development.

3. Scope of the practice

In the US, the most significant use of PPPs for social housing has been through HOPE VI. The program has invested US$5.5 billion to renovate 224 public housing estates, resulting in some 150,000 public housing units being demolished to make way for reconstruction. The financial aims typically involve mixing private debt or equity with public subsidy, and using income from sales of private dwellings to offset the capital costs of rebuilding.

4. Effectiveness/impact

Funding developments through PPPs often involves an added degree of complexity as potential cost efficiencies may be balanced by high initial transaction costs given the investment required in the tendering process, the length of time it typically takes before contracts are finalised, and the legal and administrative fees along the way.

Research estimated that bidding costs for projects with a capital value up to $250 million are typically around $2.5 million, rising to up to $6 million for a $1 billion project. Significant start-up costs are also involved in putting together teams with necessary skill sets. The typical time for a contract to be signed for projects in Australia is 17 months.

5. Guide to evidence

Evidence on the effectiveness Public Private Partnerships is found in AHURI Final Report No. 163, (2011) Partnership working in the design and delivery of housing policy and programs.

1. Explanatory statement

The recent Federal Government Affordable Housing Working Group suggested the use of social impact bonds as one of its four models to stimulate investment in affordable housing.

Social impact bonds—also known as social benefit bonds—pay a return to an investor when an agreed social benefit outcome has been achieved by a service provider. These social benefits might be anything from improving conditions for people experiencing chronic homelessness to improving employment outcomes for long-term unemployed young adults. The return to investors is partly generated by the cost savings to government, and is conditional on the service provider actually achieving good results.

Social impact bonds were developed by Social Finance Ltd in the UK, and between 2007 and 2016 that organisation has raised over £100m investment for social benefit projects.

2. A real world example

Social impact bonds have been used for transitional accommodation, such as 'Sweet Dreams', (Canada's first social impact bond which led to the creation of a an independent living environment where up to eight young mothers and their children can live safely while the mothers work on completing their education or acquiring employment) but have yet to be used for developing long-term accommodation by community housing providers.

Nevertheless, social impact bonds for other important welfare functions have commenced in Australia.

The NSW Government created Australia's first social impact bond, the Newpin bond, in March 2013 to fund UnitingCare Burnside's New Parent and Infant Network (Newpin) program.

In June 2013, $7 million had been raised to finance the bond, allowing UnitingCare Burnside to run intensive courses for parents and to restore children to their families from out-of-home care. The bond is expected to last 7.25 years.

3. Scope of practice

By 30 June 2014, the Newpin program had supported 55 families with 82 children across three facilities, including opening a new facility. During that financial year, 136 families with 228 children participated in the Newpin program, and 28 children were restored to their families.

In 2014–15, 42 children were restored to their families, and children in 35 families were prevented from entering out-of-home care.

4. Effectiveness/ impact

In its first year (2014), the Newpin bond delivered a 7.5 per cent return to investors, while improving supports for parents of children in, or at risk of entering, out-of-home care. By August 2015 the bond generated an 8.9 per cent return to investors.

The success of bonds like the Newpin bond may encourage the concept's adaption to providing forms of community housing to eligible tenants.

5. Guide to evidence

Evidence on social impact bonds is found in AHURI Inquiry Discussion Paper (2015) The Inquiry into the funding of homelessness services in Australia.

Melbourne townhousesBackground to the policy issue

The AHWG, which evolved from the Federal Governments' federation reform and tax reform processes, has a key task in investigating and responding to appropriate methods to encourage institutional investment in affordable housing.

An AHURI investigative panel saw institutional investment as the most desirable source of finance to achieve long-term growth in supply of rental housing for a number of reasons.

  • Demand is so large that no-one else (including government) has access to sufficient funds to provide the finance needed.
  • Institutional investment offers efficiency gains from scale, and proportionally lower transaction costs for a small number of large investments rather than a large number of smaller contributions.
  • Institutions are likely to view longer-term lettings more favourably and to provide a more stable and predictable source of funds than individual investors.
  • Institutional investment will be needed if a new property asset class focused on income returns rather than speculative gains is to evolve.

Unlike small-scale individual investors, large-scale institutional investors judge potential investments by their return value (such as rental yield) rather than capital gains. Currently, large-scale institutional investors are less able to realise a profitable return from residential property investment, compared with individual investors, due to tax implications.

  • Superannuation funds cannot directly borrow to buy property and, therefore, cannot claim mortgage costs (which are available to individual investors who can use negative gearing rules to reduce these costs).
  • Superannuation funds only receive a 33.3 per cent discount on capital gains as compared to the 50 per cent discount extended to individual investors.
  • Companies can borrow to buy property; however, the 30 per cent company tax rate means that the tax shelter benefits are not as attractive compared to those obtained by individual investors in the higher tax brackets (who pay 45% tax on income over $180,000).
  • Companies' business losses are quarantined and can only be deducted in future years against income from the same business or other business that belonging to the same company, whereas individual investors can deduct housing costs from income earned from other sources.
  • Company capital gains do not attract the 50 per cent discount extended to individual investors.
  • Companies, superannuation funds and individual investors that invest in many residential properties are liable for land tax.

As a consequence, the risk-adjusted rate of return on affordable housing is considered too low by institutional investors. To encourage institutional investors for affordable housing, either rates of return would need to increase (not really possible for a housing class designed for those on low incomes) or the investment risks would need to be made much lower.

The risks institutional investors see include:

  • stamp duty and land taxes, which undermine already low returns in residential real estate
  • risks associated with a new asset class, for instance lack of market information on the returns likely to be available from affordable rental housing or on the performance of organisations able to manage such assets
  • counter-party risks, for example associated with the lack of comfort that CHPs can provide to a lender because they either are limited by guarantee or have little contributed equity or retained earnings, but also risks associated with development and construction of new dwellings
  • issues of security enforcement because of perception of a negative community response to any attempt to take over affordable housing assets in cases of default
  • scale constraints arising from the reluctance of relevant players to participate in schemes involving investment packages below a threshold size of possibly around $100 million, and the preference for a number of institutional investors to be involved in any such venture to share risk
  • administrative complexity arising from the possibility of having to deal with different tax and planning regimes across jurisdictions.

There is a strong role for government in reducing the gap between the real and perceived risk associated with affordable housing. There is also a role for government in devising ways to reduce the costs for CHPs of obtaining finance from large investors.

Estimating numbers of social housing dwellings constructed in Australia

Currently it is not possible to identify exactly how many new social housing (i.e. public housing and community housing) dwellings are built each year in Australia, and therefore we can only make an estimate.

For example, in 2014–15, 3,193 non-private sector residential dwellings were built, which included public housing built for and by state housing authorities. However, from this total we must remove the 706 dwellings built by Defence Housing Australia in 2014–15. This means only up to 2,487 non-private sector built dwellings can be considered public housing dwellings.

In addition, CHP building activity is counted by the ABS as private sector building, and is not separately identified. At June 2015, the number of CHP's 'tenancy rental units' increased from 71,036 in 2014 to 73,620, an increase of 2,584 dwellings.

While not all the new CHP 'tenancy rental units' may have been from new builds (they may have been purchases of existing dwellings or transfers of rental management from governments), we can't easily identify where the newly recorded dwellings came from and so classify them as new builds.

As a consequence, combined with the 2,487 public housing dwellings, we can estimate that around 5,071 new social housing dwellings were constructed in 2014–15. This equates to a maximum 2.7 per cent of all dwellings built in Australia in 2014–15 that were social housing dwellings.

Evidence in action

Research in progress on this issue

A number of completed AHURI research projects contribute to this subject.