IS THERE ANY LIFE IN MORTGAGE MARKETS IN SOUTH AFRICA? September - - PowerPoint PPT Presentation

IS THERE ANY LIFE IN MORTGAGE MARKETS IN SOUTH AFRICA?

September 2018

AGENDA

1.

Mortgage access

2.

Borrowers

3.

Properties

4.

Lenders

5.

Governance

6.

Next steps

2

SARB data provides a long term perspective on mortgage activity – denominated in Rands

13% 24% 18% 7% 11%

• 3%

18% 20% 26% 24% 32% 43% 27% 16%

• 50%
• 57%

23% 7% 6% 16% 37% 21%

• 8%

8%

• 60%
• 40%
• 20%

0% 20% 40% 60% 100 000 200 000 300 000 400 000 500 000 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Mortgage loans paid out during the period - R millions (LHS) Year on year growth rate (RHS) Average inflation

NEW MORTGAGES: South Africa Reserve Bank

(Million Rands) Million Rands

Source: SARB

282 636 140 039 159 184 164 917 156 249 162 050 163 605 164 432 153 702 153 477 200 000 400 000 600 000 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

It is helpful to explore the number of mortgages granted as this links back to households. Deeds data and NCR data align relatively well

NCR

130 398 274 292 176 089 113 097 195405 200 000 400 000 600 000 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 * Approved taken up

OFFICE OF DISCLOSURE*

MORTGAGE LOANS GRANTED – TOTAL: 2007 – 2017

(Number of loans)

429 887 260 060 125 857 144 049 149 269 144 192 150 350 150 379 150 538

247 620 160 033 86 288 106 274 116 538 117 097 123 344 126 708 129 447 119 250 113 490

200 000 400 000 600 000 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Mortgage Mortgage linked to a sale

LIGHTSTONE

All lenders, mortgages granted to consumers Banks, mortgages

granted to

consumers All lenders, mortgages on residential properties Data anomaly

374 184

BASA data on FSC lending is available to end 2016. It shows a clear decline between 2014 and 2016

7 262 6 045 6 319 5 775 2 946 4 329 6 042 7 325 7 328 7 758 10 343 10 074 7 579 3 000 6 000 9 000 12 000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

FSC MORTGAGE ORIGINATION: 2004 –2016

R Millions

57 324 53 159 43 721 55 287 25 147 24 643 27 708 29 981 26 911 24 897 30 909 29 060 21 464 20 000 40 000 60 000 80 000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

RAND VALUE NUMBER OF LOANS Number of loans

Source: BASA

R160 637 R323 112 R440 714 R443 171 R468 854 R457 690 R491 702 R 0.00 R 100 0.00 R 200 0.00 R 300 0.00 R 400 0.00 R 500 0.00 R 600 0.00 0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20% 2004/01 2005/01 2006/01 2007/01 2008/01 2009/01 2010/01 2011/01 2012/01 2013/01 2014/01 2015/01 2016/01 2017/01 RATE ON FSC MORTGAGE LOANS MAXIMUM FSC MORTGAGE VALUE

This is despite the adjustment of income thresholds and an accommodating interest rate environment

FSC: 2004 - 2008 AFFORDABLE MARKET: 2009 – DECEMBER 2017)

Upper income threshold R7 500 R7 900 R8 200 R9 080 R9 670 R15 142 R15 498 R15 738 R16 500 R17 600 R18 600 R20 000 R20 800 R22 000

DOMINANT RATE ON NEW MORTGAGES & MAXIMUM MORTGAGE VALUE

(2004 – December 2017) Interest rate Mortgage value Source: SARB, BASA, own calculations. Maximum mortgage value assumes 25% instalment to income rate. Rate on FSC mortgages assumed to be prime plus 200 basis points

There is no single definition of the ‘affordable’ or entry-level market, which makes comparisons and consolidation difficult

27 254 27 522 28 930 29 788 22 807 18 392 14 739 12 136 8 895 6 874

20 000 40 000 60 000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

25 426

20 000 40 000 60 000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

6 143 6 094 1 994 1 887 1 618 1 039 722 762 234 228 342

24 145 17 753 8 374 8 629 11 528 11 302 11 511 11 358 11 510 8 943 7 598 20 000 40 000 60 000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 R300K - R600K Under R300K

57 324 53 159 43 721 55 287 25 147 24 643 27 708 29 981 26 911 24 897 30 909 29 060 21 464

20 000 40 000 60 000 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

NCR (Borrower income under R15 000) CITYMARK

(Mortgages with a sale)

OFFICE OF DISCLOSURE

(Borrower income \ under R15 000)

BASA

MORTGAGE LOANS GRANTED – AFFORDABLE MARKET

(Number of loans)

The primary instrument to support mortgage lending is the FLISP

1 193 2 253 2 660 17 231 14 739 12 136 8 895 6 824 18 680 30 643 49 029

• 10 000

20 000 30 000 40 000 50 000 60 000 2014 2015 2016 2017 2018 2019 2020 Number of FLISPs Mortgages granted to FLISP eligible borrowers (<R15 000) Number of FLISPs: Forecast Linear (Mortgages granted to FLISP eligible borrowers (<R15 000))

FLISPS AND MORTGAGE LOANS GRANTED TO FLISP ELIGIBLE HOUSEHOLDS

Source: SANT Human Settlements budget vote 2018

1.

Mortgage access

2.

Borrowers

3.

Properties

4.

Lenders

5.

Governance

6.

Next steps

9

Bureau data for borrowers earning between R10 000 and R15 000 indicates high levels of credit market activity and relatively high levels of arrears

53% 50% 48% 16% 13% 13% 5% 1% 0% 20% 40% 60% Retail apparel Unsecured Credit card Retail General Vehicle finance Mortgage Retail furniture Financial other

Arrears 90+ days Good

OPEN CREDIT ACCOUNTS

(Borrowers with a personal income of R10,000 – R15,000)

% Of borrowers with this account type in arrears of 90 days+ 25% 34% % of borrowers in segments with this account type 17% 16% 29% 5% 3% 3%

34% of borrowers in this income segment have at least one account in arrears

• f three months or more

Source: XDS data, own calculations

While access to credit offers consumers a pathway to asset building in theory, it can create barriers in practice

Unacceptable credit track record

31% 29% 55%

Lack of affordability

17% 40% 36% 42%

Insufficient info/ docs provided

0% 2% 3%

Unacceptable security

81% 6% 3% 2%

Ineligible applicant

2% 1% 7% 2%

Unacceptable exposure (town)

8%

Adverse credit record

1% 10% 22%

Other/ blank data Not target market

REASONS FOR DECLINE: ALL HOME LOANS 2015

Source: Office of Disclosure, 2017

While access to credit offers consumers a pathway to asset building in theory, it can create barriers in practice

REASONS FOR DECLINE: ALL HOME LOANS 2017

Source: Office of Disclosure, 2018

50,69% 38,73% 24,79% 14,32% 8,59% 3,32% 1,52% 0% 20% 40% 60%

Other Lack of affordability Unacceptable Credit Track Record Unacceptable Security Adverse Credit Record / Unacceptable loan to value ratio Insufficient Information / Documents Having Been Provided Ineligible Applicant

Patterns of credit origination in the R10 000 – R15 000 segment reflect borrower priorities and constrained choices

CREDIT FACILITIES MORTGAGES UNSECURED CREDIT SECURED CREDIT

Source: NCR Consumer credit report. Mortgage 12.5%, 240 months, vehicle 14%, 60 months, unsecured 25%, 24 months, facility 25%, 12 months

VALUE NUMBER AVERAGE SIZE

R2.1 bn 5 338 R387 102 R11.6 bn 124 812 R93 011 R15.4 bn 564 160 R27 300 R5.8 bn 979 842 R5 970

CREDIT ORIGINATION FOR INDIVIDUALS EARNING BETWEEN R10 000 AND R15 000 MONTHLY

(2017)

R3,920 R2,115 R1,405

TOTAL MONTHY INSTL. R23 m R270 m R821 m R555 m

R476

There is a noticeable difference in performance of loans written since 2009. As at December 2012 ,7.7%

• f Affordable Market loans were 30 days or more outstanding versus 4.1% of Conventional Market loans

3,7% 1,2% 2,9%

6,4% 6,1% 6,9% 6,6% 7,3% 7,6% 7,7%

0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 201106 201109 201112 201203 201206 201209 201212 201303 201306 201309 201312 201403 201406 201409 201412 201503 201506 201509 201512

% of loans 30+ days

AGEING ANALYSIS OVER TIME (CUMULATIVE)

(Mortgages originated between 2009 and 2015, mortgages from big 4 banks) AFFORDABLE MARKET LOANS Total*: 106 141 loans

1,8% 0,7% 1,7%

3,7% 3,6% 3,9% 3,6% 3,9% 4,0% 4,1%

0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 201106 201109 201112 201203 201206 201209 201212 201303 201306 201309 201312 201403 201406 201409 201412 201503 201506 201509 201512

CONVENTIONAL MARKET LOANS Total*: 545 267 loans

Note: total loans open as at December 2015 30 – 60 days 60 – 90 days 90+ days 30 – 60 days 60 – 90 days 90+ days

At Dec 2015, 1.7% of Conventional Market loans were 90 days+, 0.7% 60 – 90 days and 1.8% were 30 – 60 days 30 – 60 days 60 – 90 days 90+ days

Of all the differences between portfolios, performance by loan size is most noteworthy. It highlights noticeably higher risk for smaller loans in the affordable market

0% 1% 2% 3% 4% 5% 6% 7% 201106 201108 201110 201112 201202 201204 201206 201208 201210 201212 201302 201304 201306 201308 201310 201312 201402 201404 201406 201408 201410 201412 201502 201504 201506 201508 201510 201512

% of loans 90+ days

0% 1% 2% 3% 4% 5% 6% 7% 201106 201108 201110 201112 201202 201204 201206 201208 201210 201212 201302 201304 201306 201308 201310 201312 201402 201404 201406 201408 201410 201412 201502 201504 201506 201508 201510 201512

AFFORDABLE MARKET LOANS NPLS BY OPENING BALANCE (Mortgages originated between 2009 and 2015, mortgages from big 4 banks)

R0 – R100,000

CONVENTIONAL MARKET LOANS

Total Number of loans in Dec 2015

2,782 8,133

Total Number of loans in Dec 2015

The gap in performance narrows with larger loans

0% 1% 2% 3% 4% 5% 6% 7% 201106 201108 201110 201112 201202 201204 201206 201208 201210 201212 201302 201304 201306 201308 201310 201312 201402 201404 201406 201408 201410 201412 201502 201504 201506 201508 201510 201512

% of loans 90+ days

0% 1% 2% 3% 4% 5% 6% 7% 201106 201108 201110 201112 201202 201204 201206 201208 201210 201212 201302 201304 201306 201308 201310 201312 201402 201404 201406 201408 201410 201412 201502 201504 201506 201508 201510 201512

AFFORDABLE MARKET LOANS NPLS BY OPENING BALANCE (Mortgages originated between 2009 and 2015, mortgages from big 4 banks)

R0 – R100,000

CONVENTIONAL MARKET LOANS

Total Number of loans in Dec 2015 Total Number of loans in Dec 2015

14,599 R100,000 – R200,000 17,176

The gap in performance narrows with larger loans

0% 1% 2% 3% 4% 5% 6% 7% 201106 201108 201110 201112 201202 201204 201206 201208 201210 201212 201302 201304 201306 201308 201310 201312 201402 201404 201406 201408 201410 201412 201502 201504 201506 201508 201510 201512

% of loans 90+ days

0% 1% 2% 3% 4% 5% 6% 7% 201106 201108 201110 201112 201202 201204 201206 201208 201210 201212 201302 201304 201306 201308 201310 201312 201402 201404 201406 201408 201410 201412 201502 201504 201506 201508 201510 201512

AFFORDABLE MARKET LOANS NPLS BY OPENING BALANCE (Mortgages originated between 2009 and 2015, mortgages from big 4 banks)

R0 – R100,000

CONVENTIONAL MARKET LOANS

Total Number of loans in Dec 2015 Total Number of loans in Dec 2015

R100,000 – R200,000 R200,00 – R300,000 37,282 25,194

There is no difference in performance between larger loans granted in the affordable segment with conventional market loans

0% 1% 2% 3% 4% 5% 6% 7% 201106 201108 201110 201112 201202 201204 201206 201208 201210 201212 201302 201304 201306 201308 201310 201312 201402 201404 201406 201408 201410 201412 201502 201504 201506 201508 201510 201512

% of loans 90+ days

0% 1% 2% 3% 4% 5% 6% 7% 201106 201108 201110 201112 201202 201204 201206 201208 201210 201212 201302 201304 201306 201308 201310 201312 201402 201404 201406 201408 201410 201412 201502 201504 201506 201508 201510 201512

AFFORDABLE MARKET LOANS NPLS BY OPENING BALANCE (Mortgages originated between 2009 and 2015, mortgages from big 4 banks)

R0 – R100,000

CONVENTIONAL MARKET LOANS

Total Number of loans in Dec 2015 Total Number of loans in Dec 2015

R100,000 – R200,000 R200,00 – R300,000 51,569 495,562 2,782 14,599 37,282 8,133 17,176 25,194 R300,000+

In the old days our analysis also explored performance by area. This varies significantly

19

2006 2006 2007 2007 2008 2008 2009 2009 2010 2010 2011 2011 TABLE VIEW 0.0% 0.0% 12.5% 42.5% 26.7% 25.3% MALVERN 0.0% 3.8% 10.8% 18.5% 18.7% 19.1% VREDENBURG 0.0% 0.0% 3.5% 13.6% 17.7% 18.5% WELLINGTON 0.0% 2.7% 7.0% 8.1% 12.0% 18.4% GRASSY PARK 0.0% 4.2% 6.7% 12.1% 16.8% 18.1% CAPE TOWN 16.9% 3.7% 7.7% 13.4% 15.4% 17.7% BOKSBURG 0.0% 1.3% 4.7% 10.4% 15.2% 17.4% GEORGE 0.0% 0.0% 0.0% 9.5% 11.2% 16.2% GERMISTON 0.0% 5.3% 5.8% 5.3% 11.3% 16.0% MITCHELLS PLAIN 0.0% 0.6% 6.2% 9.8% 17.6% 15.8% NPL by suburb: 10 worst performing areas (Bonds originated between 2004 and 2008, mortgages from big 4 banks) FSC loans 2006 2006 2007 2007 2008 2008 2009 2009 2010 2010 2011 2011 MITCHELLS PLAIN 2.1% 8.6% 6.7% 10.3% 16.4% 18.0% VREDENBURG 0.6% 0.9% 3.5% 5.5% 13.2% 14.6% BRITS 1.4% 1.5% 4.3% 8.2% 13.9% 13.6% BRAKPAN 1.7% 3.8% 8.1% 11.7% 12.9% 13.3% KRUGERSDORP 2.5% 3.0% 6.4% 11.9% 13.3% 13.1% GRASSY PARK 1.6% 3.8% 5.8% 8.6% 12.6% 12.9% PORTLANDS 1.3% 4.4% 5.7% 13.1% 13.2% 12.7% GERMISTON 0.2% 2.8% 7.2% 10.4% 12.8% 12.6% BELHAR 2.2% 3.0% 7.7% 15.6% 13.2% 11.9% KEMPTON PARK 1.2% 3.5% 6.1% 10.7% 11.7% 11.9%

Source: Deeds office data sourced from the ALHDC and XDS

Non-FSC loans

In summary: Mortgages operate within a very well developed (but not always effectively regulated) credit sector Borrowing patterns and performance on other credit products can enable and curtail mortgage access Underlying supply chains – for vehicles and clothing – seem to be better at delivery than housing Loans that have been granted perform relatively well. However we have no data on triggers of default (death, disease, divorce, dispute, dismissal, degeneration, good old delinquency etc.). We are therefore unable to support vulnerable households appropriately Performance within the portfolio differs significantly by sub-market indicating scope for growth without necessarily taking on more risky borrowers What will it take to shape perceptions and expectations regarding housing journeys? What will it take to shape take-up of credit in favour of mortgages? How do we support borrowers better where there is willingness to pay?

1.

Mortgage access

2.

Borrowers

3.

Properties

4.

Lenders

5.

Governance

6.

Next steps

21

25 000 50 000 75 000 100 000 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Less than R300 000 R300 000 - R600 000 R600 000 - R1.2m More than R1.2m

Lenders have been at pains to point out the limited delivery of affordable stock. We have data on new build for the market as a whole. Curiously, we have no regularly reported administrative data on price points, despite the fact that it is collected

76 661 70 058 56 947 40 679 40 507 42 978 41 485 38 043 39666 41 527 39 014 25 000 50 000 75 000 100 000 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 STATS SA COMPLETED UNITS (RESIDENTIAL BUILDINGS) PRIMARY REGISTRATIONS BY VALUE: NON-SUBSIDY (CityMark data) Number of units Number of units 80 239 78 568 26 903 32 424 38 651 44 698 50 622 52 632 50 670 53277 25 000 50 000 75 000 100 000 2007/08 2008/09 2009/10 2010/11 2011/12 2012/13 2013/14 2014/15 2015/16 2016/17 Number of units NHBRC NON-SUBSIDY ENROLMENTS (HOUSING STARTS) 37 131 39 391 40 919 40 132 46 358 33 244 32 562 29 073 29 654 55 459 78 566

9 818 1 289

At the end of 2017, there were 1,888,979 RDP and BNG houses on the deeds registry: 30% of the total residential property market in SA

23 Source: CityMark

But this is to ignore existing housing stock. There are almost 1.9 million RDP houses on the deeds registry. While not all are mortgageable (because of informal transactions, intestacy and poor governance / limited compliance with by-laws), many are

Churn rates increase with values. The proportion of transactions that are bonded is limited in lowest value bands

11 621 27 149 59 984 64 157 1 363 13 623 34 958 36 684

10 000 20 000 30 000 40 000 50 000 60 000 70 000 Under R300 000 R300 000 - R600 000 R600 000 - R1.2m More than R1.2m Repeat transactions Bonded repeat transactions REPEAT TRANSACTIONS AND BONDED TRANSACTIONS: 2017

0.5% 1.8% 4% 5.5% Churn rate 11.7% 50.2% 58.3% 57.2% Bonded %

2 216 1 484 1 512 1 167

Source: CityMark

In summary: It would be useful to have better data on new build. It exists and should be shared That new stock numbers are disappointing does not mean there is nothing to mortgage – even if there has been a dilution of mortgageability What will it take to shape perceptions and expectations regarding housing markets? What will it take to shape provision of credit in favour of mortgages, especially in entry level resale markets?

1.

Mortgage access

2.

Borrowers

3.

Properties

4.

Lenders

5.

Governance

6.

Next steps

26

To explore mortgage profitability in the affordable and conventional market, we built a discounted cash flow model of a mortgage at the point of sale to understand profitability and assess the relative magnitude of value drivers

INPUTS

(revenue and expense drivers)

MODEL

(combines inputs in cash flow model)

OUTPUTS

(findings) Loan Characteristics Fee Income NPV of Cash Flows Fee Expense Return Bad Debts Loss Severity Prepayment Rates

Financial projection model

26 28 31 10 20 30 40

Risk discount rate : 8.4%

19 17.5 16.5

• 8

NPV

• 8

Inception End Year 1 End Year 2 End Year 3

• 10

R146

50 60

The model uses six segments – three in the affordable and three in the conventional market

39, 39,3% 3% 36,9%

• 5,3%
• 0,02%

7,7% 35, 35,2% 2% 36,1%

• 5,2%
• 0,07%

4,4% 32, 32,4% 4% 34,2%

• 4,9%
• 0,02%

3,2% 30, 30,4% 4% 30,6%

• 5,0%
• 0,02%

4,8% 21, 21,6% 6% 23,8%

• 4,7%

2,6% 14, 14,5% 5% 17,0%

• 4,5%
• 0,00%

1,9%

Return per R1 loaned

Breakdown:

AFFORDABLE LOW AFFORDABLE MEDIUM AFFORDABLE HIGH CONVENTIONAL MEDIUM CONVENTIONAL LOW CONVENTIONAL HIGH

Loan size R110,000 R240,000 R380,000 R210,000 R680,000 R1.8 million Term 20 years 20 years 20 years 20 years 20 years 20 years LTV 90% 90% 90% 85% 85% 85% Interest rate 12.5% 12.5% 12.5% 11.5% 10.5% 9.5%

• 0.00%

Interest profit/ loss Equity cost Loss given default Fee profit/ loss

The model output generates a ROI, broken down into key value drivers. Given the assumptions about the loan characteristics as well as costs and performance, it appears that higher interest margins and lower probability of settling early more than compensate for higher default

52, 52,8% 8% 50,1%

• 6,6%

9,3% 55, 55,6% 6% 56,9%

• 7,5%

6,1%

• 36,

36,1% 1%

• 7,9%
• 4,2%
• 9,5%
• 14,5%

14, 14,2% 2% 16,0%

• 5,3%

3,4% 48, 48,8% 8% 48,0%

• 6,2%

7,0% 8, 8,5% 5% 5,9%

• 1,1%

3,6% 27, 27,0% 0% 24,9%

• 3,9%

5,9% 42, 42,1% 1% 40,0%

• 5,6%

7,8%

Return per R1 loaned

Interest profit/ loss Equity cost Loss given default Fee profit/ loss Breakdown:

PERFECT PAYER ERRATIC PAYER DEFAULT MONTH 12 DEFAULT MONTH 120 DEFAULT MONTH 60 PREPAYMENT MONTH 12 PREPAYMENT MONTH 120 PREPAYMENT MONTH 60 DEFAULTERS (RESULTING IN REPOSSESSION) EARLY RE-PAYERS

While the primary measure of performance that is typically quoted is the proportion of loans that are 90 days or more in arrears, it obviously makes a big difference when in the life of the loan the default occurs. Likewise with early repayment as illustrated below for a R110 000 affordable loan

20,1% 4,2% 1,1%

• 2,5%
• 10,6%

CHANGE IN ROI: AFFORDABLE LOW SEGMENT

10% INCREASE IN INTEREST RATE 10% INCREASE IN DEPOSIT FUNDING 10% INCREASE IN LTV 10% INCREASE IN RECOVERIES ON DEFAULT 10% INCREASE IN THE COST OF FUNDING

% change in ROI

Interest rate increase from 12.5% to 13.75% Deposit to debt funding ratio changes from 70%:30% to 77%:23% Recoveries on default increases from 60% to 66% LTV increases from 90% to 99% The cost of deposit funding, debt funding and equity increase by 10%

While cost of funding is critical, the interest margin impacts significantly on profitability. Why don’t lenders charge more?

Mortgage lending rates are not much higher than risk free government bonds. Margins are thin

9,27 8,73

6 8 10 12 14 16 18

2004/01 2005/01 2006/01 2007/01 2008/01 2009/01 2010/01 2011/01 2012/01 2013/01 2014/01 2015/01 2016/01 2017/01

Yield on loan stock traded on the stock exchange: Government bonds - 10 years and over Predominant rate on new mortgage loans: Banks - dwelling units (home mortgage rate)

MORTGAGE LENDING RATES VS LONG TERM RISK FREE RATES

Percent While there is no comprehensive data on the distribution of rates charged by banks on all mortgages they originate, based on discussions with lenders, it appears that loans tend to be written in the narrow range between Prime less 100 or 150 basis points and Prime plus 200 or on the rare occasion 300 basis points. This is far lower than the current regulatory maximum rate on a mortgage of 12 percentage points above the repo rate (the repo rate is typically 3.5 percentage points below prime rate, and the maximum rate on mortgages would therefore be prime plus 800 basis points).

Source: SARB

How do lenders decide how to price mortgages, and when to ration or price for risk?

33

Mortgages, adjacencies & economies of scope Rationing and reputational risk Regulatory uncertainty

• Do banks price the

product as a standalone, or do they look at customer profitability (ie price the mortgage to get the credit card, VAF account and bank account)?

• Are there some

risks that are simply not worth taking?

• How do you evict

a household on the lowest rung of the housing ladder?

• Given the long

term nature of mortgage lending, uncertainty with regard to credit market regulation, not to mention all

• ther regulations,

will impact on lender incentives

Other?

• FSC – target driven

number chasing or real transformation?

In summary: Default is an important, but not the only, driver of value for lenders However, lenders can control (select) for it - which is why they focus on it Based on model assumptions, where there is an additional premium, lenders can more than compensate for higher default risk Mortgage profitability, measured in terms of expected profits on newly originated loans, does not look that good overall What shapes the way lenders price / ration mortgage credit? What do we want them to do?

1.

Mortgage access

2.

Borrowers

3.

Properties

4.

Lenders

5.

Governance

6.

Next steps

35

In many parts of the country, there is no state monopoly on violence and effectively no governance. Can lenders operate there?

MURDERS: 2016 MURDERS: WORST 10 PRECINCTS: 2017 Precinct Province Murders: 2017 Nyanga Western Cape 308 Umlazi Kwazulu-Natal 223 Philippi East Western Cape 205 Inanda Kwazulu-Natal 203 Delft Western Cape 195 Khayelitsha Western Cape 192 Kraaifontein Western Cape 186 Gugulethu Western Cape 182 Mthatha Eastern Cape 160 Mfuleni Western Cape 157 MURDERS: 2017 DISTRIBUTION OF RESIDENTIAL PROPERTIES BY MARKET SEGMENT: 2017

Source: CAHF CityMark data, Crime statistics from https://crimestatssa.com/

Some properties in Delft are advertised on Gumtree as cash only sales. One agent mentioned the reason for this is because some homes in the area are not in adequate condition to qualify for mortgage loans

37

Why is this house advertised as a cash sale? Agent: With some of these homes they don’t have proper ceilings, so then there is no use sending a bond buyer to look because the bank has certain criteria, a check list…when we go evaluate the house, we can immediately see when it won’t qualify for a bond.

• Agent, NMKM Properties Group

Why is this house advertised as a cash sale? Agent: Some people are in a hurry to leave…to go to the Eastern Cape, they don’t want to wait for the bond

• process. For others they also know that the bank

doesn’t give bonds for some homes so they have to take cash…

• Miriam, arranges viewings on behalf of an agent

Telephonic interviews with agents

FORMAL INFORMAL

DIRECT COST: BUYER Conveyancing fee: R7 382.40* None DIRECT COST: SELLER Rates arrears: ?? Certificates: R1 500 Certified copies of title deed (if required) None ACCESS CONSTRAINTS Clean title deed Pre-emptive clause Other documents Search costs Travel costs None TIME Uncertain: up to ten months, maybe more than a year Immediate RECOURSE AND PROTECTION Complaints process: submit a written complaint to the Cape Law Society Social sanction

38 38

Remind me again, why should I transact formally?

q There are very many reasons to sell a low-cost house informally: we must make the reasons to sell a house formally relevant to low value transactions. q Slow response times from regulatory bodies (the City, the Province, the Deeds

• ffice)

q Limited trust in formal processes and limited awareness of how they protect buyers and sellers q High transaction costs q As a consequence there is very limited consumer protection and very high risk associated with property transactions

HOUSE FOR SALE: R140 000

* Note: on this transaction there was a mortgage of R84 000. Bond attorney costs were R7 123.60

FLISP has not been implemented well. It is typically applied after the fact, and is not suited to secondary market transactions

1. OFTEN DOES NOT SUPPORT AFFORDABILITY (Revised FLISP?)

• 2. LENGTHY PROCESS, POORLY ALIGNED WITH MARKET NEED

(A case study: Cape Town)

Pre-approval from Bank Signed offer to purchase 7 September 2016 Mortgage application 7 September 2016 Payment to seller (bridging finance) 9 September 2016 Vacant possession by buyer 10 September 2016 Mortgage approval letter 22 September 2016 Delivery of subsidy application forms 13 October 2016 Letter of undertaking: Province 16 March 2017 Electrical certificate 24 April 2017 Documents lodged at deeds registry 26 May 2017 Mortgage payment received 26 May 2017 FLISP payment received 28 June 2017

Sept 2016 Oct 2016 Mar 2017 Apr 2017 May 2017 June 2017

40 Source: City of Cape Town Valuation Roll (2015); pilot survey by TSC community co-ordinator

Two neighbours, two very different situations…

‘Expropriation’ case Registered owner living in property

The The Trans nsaction n Suppo upport Cen entre e is a pi pilot, action-re researc rch initiative in in Khayelit litsha, in in the Desmond Tutu Sport & Recreatio ion Cent ntre The TSC provides free, hands-on assistance and advice for individuals looking to buy and sell residential properties through EFFICIENT, SAFE & LEGAL processes. At the same time, the TSC documents the progress of transactions, to highlight potential policy, legislative and administrative issues for attention.

The Transaction Support Centre is NOT an estate agent or a conveyancer (lawyer). We do not:

• 1. Earn any commissions on sales conducted

through the TSC

• 2. Charge any fees for our advice or assistance

The TSC connects buyers & sellers to trusted lawyers, formal estate agents and other service providers who can help them do things the right way. In this way, the TSC supports effective, legal transaction processes, and works to overcome the attraction (and practice) of informal sales.

HOW WE HELP BUYERS

Understand affordability by assessing: 1. Credit status 2. FLISP subsidy eligibility 3. Loan eligibility 4. Access to savings / other capital Find a house to buy We connect buyers & sellers by directing them to registered estate agents, online property websites, and local networks Sign offer to purchase We help buyers understand the offer to purchase and the other legal contracts and forms they need to sign, as well as the role of the lawyers involved Apply for a bond (mortgage loan) We help guide buyers through the mortgage application process (if applicable) Apply for FLISP subsidy We help guide buyers through the subsidy application process (if applicable) Transfer ownership We support buyers up until the sale is finalised and ensure that they have a legal title to their new home

HOW WE HELP SELLERS

Get the home ‘sale ready’: 1. We help make sure the title deed is in the sellers name 2. Get rates clearance certificate 3. Get electrical compliance certificate Find a buyer We connect buyers & sellers by directing them to registered estate agents, online property websites and local networks Sign offer to purchase We help sellers understand the offer to purchase and the other legal contracts and forms they need to sign, as well as the role of the lawyers involved Transfer the property We support the seller up until the sale is finalised and ensure that the seller receives payment when the property is transferred

VISIT US:

Desmond Tutu Sport & Recreational Hall, Paul Ave, Khayelitsha Monday - Saturday

CONTACT US:

061 863 7169 / 081 783 0281 http://housingfinanceafrica.org/projects/transaction-support-centre/

In summary: Administrative barriers are significant and impact on choices Governance is a real issue Housing and good process is contagious. So is bad process Mortgage lenders, municipalities and property owners have aligned incentives What shapes the way lenders price / ration mortgage credit? What do we want them to do?

1.

Mortgage access

2.

Borrowers

3.

Properties

4.

Lenders

5.

Governance

6.

Next steps

48

Detail

Causes of default

§

Explore SARS data on labour market churn to understand this better

§

Monitor governance: what are the KPIs?

§

What other local area dynamics are critical? Candidate areas for greenlining

§

What are the characteristics of areas that could absorb sustainable lending? Property market performance

§

Existing stock including RDP house

§

New build

• Engage at the most

senior level with lenders

• Position housing clearly

as the focus of financial inclusion and transformation

• bjectives
• Share greenlining

strategy

• Identify key barriers that

limit lender involvement and develop plan to mitigate risks

• Develop mechanisms to

support borrowers in line with primary causes of default

• Support and streamline

the transactions process

• Identify key city-level

interventions in areas targeted for greenlining

• This includes

condonation of building activity, arrears management, more accessible city services to enable compliance

• Ensure rapid subsidy

approval and payout

• Efficient and clear

application systems

DATA GAPS LENDERS BORROWERS CITY LEVEL SUPPORT EFFICIENT SUBSIDY ACCESS

What would it take to expand mortgage access?