Economic Impact of Limiting the Tax Deductibility of Negatively Geared Residential Investment Properties

Transcription

1 Economic Impact of Limiting the Tax Deductibility of Negatively Geared Residential Investment Properties MARCH

2 BIS Shrapnel Pty Limited March 2016 The information contained in this report is the property of BIS Shrapnel Pty Limited and is confidential. All rights reserved. No part of this report may be reproduced or transmitted in any form, nor may any part of or any information contained in this report be distributed or disclosed to any person who is not a full-time employee of the Subscriber without the prior written consent of BIS Shrapnel Pty Limited. The Subscriber agrees to take all reasonable measures to safeguard this confidentiality. Subscribers may not, under any circumstances, use information in this report for promotional purposes. Disclaimer: BIS Shrapnel has taken great care to ensure accuracy and balance in this report and the service it represents, but does not warrant the completeness or infallibility of the information. The entire contents are intended as general market information only and BIS Shrapnel implies no specific advice pertaining to the situation of any particular user and no warranty either expressed or implied is made as to the applicability of the information to the requirements or circumstances of any individual recipient. The opinions and forecasts are contingent by nature and materially different actual outcomes may eventuate. BIS Shrapnel Pty Ltd and its staff do not accept any liability for any loss or damage whatsoever arising out of the use or dissemination of all or any part (whether in printed, online, verbal or any other form) of this report or service, and intends by this statement to exclude any such liability. Job number: B6259 BIS Shrapnel contact: Kim Hawtrey Level 8, 99 Walker Street North Sydney NSW 2060 Australia T: +61 (02)

3 Contents EXECUTIVE SUMMARY ABOUT THE REPORT ASSUMPTIONS AND APPROACH Base case vs Alternative case How and when the policy is altered Financial impact on the typical negative geared investor IMPACT ON THE RENTAL MARKET Impact on investment returns Impact on rents and prices and development IMPACT ON RESIDENTIAL BUILDING ACTIVITY BIS Shrapnel s base case building outlook Investors and dwelling building Forecast dwelling commencements after the change initial phase Medium and long term dwelling forecasts Impact on dwelling stock deficiency ECONOMIC GROWTH IMPACT Impact on output Impact on jobs IMPACT ON HOUSEHOLDS Tenants and the Australian rental market Welfare impact on households of limiting negative gearing FISCAL IMPACTS Government revenues Tax receipts Stamp duty revenues GST receipts Other taxes and levies Capital gains tax Government outlays Summary of fiscal impact...28 BIS Shrapnel Pty Limited 2016

4 Tables Table 1: Tax benefit from negative gearing Sydney^... 8 Table 2: Tax benefit from negative gearing and required additional rent and yield to offset tax benefit, Table 3: Median unit rent and price forecasts, Sydney...11 Table 4: Projected median unit rent and price differential by 2026 with full and limited negative gearing by city...12 Table 5: Landlords average tax deduction...14 Table 6: Impact on dwelling commencements Australia, number, 5 year average...17 Table 7: Estimated increase in households (2011 numbers) experiencing housing stress due to rental increases resulting from limiting negative gearing...21 Table 8: Indicative impact on income tax revenue...24 Table 9: Indicative impact on stamp duty revenue...25 Table 10: Revenue estimate from council rates Residential property...26 Table 11: Indicative impact on council rates revenue (residential) 2020/ Table 12: Revenue estimate from land/other taxes Residential property...27 Table 13: Indicative impact on land/other tax revenue (residential) 2020/ Table 14: Projected difference in capital gains tax paid by 2026 by an individual investor - with full and limited negative gearing, by city...28 Table 15: Indicative impact on overall tax revenue annual average 2016/ / Charts Chart 1: Unit prices with full and limited negative gearing vs development costs, Sydney...11 Chart 2: Median unit rents, prices and yields with full and limited negative gearing, Sydney...12 Chart 3: Investor demand helps drive dwelling commencements...15 Chart 4: Removing negative gearing will reduce dwelling commencements...16 Chart 5: Tenure by dwelling type and state, five mainland states...19 Chart 6: Tenure by age of household reference person five mainland states...19 Chart 7: Household income by tenure type, five mainland states...20 Chart 8: Type of household by tenure type, five mainland states...20 BIS Shrapnel Pty Limited 2016

5 EXECUTIVE SUMMARY Negative gearing applies if the annual cash costs from investing in a leveraged property exceed the annual cash revenue generated. The current base case, under Australian tax law, allows the resulting cash loss to be deducted from the investor s income from other sources, reducing their overall taxable income and final tax bill. This Report measures the effect of limiting negative gearing, an alternative case that assumes: Abolition for established dwellings removal of tax deductibility of losses on established residential property against general income New properties exempt the change applies to established dwellings only; new dwellings continue to attract concession as per usual Grandfathering the new policy applies to purchases of property made on or after 1 July 2016, but purchases of property made before 1 July 2016 would not be affected Deductibility within property portfolio no restriction on negative gearing deductions against another property owned by the same taxpayer No change to related taxes - capital gains tax and stamp duty remain unchanged No change to other asset classes negative gearing offset remains for shares, etc In effect, the current negative gearing tax provision would be replaced by neutral gearing, for established properties. We find that limiting tax deductibility of negatively geared residential investment properties would have consequences that go well beyond any tax saving to the Federal budget: Rents will rise by up to 10% ($2,600) per annum New home building will shrink by around 4% nationally, or 7,200 dwellings a year GDP would shrink by around $19 billion per annum on average, equating to some 1% of Australia s $190 billion annual income 175,000 fewer jobs would be created over the next 10 years, resulting in the unemployment rate rising from 5.8% to 5.9% Government revenue across a range of taxes would shrink by $1.65 billion per annum 70,000 extra households would be pushed into housing rental stress If the government were to compensate these stressed households, it would require an additional subsidy outlay of $650 million per annum. In other words, the impact would go well beyond any saving of the income tax concession, to a multitude of unintended consequences. BIS Shrapnel Pty Limited

6 Why would limiting negative gearing have this adverse impact? The policy intervention would raise rents. Developers will find that the expected return from developing new dwellings is now lower, relative to cost, diminishing their ability to build new homes. Lower rental stock means higher rents. Over time, rents are projected to be between 1.7% and 10% higher across the capital cities by (Section 3). The housing shortage will grow. The policy change will create a discouraged investor effect that will dampen investor demand for housing. Although pre-existing investments would continue to receive the original tax benefit, current investors will now have a strong incentive to hang on to their grandfathered investment property, in order to prolong the legacy of the old regime. And while negative gearing would still apply to new dwellings, it is doubtful that this would attract significant investment funds into new dwellings: when you later came to sell a property acquired after 1 st of July 2016, the next owner won t qualify for negative gearing and therefore you would not be able to sell it without taking a price fall. This will lead to weaker new home building of around 4% nationally, or 7200 dwellings a year, resulting in a greater cumulative housing stock deficiency by the end of the forecast horizon. The impact would be concentrated in Sydney and Melbourne, and to a lesser degree Brisbane and Perth. Apartment building would bear the brunt, more so than detached houses. (Section 4). Lower new dwelling building numbers will mean real GDP is forecast to shrink by a cumulative $190 billion over the 10 years. The largest decline in output by industry sector is expected in dwelling building, which would be lower by $57 billion over the next ten years. In turn, sectors which service or supply inputs to the residential building industry will be slower, with a flow on to the rest of the economy. We estimate that around 175,000 fewer jobs will be created in the economy under the alternative scenario, and the unemployment rate will rise from 5.8% to 5.9%. (Section 5). More households would be pushed into housing stress. Rental dwellings are especially important to those on lower incomes, and the young. An additional 70,000 households would be pushed into rental stress. If the government were to compensate them, we estimate that the subsidy will range from $4 to $24 per week per household in each state, and amount to a total subsidy of $650 million per annum. (Section 6). Government tax revenues would shrink. The Federal government will save $2.1 billion per annum as fewer individuals deduct losses from property, but will incur an income tax loss from lower home building of $1.8 billion. Stamp duty revenue will fall by around $1.1 billion per annum; GST collections will drop by $0.2 billion a year. Altogether, total tax revenue is likely to fall by $1.65 billion per year. (Section 7). At the end of a 10-year horizon the market would find a new normal, but the legacy will be a lost decade. Rents will sit 2-10% higher depending on the city. The dwelling stock deficiency will be 60,000 homes (instead of 20,000). GDP growth and tax revenue will track lower. And more people will have been pushed over the threshold into housing unaffordability, resulting in a welfare loss predominantly borne by those in lower income brackets. 2 BIS Shrapnel Pty Limited 2016

7 1. ABOUT THE REPORT 1.1 Aims This Report provides forecasts of the likely economic impact of eliminating negative gearing tax concessions for fresh investment in established residential real estate. The analysis provides estimates of the effect on residential building activity, rents, economic growth, and fiscal revenues. BIS Shrapnel has been commissioned to trace the economic impact of limiting negative gearing provisions in the Australian tax code for investment in residential real estate. The key objective of the Report is to assess how limits on negative gearing provisions would affect the economy. The Report aims to quantify the implications for the housing market, economic growth, household welfare, and for government revenue. To analyse negative gearing that operates via the housing sector, a core requirement is to measure the likely effect on rents, rental yields and house prices, and potential impact on future residential building activity. The effects would work themselves out over time. Hence the Report gives the likely phasing of activity over the short, medium and long term a total forecasting horizon of 10 years. 1.2 Scope The Report clarifies the logic of the role negative gearing plays in the house building and rental markets and the impact of removal, compared with the current situation as base case. The analysis tracks the consequences of the change in policy and quantifies the impact on different parts of the market. The Report provides two scenarios ( base, alternative ) for forecast levels of: Housing cost - rents and dwelling prices before and after the policy change, and the impact on rental households Dwelling supply - new dwelling commencements, over a ten-year horizon. The number of commencements by state and nationally will be provided. We need to do this by state as state cycles are at different stages and will react differently. Economy - macroeconomic indicators (e.g. GDP, employment) and government revenues likely to be impacted We explain the fiscal impact on governments and estimate a broad magnitude of impact, at Federal level (e.g. loss of income tax revenue from lower building activity) and State level (e.g. loss of revenues to state governments via reduction in stamp duty). The reality is that the impact would go well beyond any saving of the income tax concession, to a multitude of unintended consequences. BIS Shrapnel Pty Limited

8 2. ASSUMPTIONS AND APPROACH 2.1 Base case vs Alternative case The approach in this Report is to compare our current status quo forecasts ( base case ) with an alternative forecast that assumes the limiting of negative gearing ( alternative case ). The story follows the logic from rental yields, through to building activity, and on to the wider macroeconomic impact. The restriction of the negative gearing tax offset, over time, will put upward pressure on rents as landlords seek to protect their after-tax rental yields. However, the required rise in rents will not happen overnight, causing a drop in investor demand and less investor support for new development, producing a softer outlook for dwelling building. This restriction on the supply of housing will eventually provide the basis for rents to rise, lifting yields and eventually bringing investors and developers back into the market. Detrimental effects on the economy and vulnerable households will follow. The end result will be a revenue shortfall for government, and a welfare loss to households. 2.2 How and when the policy is altered Gearing is not restricted to property investments; it can apply to any type of asset. It involves borrowing to invest in an asset, such as property. A property investment has positive gearing if the annual cash costs (loan interest plus other outlays such as building maintenance) are less than the annual cash revenue (e.g. rent) generated, resulting in a cash surplus. Negative gearing applies if the annual cash costs exceed the annual cash revenue generated. Under current Australian tax law, the resulting cash loss can be deducted from the investor s income from other sources, reducing their overall taxable income and final tax bill. To analyse the effect of limiting negative gearing tax benefits, we make the following assumptions about exactly how and when the policy would be changed: Abolition for established dwellings removal of tax deductibility of losses on established residential property against general income New properties exempt the change applies to established dwellings only; new dwellings continue to attract concession as per usual 1. Grandfathering the new policy applies to purchases of property made on or after 1 July 2016, but purchases of property made before 1 July 2016 would not be affected Deductibility within property portfolio no restriction on negative gearing deductions against another property owned by the same taxpayer No change to related taxes - capital gains tax and stamp duty remain unchanged No change to other asset classes negative gearing offset remains for shares, etc. The current negative gearing tax provision is assumed to be replaced by neutral gearing (on established properties) 2. 1 If new dwellings were also included, the impact would be somewhat greater than the estimates in this Report 2 This would still allow tax deduction of losses on a residential property against the taxpayer s taxable income from their total property portfolio, but only up to a limit equal to the (current and future) income from that portfolio, or when the property is sold, against capital gain (i.e. deferral of tax deductions, with some leakage). 4 BIS Shrapnel Pty Limited 2016

9 2.3 Financial impact on the typical negative geared investor The net cashflow of a property (or property portfolio, if the investor holds more than one property) is normally calculated as the annual rental income minus cash expenses 3. if the portfolio (which for many investors comprises only one property) is cashflow positive, the investor is positively geared and the taxpayer expects to pay tax on the net proceeds, as part of their overall annual income tax assessment if the portfolio is cashflow negative, the taxpayer is negatively geared and, under current legislation, is able to claim the loss as an offset against their general income from other sources. This will effectively reduce their taxable income at the margin, and so the benefit of negative gearing applies at their marginal tax rate. It follows that any benefit foregone following the removal of negative gearing would impact at the investor s marginal tax rate. Suppose negative gearing provisions were removed from 1 July 2016 and consider the case of a typical negative geared investor, noting: according to the most recent figures published by the Australian Tax Office (ATO) in April 2015, the average reported loss across all those claiming a negative gearing property deduction is around $10,000 per annum 4. BIS Shrapnel predicts that in June 2016 the national median detached house price will be $701,000 and median attached dwelling (multi-residential) price will be $510,000 we forecast that the interest rate for investor standard variable home loans will range between 4.25% and 5.70% in June 2016 Because more investors tend to buy an apartment rather than a detached house, we base our typical case on a multi-residential dwelling. In round figures, based on national averages, currently the typical investor s situation is: an investor borrows $500,000 to buy an investment property, and pays interest at 5% per year to the bank (interest-only loan), incurring a total interest outlay per year of $25,000 the investor s net rental income for the year after expenses such as repairs, maintenance, insurance, agents' fees, council rates, etc. is $15,000 the individual faces a marginal tax rate of 35% 5. 3 Cash expenses can be claimed immediately against current year s income. These could include loan interest, leasing fees, strata fees, repairs, maintenance, council rates, accounting fees, advertising, bank charges, gardening, insurances, government charges, property management, water/utilities, sundries. Depreciation and capital works (improvements, alterations, additions) are typically deducted over a number of years. 4 The actual figure was $9,558 for the 2012/13 tax year. In the year prior, 2011/12, it was $10,946. The average claim may vary slightly from year to year, but has typically been around the $10,000 mark in recent times. Higher house prices can push up the average claim through higher costs, while lower interest rates will tend to work in the other direction. For 2015/16, we assume higher house prices roughly offset the effect of lower interest rates, and retain the working estimate of $10,000 as the average claim across all negatively geared investors. For comparison purposes, the national median detached house price in June 2013 was $566,300 and median attached dwelling price was $457,200 (source REIA) while the interest rate in June 2013 for investor standard variable home loans ranged between 4.74% and 6.26% (source Canstar). 5 According ATO data, the largest group (40%) of negative gearing claimants are in the $37,000-$80,000 income bracket (top marginal tax rate = 32.5%), and the next largest in the $80,001-$180,000 range (MTR = 37%). We use 35% as an assumed rate. BIS Shrapnel Pty Limited

10 Because the net income from the property ($15,000) is less than the interest outlay ($25,000), the investor can currently offset the net loss ($10,000) against income from other sources (such as wages) before calculating tax payable - making for an expected tax saving of $3,500. Negative gearing Example Net rental income $15,000 less Interest + Operating costs -$25,000 = Net gain /loss -$10,000 (Tax saving*) ($3,500) *$3,500 = ($10,000 deduction from taxable income) x (35% marginal tax rate) After the change, however, the tax benefit - at least in the current year, against current wage income - would disappear for established dwellings. The offset would, in most cases, now be deferred 6. Landlords will seek to increase rent by $3,500 p.a. to compensate. As the change is assumed to be grandfathered, only new investment decisions (in existing dwellings) would be affected. Pre-existing investments would continue to receive the original tax benefit. By the same token, existing investors will nevertheless sharply change their behaviour, becoming far less likely to flip properties. They will now have a strong incentive to hang on to their grandfathered investment property, in order to prolong the legacy benefits from the old negative gearing regime. This will severely limit the number of properties for sale, and reduce stamp duty revenues of state governments. Negative gearing would still apply to fresh investment in new dwellings. However, it is doubtful that this would attract significant investment funds into new dwellings. When you later came to sell a property acquired after the 1 st of July 2016, the next owner won t qualify for negative gearing and therefore you would not be able to sell it without taking a price fall. Moreover, new dwellings are more expensive and many would-be investors will face a price barrier. The discouraged investor effect will be strong, even for new dwellings. 6 Cash expenses could no longer be claimed immediately against current year s income. However, we assume they would now be treated as a capital expense and accumulate along with depreciation and capital works, to be deducted a number of years later against capital gains and/or positive geared income from the property, if and when these eventuate. Even so, this would substantially defer the financial benefit for tax purposes, potentially for 5-10 years or more, and have a serious deterrent effect on new investment for most average investors, who are in the $30,001-$80,000 tax bracket and currently rely heavily on immediate deduction against current year wage income. A current cash amount is replaced by a future amount, which is worth less in today s dollars because it must be time discounted (Net Present Valued). Uncertainty also rises: a certain cash amount is replaced by an uncertain amount contingent upon future rental income and dwelling prices; a Known Value replaced by an Expected Value with a probability < 100%. 6 BIS Shrapnel Pty Limited 2016

11 3. IMPACT ON THE RENTAL MARKET A disconnect exists in the residential market between rent formation and price formation. Rents are formed by the interaction of demand and supply in the rental market. Prices are formed by the interaction of demand and supply in the broader sales market, including owner occupiers, subject to the constraint that prices need to be above development costs to justify developer building. While prices can fall below development costs for a period, the consequent lack of building creates a shortfall that drives up prices and subsequently supply. Placing the proposed limitation on negative gearing would result in a short run correction in real prices due to lower investor demand. However, it will result in higher rents than would be expected with negative gearing. Landlords will require higher yields to compensate for the lost negative gearing concessions. Can prices fully correct to accommodate the higher yields at current rents? This is unlikely 7. At current rents a fall in prices would reduce the feasibility of development, causing construction to fall and rental supply to dry up, which in turn will drive up rents at a greater rate than otherwise. Moreover, development will not re-start until prices rise sufficiently to support future development and rental supply. The impact will be mostly felt in the apartment sector. Banks require sufficient pre-sales in multi-unit residential developments (primarily to investors) to underwrite the finance for a development. Previous research undertaken by BIS Shrapnel suggests that owner occupiers are more comfortable purchasing units after the dwelling is completed and onsold. Any change that negatively impacts on investor demand will have a corresponding impact on multi-unit residential dwelling supply, which accounted for 45% of new dwelling starts in 2014/15. The limitation of negative gearing exclusively to new dwellings is not expected to have a significant impact on the market for new dwellings: Firstly, new dwellings allow a higher level of negative gearing concession via depreciation benefits, and are therefore already attracting their share of investors influenced by the prospect of greater negative gearing deductions. Secondly, the inability to fully negatively gear an established dwelling would mean the price of new stock will immediately fall as subsequent buyers cannot receive the same negative gearing concession. This will deter many investors from investing in new dwellings. Consequently, any impact that the limitation of negative gearing may have on established home prices is likely to have the same relative price impact on the new home market. 3.1 Impact on investment returns While negative gearing by its nature generates a loss, the loss itself is offset against other taxable income to generate a tax saving. This tax return represents a rebate to the landlord that would otherwise not be available if negative gearing was not in place. To generate the equivalent return, a landlord would require additional rent commensurate with the loss of the rebate. 7 We expect the bulk of the initial adjustment to occur via rents rising, rather than dwelling prices falling. Grandfathering means most existing investors will become reluctant to sell after 1 July Nominal house prices typically are sticky in a downwards direction in Australia. House prices are a mark-up on construction costs, which provide a floor. And during when negative gearing was briefly removed, the effect showed up in higher rents (see discussion of on page 37). BIS Shrapnel Pty Limited

12 Table 1 shows a simple first year calculation of the tax concession generated by negative gearing based on indicative medians for rents and prices for multi-residential dwellings in Sydney. The analysis is based on the following assumptions: rent is based on the two-bedroom median multi-unit residential dwelling rent and median multi-unit residential dwelling price for Greater Sydney the purchaser borrows on a 90% loan-to-value ratio. The interest rate is for an interest only loan at the prevailing standard variable rate at June 2015 of 5.45% outgoings (i.e. owner s corporation, rates, etc.) reduce the yield by an effective 0.75 percentage points the landlord purchases an established dwelling i.e. no additional deductions for building depreciation, depreciation of fixtures, fittings, etc. the landlord s pre-tax income is equivalent to average weekly male full time earnings for NSW. Table 1: Tax benefit from negative gearing Sydney^ Median rent (per week) $510 Median rent (annual) $26,520 Median price $675,000 Indicative yield 3.9% Deductions Outgoings 0.75% of yield $5,063 Interest 5.45% pa. $33,109 Net income/loss on rent -$11,651 Average earnings (per annum) $81,095 Tax on gross income $17,903 Tax on taxable income after negative gearing $14,116 Tax concession $3,787 Annual rent required to compensate for loss of negative gearing $30,307 Required yield to compensate for loss of negative gearing 4.5% ^The analysis is based on a June quarter 2015 estimate The table indicates that the current initial rental yield for an indicative multi-unit residential investment is 3.9%. At this yield, the landlord experiences a taxable loss of $11,651 in relation to his investment. This results in an annual tax concession of $3,787 dollars per annum after claiming the loss against wages income. To estimate the yield required, we assume that the landlord compensates for the loss of tax concessions. On that basis, if negative gearing is removed a landlord would seek an increase in rent of $3,787 per annum ($73 per week) to generate a commensurate return. This translates to a higher indicative yield of 4.5% required by the landlord to compensate for the lack of the ability to negatively gear. Arguably, in a more normal interest rate environment, this yield would also be higher. Across the other capital cities, the annual negative gearing benefit ranges from $506 per annum (equivalent to $10 per week) in Hobart, to $2,936 ($56 per week) in Melbourne (Table 2). In effect, this concession is a subsidy currently provided by the Federal Government to allow landlords to offer their dwelling at a more affordable rent. 8 BIS Shrapnel Pty Limited 2016

13 Table 2: Tax benefit from negative gearing and required additional rent and yield to offset tax benefit, 2015 Median rent Current yield Tax Benefit Yield required to offset tax benefit $ per week % $ per week % Sydney $ % $73 4.5% Melbourne $ % $56 4.4% Brisbane $ % $32 4.9% Adelaide $ % $29 4.7% Perth $ % $30 5.3% Canberra $ % $28 5.2% Hobart $ % $10 5.3% Darwin $ % $20 5.2% Australia $ % $50 4.7% 3.2 Impact on rents and prices and development The higher yield required by an investor to compensate for curtailing negative gearing reduces the price the investor is willing to pay. At a required yield of 4.5% (compared to 3.9% with negative gearing), the justifiable price to an investor based on a current rent of $510 per week (or $26,520 per annum) is $591,000, which is 12.5% below the current median of $675,000. This difference varies across the capital cities and ranges from 3.4% in Hobart (where yields are already higher) to 13.2% in Melbourne. The correction in prices will be constrained by development costs. Developers need a minimum level of prices to underwrite development costs. Any decline in prices will cause new building to fall. Most development costs (i.e. construction, professional fees, etc.), are relatively fixed, and will continue to rise, and while there will be variability in the site purchase price, there are limits to which site values can fall to assist feasibility in a weaker market: below a certain value, the site becomes more attractive for existing or other uses i.e. industrial, commercial or retail price benchmarks set at the top of the market mean that many existing landholders are reluctant to sell until the previous benchmarks are again reached as the easier sites to develop are progressively absorbed, greater costs will be incurred in aggregating future sites to be suitable for development. This was evident in the Sydney market through the mid-2000s. Prices fell below development costs over 2004 to 2009 (Chart 1). Fewer sites became conducive to development and there was a collapse in new unit construction over the period as developers retreated from the market. A similar outcome is expected as the limitation on negative gearing causes underlying prices to again fall below the cost of development. New dwelling building will fall until dwelling prices, justified by the higher required yield, reach a level to facilitate profitable development. The decline in new supply will not only cause the rental market to tighten more quickly (and therefore rents to rise), but constrain an upturn. In the absence of negative gearing, rents will need to rise to a higher level in order to support the yields, and consequently the prices, to underwrite future rounds of new dwelling building. BIS Shrapnel Pty Limited

14 While lower prices might create more owner occupiers, there will still be less supply for growth in the rental market, as it will be the investors who need to underwrite pre-sales in apartment projects which banks will require before underwriting the finance. This will cause the rental market to tighten and lead to rising rents. There is some flexibility for site values to be suppressed to support development costs. However, the scope for reduction is limited, not only for the reasons outlined above, but also due to the impost of government charges, which influence the development equation. At the same time construction costs and professional fees will continue to rise. Consequently, a sufficient level of price growth is still required over time to meet at least the growth in these development costs. The forward impact is shown in Chart 1. While the limitation on negative gearing from July 2016 will reduce the overall justifiable price to investors by 12.5% in Sydney, and below the cost of development, the fall is not expected to play through in full. In reality, prices will be somewhat supported by owner occupiers and positively geared investors, while other owners would not look to sell at a loss. Nevertheless, disinterest from negatively geared investors will cause prices in a market without negative gearing (blue line) to fall below that otherwise expected in a market with negative gearing (purple line). It will also reduce pre-sales to investors until rental yields become more attractive. Even after allowing for stronger rental growth, rents aren t expected to reach a level to justify prices (black line) that sufficiently underwrite development costs until As a result, new dwelling building will be constrained, with only easy projects being able to proceed. This lack of supply will cause a sharper rises in rents to occur than under negative gearing with its higher level of construction. Stronger rental growth is also expected through to at least 2025 given the upturn into this period will start from a larger dwelling deficiency. The rise in rents will eventually see yields reach a point that generates price levels to underwrite the next round of construction. 10 BIS Shrapnel Pty Limited 2016

15 Chart 1: Unit prices with full and limited negative gearing vs development costs, Sydney $1,200,000 $1,000,000 Development cost less land Land Justifiable price - limited negative gearing Price forecast - full negative gearing Price forecast - limited negative gearing $800,000 $600,000 $400,000 $200,000 $ Consequently, while over the longer term unit prices with restricted negative gearing are estimated to be 6% below where they would otherwise be, rents will actually be 6% higher to facilitate the higher yield for investors (Table 3). Table 3: Median unit rent and price forecasts, Sydney As at June Rent full NG Prices before Yield before Rent after Prices after Yield after $ p.w. % ch $'000s % ch % $ p.w. % ch $'000s % ch % BIS Shrapnel Pty Limited

16 Chart 2: Median unit rents, prices and yields with full and limited negative gearing, Sydney $ per week 800 Rent before 750 Rent after as at June $ '000s Prices before Prices after as at June Per cent 4.9 Yield before 4.7 Yield after as at June Using the same underlying methodology, Table 4 summarises the total difference in rents and prices anticipated by 2025/26 across each of the capital cities, under full versus limited negative gearing. Rents are projected to be anywhere between 1.7% and 10% higher across the capital cities by 2025/26, when most of the impacts of the change to negative gearing have washed through. The impact on prices is projected to range from 1% to 6% lower. The impact on rents and values is a function of current yields and prices. A lower yield supported by negative gearing means a higher rent is required to compensate for the magnitude of the tax concession. A lower current unit price means that, for there to be significant building, prices cannot fall below development costs. The rising cost of development will put a floor under prices. It also means that rents will need to rise by a higher rate as prices will be less able to correct. These factors combine for Adelaide (+10%) and Melbourne (+9.6%) to experience the greatest projected increase in rent levels by Table 4: Projected median unit rent and price differential by 2026 with full and limited negative gearing by city Estimated unit rent 2026 ($ per week) Difference (%) Estimated unit price 2026 ($'000s) Difference (%) full NG limited NG full NG limited NG Sydney % % Melbourne % % Brisbane % % Adelaide % % Perth % % Canberra % % Hobart % % 12 BIS Shrapnel Pty Limited 2016

17 4. IMPACT ON RESIDENTIAL BUILDING ACTIVITY Dwelling construction plays a vital role in the Australian economy. It directly creates jobs, indirectly generates upstream business for a vast range of supplier industries (bricks, concrete etc.), and results in taxation revenue for both Federal and State governments. Any assessment of negative gearing therefore, should take account of the potential impact on future dwelling construction activity. This section compares our base residential building forecasts which act as a control scenario - with an alternative forecast that assumes the modification of negative gearing. 4.1 BIS Shrapnel s base case building outlook BIS Shrapnel s current baseline dwelling forecasts are shown in Table 5 8. Underlying demand, which runs off population growth and household formation, underpins the dwelling forecasts. When set against dwelling completions, we derive an estimate of the dwelling stock deficiency. Residential building commencements are estimated to have recorded a third consecutive year of double-digit percentage growth in 2014/15, taking new dwelling starts to a record high of over 210,000. Growth has been led by Victoria and New South Wales, with Queensland, Western Australia and the smaller states/territories also contributing. The current accommodative monetary policy stance unlocked supportive underlying fundamentals in these markets and underpinned a strong supply response, driven by significant demand from the investor segment of the market. The 2014/15 result will represent the peak in activity for this cycle, with new dwelling commencements set to decline in 2015/16. Interest rates will remain at accommodative levels but after such a strong level of activity the heat will begin to come out of key property markets and activity will begin to decline from its current unsustainable highs, particularly in the other dwellings sector. The decline in activity will then accelerate over 2016/17 and 2017/18, with New South Wales joining the rest of the country in seeing residential building weaken. Interest rates will begin to rise in late 2016, which will negatively impact on confidence, particularly in the investor segment of the market. Interest rates will begin to fall once more in 2018 in response to signs of economic weakness. This will arrest the decline in new dwelling starts and underpin another pick up in 2019/20, although with most markets being more balanced the upturn will not be as strong. This will leave new dwelling starts at 178,550 for 2019/20 and the annual average for the five years to 2020 at 179,220, an increase on the already strong previous five-year period of 4%. Beyond 2019/20 we are forecasting further strong results for residential building. New dwelling starts will be slightly softer over the five years to 2025 after experiencing such a strong period from 2016 to 2020, but will still average a solid 172,260 annually. 8 An overview of BIS Shrapnel s building forecasting methodology is attached in the Appendix. BIS Shrapnel Pty Limited

18 4.2 Investors and dwelling building Negative gearing operates via investors, who in turn have an important influence on the building cycle (Chart 3). Investor demand has rivalled occupier demand in driving the current dwelling cycle, with housing loans to investors up by a strong 22% in the year to April 2015 (MAT). The adverse effect on residential investor demand and building activity if negative gearing were to be limited, accordingly, is likely to be significant. Before 1 July 2016, as previously noted (page 9), an investor can offset the net loss ($10,000) against income from other sources (such as wages) before calculating tax payable - making for an expected tax saving of $3,500. After the change, however, the tax benefit (at least in the current year, against current wage income) would disappear for established dwellings, and the effect can be expressed as a rise in the investor s effective cost of funding: Table 5: Landlords average tax deduction Before After Mortgage interest $25,000 $25,000 less Tax saving -$3,500 -$0 = Effective funding cost $21,500 $25,000 (%) (4.3%) (5%) The average negative-geared investor s after-tax effective mortgage interest rate in the current year rises by an estimated 0.7% after the abolition of negative gearing provisions. By raising investors effective borrowing rates, the policy change (in the absence of any compensating increase in rents or fall in dwelling prices in the short term) will reduce effective after-tax yields and, other things equal, will dampen investor demand for housing. This will become evident in a slowdown in investor demand for home loans. Based on BIS Shrapnel analysis, we estimate that every 1% increase in the mortgage interest rate produces a 9.7% fall in investor loan approvals. And that this happens with a 4-quarter lag, on average. In round figures, therefore, based on the average after-tax effective mortgage interest rate rising by an average 0.7% after the abolition of negative gearing provisions, we would expect to see a 7% drop off in investor loan commitments, on account of the higher funding cost. We predict this will start to happen within 12 months. As investors pull back from the market, the number of dwelling commencements will soften. Investor loans are a strong driver of dwelling starts, as shown in Chart 3. What is the magnitude of this investor-to-dwelling-starts effect? Based on BIS Shrapnel analysis, we estimate that every 1% decline in investor loan approvals results in a 0.8% fall in dwelling commencements. And that this happens with a 4-quarter lag, on average. Therefore, based on investor loan approvals falling by 7% after the abolition of negative gearing provisions, we would expect to see roughly a 5% drop off in dwelling starts, compared to what would have happened under negative gearing. We predict this effect will start to impact dwelling approvals within 12 months, other things equal. Chart 4 summarises the logic. Removal of negative gearing would raise effective interest rates by an estimated 0.7%, leading to (lagged) 7% decline in investor new loan demand, resulting in a (phased) 5% fall in dwelling commencements (before subsequent adjustment effects eventually start to reverse the decline, in the medium term). 14 BIS Shrapnel Pty Limited 2016

19 Chart 3: Investor demand helps drive dwelling commencements 220,000 MAT No. Australia MAT $' ,000 55,000 MAT No. New South Wales MAT $'000 50, , , ,000 50,000 45, , , , ,000 90,000 45,000 40,000 40,000 35, , ,000 80,000 35,000 30, , , ,000 70,000 60,000 30,000 25,000 25,000 20,000 Total Dwelling Commencements (LHS) Lagged (1yr) Total Investor Loans (RHS) Source: ABS, BIS Shrapnel Total Dwelling Commencements (LHS) Lagged (1yr) Total Investor Loans (RHS) Source: ABS, BIS Shrapnel 65,000 MAT No. Victoria MAT $'000 30,000 45,000 MAT No. Queensland MAT $'000 18,000 60,000 28,000 43,000 41,000 17,000 55,000 50,000 45,000 40,000 26,000 24,000 22,000 20,000 18,000 39,000 37,000 35,000 33,000 31,000 29,000 27,000 25,000 16,000 15,000 14,000 13,000 12,000 Total Dwelling Commencements (LHS) Lagged (1yr) Total Investor Loans (RHS) Source: ABS, BIS Shrapnel Total Dwelling Commencements (LHS) Lagged (1yr) Total Investor Loans (RHS) Source: ABS, BIS Shrapnel 14,000 13,000 12,000 11,000 10,000 9,000 8,000 MAT No. South Australia MAT $'000 5,000 4,800 4,600 4,400 4,200 4,000 37,000 35,000 33,000 31,000 29,000 27,000 25,000 23,000 21,000 19,000 MAT No. Western Australia MAT $'000 14,000 13,000 12,000 11,000 10,000 9,000 8,000 7,000 3,800 17,000 7,000 Total Dwelling Commencements (LHS) Lagged (1yr) Total Investor Loans (RHS) Source: ABS, BIS Shrapnel Total Dwelling Commencements (LHS) Lagged (1yr) Total Investor Loans (RHS) Source: ABS, BIS Shrapnel BIS Shrapnel Pty Limited

20 Chart 4: Removing negative gearing will reduce dwelling commencements Effective interest rates Investor loans Dwelling commencements % $m no. time time time 4.3 Forecast dwelling commencements after the change initial phase The before and after dwelling forecasts below reflect these further considerations: the adverse effect on building activity will be worse for Attached Dwellings (especially high rise, 4 storeys and above) than for Detached Houses, because the lion s share of investor money goes into apartments the negative impact will be felt most strongly in Sydney and Melbourne, and to a lesser degree Perth and Brisbane, the markets where investors are concentrated owing to the normal lags from approval to commencement to completion, the slowdown will initially be camouflaged by momentum effects from approvals already in the system there could be a pull-forward effect as investors rush to get into the market prior to the policy change kicking in on 30 June 2016, leaving a vacuum in its wake after the 1st of July Limiting negative gearing would reduce dwelling supply. Our dwelling forecasts contrasting the Base and Alternative cases reveal that the adverse impact on dwelling commencements from the policy change will be especially marked in the first half decade (years 1-5), as rental yields drop from their former level. The variance will be worst in this period, when total commencements will reach 4.4% lower under the alternative case. Sydney (- 6.3%) and Melbourne (-5.4%) will see sharper variances, at their nadir, than the national average. Apartments (-9%) at their low point will drop by more than detached houses (-1%). Around 7,200 fewer new dwellings will be built at the point of greatest impact following the policy change. Over 5,000 of these will be in Sydney and Melbourne. And about 5,500 will be attached dwellings (apartments etc.). 4.4 Medium and long term dwelling forecasts In the later period of the ten year forecast horizon, rising rents (and/or lower dwelling prices) as outlined in Section 3 - can be expected to lead to higher yields, eventually bringing freshman investors back in and restoring dwelling building activity. New investors cash rental income needs to rise by enough to fully offset the tax loss incurred upon the removal of negative gearing benefits, and compensate for the loss of yield 9. 9 As noted earlier, although cash expenses would now be treated as a capital expense and accumulate to be deducted a number of years later against capital gains and/or income from the property, the resulting delay and uncertainty would have a serious deterrent effect on new investment for most Mum & Dad investors, who form the majority of landlords in Australia. 16 BIS Shrapnel Pty Limited 2016

21 How long this adjustment will take is a significant question? We have synchronised the phasing between rents/yields (Section 3) and dwelling building (Section 4). The requisite rise in rents/yields will not happen overnight, because the vast stock of existing rental properties will still enjoy negative gearing and so the supply of rental properties will be impacted only at the margin, initially. It will take time for the reduction in new dwelling building to start to incrementally affect vacancy rates, and hence rents. The adjustment process is predicted to take a decade to reach completion. After 2020, the effect of the policy gradually starts to diminish, as rental yields slowly normalise. Critically, the justifiable price under limited negative gearing will not exceed the development cost until 2022 (Chart 1), so developers are not expected to start restoring their pre-intervention desired building rate until Allowing for developer and approval lags, it will likely be another 1-2 years before dwelling building normalises back to the base case, by 2025/26 in other words. By Year 10, dwelling starts under the Alternative case are back to where they would have been under the Base case. Table 6: Impact on dwelling commencements Australia, number, 5 year average Base case Alternative case In the long run, with the passage of time, grandfathered negative gearing will eventually wash out of the system entirely, as mortgages are paid off, legacy properties are sold to fund retirement, investors pass away, and properties move from negative to positive net market returns. However, this will be a long and drawn out process extending out across the 10 year forecast horizon. 4.5 Impact on dwelling stock deficiency Five-year average 52,820 45,350 39,370 9,840 21, ,910 52,830 45,960 35,510 9,420 21, ,630 50,580 43,810 38,100 9,660 21, ,420 51,420 44,980 35,210 9,400 21, ,170 Variance number -2,240-1,540-1, ,490-1, ,460 Source: ABS, BIS Shrapnel BIS Shrapnel estimates the underlying demand for dwellings, based on household formation. This is then set against annual new dwelling completions to arrive at a measure of the dwelling stock deficiency. Our forecasts show the deficiency would be increased after the policy change 10. The policy change would substantially increase Australia s national home building shortfall by the end of the 10 year forecast horizon, of almost 60,000 unbuilt dwellings compared to just under 20,000 in the base case. 10 Forecasts assume the policy shift has no impact on the rate of household formation, with the effect on household formation of higher rents being roughly offset by lower dwelling prices, over the 10-year horizon. BIS Shrapnel Pty Limited

22 5. ECONOMIC GROWTH IMPACT 5.1 Impact on output Real GDP is forecast to shrink by a cumulative $190 billion over the 10 years, under the alternative scenario. The largest decline in output by industry sector (in value terms) will be in dwelling building, which is estimated to be lowered by $57 billion over the ten years. In turn, reduced output of dwelling building means a lower contribution to value-added by sectors which service or supply inputs to the residential building industry. We estimate that output (gross value added) in these sectors will contract by: business services will contract by - $49.4 billion manufacturing -$28.5 billion heavy and civil construction -$9.5 billion non-residential building -$7.6 billion utilities -$3.8 billion transport -$7.6 billion wholesale and retail trade -$7.6 billion household services -$5.7 billion oil and gas extraction -$5.7 billion other (mainly agriculture and non-ferrous metal ore mining) - $7.6 billion. We used the links from demand for housing to activity in other domestic industries using structural relationships embedded in input-output tables. Dwelling building has a spill over effect to other parts of the economy, therefore lower building will result in lower output of several related sectors including business services, manufacturing, heavy and civil construction, nonresidential building, utilities and retail and wholesale trade. To put the result in perspective, a contraction in real GDP of $19 billion per annum would shave about 1% off real national product, which is forecast to average $1,900 billion per year between 2017 and Under the alternative scenario, real GDP growth is forecast to average 2.9 per cent per annum, 0.2 percentage points lower than the base case of 3.1 % per annum. 5.2 Impact on jobs In terms of employment, we estimate that around 73,000 jobs will be lost in the dwelling building sector due to lower construction. However, the overall decline in employment (i.e. including falls in related sectors) is much larger: we estimate that around 175,000 fewer jobs will be created in the economy under the alternative scenario, about 1.4% of the total number of jobs (12,500,000). Unemployment on average will rise from 5.8% to 5.9%. 18 BIS Shrapnel Pty Limited 2016

23 6. IMPACT ON HOUSEHOLDS 6.1 Tenants and the Australian rental market The prevalence of rental occupancy in the Australian market has growth over the last decade and a half. Private rental dwellings, that is dwellings owned by private landlords, accounted for 27.2% of the total private dwelling stock in the five mainland states; up from 25.0% in 1996 (Chart 5). The highest percentage of renters are in Queensland (30.9%), followed by New South Wales (26.9%). High density dwellings (apartments in buildings four storeys and above) have the greatest percentage of renters, and therefore investor owners, at 59.2% of the total stock. Public rental dwelling stock fell from 5.3% of total dwellings in 1996, to 4.1% in Chart 5: Tenure by dwelling type and state, five mainland states 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% NSW VIC QLD SA WA 5 states Separate house Medium Density High density Total Owned Being purchased Rented Owned Being purchased Rented Source: Australian Bureau of Statistics 2011 Census Those most exposed to the rental market are in the youngest age adult categories, with 54.2% of households with a household reference person aged 20 to 34 years old in rental dwellings (Chart 6). This is up from 50.2% in 1996 and does not allow for the increased percentage of 20 to 34 year olds who are also remaining in the family home for a longer period. Chart 6: Tenure by age of household reference person five mainland states 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% years & over Owned Being purchased Rented Source: Australian Bureau of Statistics 2011 Census BIS Shrapnel Pty Limited

24 In the following age cohort, just under one third or 29% of 35 to 49-year-old households are renting. Importantly, only 20% of these households were renting at the 1996 Census, indicating an increased propensity of the youngest households to remain in rental as they age. Rental dwellings provide an affordable option into the housing market from those on lower incomes (Chart 7). Around 64% of rental households earn a household income less than $65,000 per annum, compared to 44% in owner occupied households. Chart 7: Household income by tenure type, five mainland states Percent (%) Private Rental Owner Occupier Negative Income Nil Income $1-$10,399 $10,400-$15,599 $15,600-$20,799 $20,800-$31,199 $31,200-$41,599 $41,600-$51,999 $52,000-$64,999 $65,000-$77,999 $78,000- $103,999 $104,000- $129,999 $130,000- $155,999 $156,000- $181,999 $182,000- $207,999 $208,000- $259,999 $260,000 or more Source: Australian Bureau of Statistics 2011 Census Chart 8: Type of household by tenure type, five mainland states Rental households also provide an affordable entry into the housing market for single person households (Chart 8). The lower level of household income of rental households is reflected in lone person households being the most prevalent occupants of rental dwellings. Around 39% of rental dwellings are occupied by a lone person, compared to 24% across the total dwelling stock. In contrast, family with children households are more prevalent across the whole housing stock. This demand from lone person tenants is most likely to be met by smaller dwellings such as apartments. 20 BIS Shrapnel Pty Limited 2016