un habitat guide to housing finance strategy

8/7/2019 UN Habitat Guide to Housing Finance Strategy

1/121

HS/ISBN: (series)

ISBN:

UNITED NATIONS HUMAN SETTLEMENTS PROGRMMEP.O. Box 30030, GPO Nairobi 00100, KENYATelephone: 254-20-7623120;

Fax: 254-20-7624266/7 (Central Ofce)E-mail: [email protected]

Website:http//www.unhabitat.org

Printed in Nairobi

AGUIDET

OPREPARINGAHOUSINGFINANCESTRATE

GY

GUIDE TO PREPARING A

HOUSING FINANCE STRATEGY

A Housing Finance Strategy is a plan or deploying the resources available(and i needed increasing them) to inance the demand or housing bydierent segments o society. A Housing Finance Strategy is inluenced bynational, regional and sub-regional agendas and is linked to a wide rangeo other strategies and actors.

This guide sets out the key elements o a housing inance strategy andthe process o developing a housing inance strategy, including the studyo housing demand, identiying available resources, market orces andstrategies to accomplishing housing inance goa ls and objectives. It providesassistance or member states on how to prepare successul housing inancestrategies and action plans. It guides member states in their uture housing

inance planning and better positions themselves to meet market potentialsand address the gap between demand and supply in housing inance.

THE HUMAN SETTLEMENTS FINANCING

TOOLS AND BEST PRACTICES SERIES


2/121

GUIDE TO PREPARING A HOUSING

FINANCE STRATEGY

Nairobi, 2009


3/121

Te Human Settlements Financing ools and Best Practices Series

Guide to Preparing a Housing Finance Strategy

First published in Nairobi in 2009 by UN-HABIA

Copyright United Nations Human Settlements Programme Nairobi, 2008

HS/ 1070/08E

ISBN: 978-92-1-132022-0 (series)

ISBN: 978-92-1-132064-0

Disclamer

Te designations employed and the presentation o the material in this publication do notimply the expression o any opinion whatsoever on the part o the Secretariat o the United

Nations concerning the legal status o any country, territory, city or area or o its authorities, orconcerning the delimitation o its rontiers o boundaries.

Views expressed in this publication do not necessarily reect the views o the United NationsHuman Settlements Programme, the United Nations, or its Member States.

Excerpts may be reproduced without authorization, on condition that the source is indicated.

Cover photo: Xing Zhang/UN-HABIA

Acknowledgements

Director: Oyebanji Oyeyinka

Principal Editor and Manager: Xing Quan Zhang

Principal Author: Raymond Struyk

Contributors: yler Yang, Marisol Ravicz, Nadezhda Kosareva, StigJensson and John Webber

English Editor: Ingrid Uys

Design and Layout: Anne Musotsi


4/121

And or this, we

need innovative

governance, and

local thinking and

reporting i we areto bring hope to

the urban poor.

Equally impor-

tantly, we need to

support our towns

and cities, indeed

our countries, to adopt pro-poor policies and

strategies that will obviate the need or urther

slum creation.

It is against this background, that the Human

Settlements Financing ools and Best

Practices series ocuses on the development

o know-how, knowledge and tools in human

settlements nancing, rom which Member

States can learn in delivering aordable hous-

ing to the poor.

Dr. Anna ibaijukaUnder-Secretary-General andExecutive DirectorUN-HABIA, Nairobi, 2008

Te global housing crisis, especially in the

developing world, is getting worse by the day

making the right to adequate shelter a quest

that is becoming more and more difcult to

meet, despite the targets set by the MillenniumDevelopment Goals.

Such is the rate o urbanization the inux o

people into towns and cities, and their natural

growth that the world has now reached a

point where or the rst time now, hal the

global population lives in towns and cities.

By the year 2050, six billion people two-thirds o humanity will be living in towns

and cities. And as urban centres grow, the locus

o global poverty is moving into towns and

cities, especially into the burgeoning inormal

settlements and slums, o the developing

world. In the developing world, this is hap-

pening so ast that slums are mushrooming in

what is termed the urbanization o poverty.

Tis makes it imperative that we use every

means at our disposal to ensure that we at

UN-HABIA, and our partners, keep ap-

plying ourselves to arget 11 o the Goals to

achieve signicant improvement in the lives o

at least 100 million slum dwellers, by 2020.

FOREWORD


5/121

TablE OF COnTEnTs

FOREWORD I

LIST OF TABLES, FIGURES AND BOXES IV

PART I CONCEPT AND CONTENT 1

CHAPTER 1 WHY A STRATEGY? 3

Vritio i the Depth o Houig Fice Deveopmet 3

Wht Doe low Voume o Houig Fice Me? 6

Orgizig Houig strtegy 6

The strtegy Mut e Compreheive 7

How Wi the strtegy e Fufed? 9

Prior Coditio or Preprig strtegy 11

Thee Guideie 12

CHAPTER 2 WHAT A STRATEGY DOES 13

a Iutrtio 13

leo 16

CHAPTER 3 STRATEGY OVERVIEW 17

Techic Eemet 17

The Poitic Proce 18


6/121

PART II TECHNICAL ANALYSIS 19

CHAPTER 4 THE DEMAND FOR HOUSING FINANCE 21

Preprig Demd Etimte 21

Cotructio Period Fice 22

acce to Houig Fice 25

CHAPTER 5 HOUSING FINANCE SUPPLY 29

The big Picture 29

Wht to ler aout Ech Type o ledig 31

addreig leder Efciecy d Rik Mgemet 34

Fudig source or leder 39

Govermet Progrm d Iititive 44

CHAPTER 6 OPTIONS FOR ADDRESSING THE GAP 55Icreig the Voume o Houig ledig 56

Icreig the Fud avie or Houig ledig 59

PART III ORGANIZING AND IMPLEMENTING A STRATEGY

DEVELOPMENT 61

CHAPTER 7 INSTITUTIONAL FRAMEWORK WHO DOES WHAT 63

Who shoud led the strtegy Preprtio? 63

Who shoud Prticipte? 64

How shoud the Techic ayi e Orgized? 65

Who shoud Py or the strtegy Deveopmet? 66

How to Determie the Recommedtio? 67

Defig the strtegy 68


7/121

v

CHAPTER 8 THE ACTION PLAN 71

The P Competed 71

Mgig Impemettio 72

CHAPTER 9 MONITORING RESULTS 75

Demd or Houehod side 75

suppy side 77

Who Prepre the ayi d Report 78

Reviewig Reut 79

REFERENCES 81

ANNEXES 85

a. suppemet Exhiit 85

b. Etimtig the Demd or Houig Fice 103

LIST OF TABLES

1.1 Exmpe o Mortgge Fice Ihiitor d Poie

Govermet step to Reove Them

2.1 Houe acquiitio i Dr e sm i 1990

2.2 source o Fud or Dweig Cotructio

4.1 shre o Houehod Uig borrowig Fud i the Pt 12 Moth

5.1 Poie atertive Ficig source y Houig Ivetmet Type

5.2 Iutrtive Type o Iormtio to Coect o leder Requiremet orHouig lo to Idividu

5.3 net Iteret Mrgi i Itertio Comprio

5.4 seected Rik Mgemet Optio

5.5 Optio or attrctig Fud rom Cpit Mrket

or Mortgge ledig


8/121

v

5.6 Rtig o Govermet Itervetio to Icree Homeowerhip

d attimet o Good Quity Houig

5.7 Poie Govermet Houig Fice Mrket Deveopmet actio

6.1 Chrcteriztio o Houig Fice Gp i Egypt y Teure-loctioGroup d source o Fice

6.2 seected Chege to Icreig Houig ledig

Voume d Poie Repoe

6.3 Ide or attrctig Whoee source o Fice

6.4 Exmpe o actio i Houig Fice strtegy

7.1 Exmpe o actio i Houig Fice strtegy

8.1 actio P Exmpe

9.1 Key Iormtio Item or Houehod survey

a.1 2007 Govermet Updte o seected Eemet o

the 2003 Tzi-Un Hitt P o actio

a.2 summry o Rui Federtio aorde Houig Mrket Iititivesttu o legitive Propo, Jury 2005

b.1 Iutrtive Preettio o au Houig Fice Demd i Ciro,2006

lIsT OF FIGUREs

1.1 Rtio o Outtdig Mortgge Det to GDP or seected Coutrie

1.2 Mortgge Depth v. Tot Credit Depth

1.3 Mortgge Depth v. ledig Efciecy

1.4 The Proce o strtegy Deveopmet

5.1 atertive or Commerci bk to Work i the Microfce Mrket


9/121

v

lIsT OF bOXEs

2.1 Outie o the Tzi Houig Fice strtegy

4.1 Kowedge o the Mortgge lo i Ciro

4.2 Moie bkig Reiveted

5.1 ledig Iteret Rte Mke-up o lti americ MFI

5.2 Pricip Rik i Dweig Purche or Progreive Cotructio Fice

5.3 The Chie approch to Houig suidie

b.1 aeig Demd or Houig d Houig Fice: a Exmpe orIdoei, y Mrj Hoek-smit

b.2 aeig Demd or Houig d Houig Fice: a Exmpe orMocow d budpet, y Jze Heged et .

lIsT OF abbREVIaTIOns

ba bker aocitio

Cb Credit ureu

CGaP The Couttive Group to ait the Pooret [a MicrofceProgrm]

EsFE Europe Fud or southeter Europe

GOT Govermet o Tzi

MFI Micro Fice Ititutio

nGO no Govermet Orgiztio

PTY pymet-to-icome rtio

sME m d medium eterprie

VRM Vrie iteret rte mortgge


10/121

PaRT I

COnCEPT anD COnTEnT


11/121


12/121

WHY a sTRaTEGY?

Te cold reality is that in many countries

amilies who want to construct, purchase, or

improve their housing situation, are inhibited

because they cannot borrow unds. Loans thatare available are channeled to middle- and

higher-income amilies. In these countries

most amilies, especially the poor, improve

their housing only as their savings will permit

as they are unable to leverage their incomes

through loans.

Te eect o this situation, combined with

extremely low income levels, are amiliesin cities and in the countryside who live in

extremely difcult conditions characterised

by poor quality dwellings and an absence o

water and sewerage services that much o the

world takes or granted. Te rapid growth o

cities and the increasing urbanisation in the

developing world are well documented, and

it makes clear this situation can only become

worse in the years ahead unless it is addressed

(Okpala et al. 2005). In the ollowing ew

pages, the current global housing nance situ-

ation is briey highlighted and a case is made

or the use o a housing nance strategy to

overcome this severe problem.

VaRIaTIOns In THEDEPTH OF HOUsInGFInanCE DEVElOPMEnT

How restricted is the availability o housing

nance in some countries? Characterising di-

erences among countries in the development

o their housing nance systems is a challenge

because o the limitations on data and market

ragmentation and segmentation. As discussed

urther below, lending or housing in some

transition countries and many developing

countries is segmented primarily between or-mal lending - mortgage lending similar to that

in industrialised countries - and micro lending

that supports the incremental construction

and improvement o dwellings or occupancy

by households who either cannot qualiy or

or cannot access ormal mortgage loans.

Unortunately, comprehensive data on

national housing nance development are

completely lacking in many countries. Te

comparative data available are or ormal

lending, particularly the ratio o the outstand-

ing volume o mortgage loans to a countrys

GDP.

01


13/121

Te ratio is reported or an array o countries

in Figure 1.1, using data generally around

2004. It is a rough but serviceable indicator.It is rough because it excludes: micro nance

or all kinds o housing;1 a good deal o the

nancing or rental housing which is oten

equity nanced; and, government nance or

housing unless it takes the orm o low interest

rate mortgage loans.

Te broad pattern shown in Figure 1.1 is

instructive, with expected dramatically higher

values or industrialised countries. Malaysiasratio tops 20 percent in large part due to such

lending being a long-term government prior-

ity maniest in the operation o its secondary

acility, Cagamas, and other actions.

While many countries have ratios in the 10-

15 percent range, there is a rough negative

relationship between per capita income and

mortgage market depth.

Surprisingly, the depth o mortgage lending

is not closely related to either a countrys

nancial depth or to its nancial efciency.

Tis is important because it reects the reality

that mortgage market development does not

simply ollow broad nancial market develop-

ment. Tis point is explored urther in the

ollowing paragraphs.

Figure 1.2 details mortgage lending and total

lending as a share o GDP or 11 countries

on our continents. Four o the countries that

have mortgage lending equal to 3 to 5 percent

The distortion rom exclusion o this source may be modest. For example,

data or South Arica, which has a vibrant micronance sector, show

that micro lenders in 2006 accounted or only about 2 percent o

outstanding credits. Separate data or housing loan are not available

but micro lenders are unlikely to have no more than a 2 percent share

o such loans. Honohan, P., and Beck, T. (2007). Making Finance Work

or Arica. Washington, DC: World Bank.

o GDP have approximately the same total

credit to GDP ratio as our countries or whom

mortgage lending is equal to 10 to 14 percento GDP. Further, in Brazil, where mortgage

lending is equal to 5 percent o GDP, total

credit equals almost 100 percent o GDP. By

contrast, in Jordan the mortgage to GDP ratio

is more than double that o Brazil - 11 percent

in Jordan, but total credit to GDP is slightly

below Brazils level - 92 percent in Jordan.

Similarly, Figure 1.3 shows that the depth o

a countrys mortgage market is not stronglyrelated to its nancial market efciency (as

measured by the spread between the bank-

ing systems lending and borrowing rates).

Te three countries with an approximately 4

percent mortgage to GDP ratio have almost

the same lending rate spread as our countries

with a 10 to 14 percent mortgage to GDP

ratio. Further, Croatia and China have almost

the same mortgage to GDP ratio but Chinaslending spread is less than 4 percentage points

whereas Croatias is 10 points.

Finally, Croatia and Columbia have almost

double the mortgage to GDP ratio as Slovenia,

the Slovak Republic and Poland but also have

much higher lending rate spreads.

Why is this? Mortgage depth can be much

more strongly inuenced by government

housing and housing nance policies than by

general nancial development. As discussed

below, well intentioned but erroneous policies

such as government imposition o interest

rate controls on mortgage lending by private

as well as public lenders to make loans aord-

able can have the eect stunting mortgage

market development even while overall nan-

cial markets mature and deepen.


14/121

FIGURE 1.2 MORTGaGE DEPTH Vs. TOTal CREDIT DEPTH

Sources: Chiquier (2006) and World Bank World Development Indicators.

0

20

40

60

80

00

20

40

Mortgage % GDP

TotalCredit%GDP

Brazil

China

Colombia

Croatia

India

Jordan

Latvia

Mexico

Poland

Slovak Republic

Slovenia

0 2 4 6 8 0 2 4

FIGURE 1.1 RaTIO OF OUTsTanDInG MORTGaGE DEbT TO GDP In sElECTED COUnTRIEs

Source: Chiquier (2006)

Depth Residential Mortgage Markets

0%

10%

20%

30%

40%

50%

60%

70%

Usa

EU-15

Cnd

Hong-Kong

Myi

Etoni

ThindChie

ltvi

CoomiChin

Mexico

Jordn

Croti

boivi

Hungry

Morocco

Tunii

CzechRep.

brziIndi

Pond

sovki

soveni

bngdeh

IrnPeru

ageri

sudiari

PkitnGhn

DebtAs%GDP


15/121

FIGURE 1.3 MORTGaGE DEPTH Vs. lEnDInG EFFICIEnCY

Sources: Chiquier (2006) and World Bank World Development Indicators.

-

2.0

4.0

6.0

8.0

0.0

- .0 6.0 9.0 2.0 5.0

Mortgage % GDP

lendingRatespread(%)

China

Colombia

Croatia

Jordan

Latvia

Mexico

Poland

Slovak Republic

Slovenia

is the better-o amilies. Incremental housingdevelopment carried out on a cash-only basis

is not only slow; it tends to be quite inefcient

- the tendency is to begin with poor quality

materials that need repeated replacing and

possibly to demolish when the next stage

o construction is carried out (UN-Habitat,

2005; 159). In sum, a low volume o nance

means unnecessarily poor living conditions

and extra housing costs to the poorest seg-

ments o a countrys population.

ORGanIsInG a HOUsInGFInanCE sTRaTEGY

A housing strategy is a plan or deploying

available resources (and i needs be, increas-

ing them) to nance the demand or housing

by dierent segments o society. Te purposeo a strategy is to get the most rom available

resources. Preparing and implementing a strat-

WHaT DOEs a lOW VOlUMEOF HOUsInG FInanCE MEan?

Loans or housing permit amilies to leverage

the unds to spend each month on housing so

that they can purchase a unit sooner or reach

the next level o incremental development o

their dwelling. Te same monthly payments

can go to loan repayments or they can be

saved or years until the amily can purchase

the housing unit. In economists jargon, loans

pre-empt housing demand.

Where nance is limited, housing construction

is slow, thereby driving up the price o housing

services - both rentals and units or purchase

- and this makes it even more difcult or most

amilies to save to obtain good quality hous-

ing. Critically, when housing nance is scarcethere is a denite pattern o who receives

most o the loan unds that are available: it


16/121

egy is a task that requires broad stakeholder

involvement, a good deal o technical analysis,

and strong political leadership.

Te emphasis on stakeholder involvement and

getting the process correct is clearly evident in

the steps in the strategy development proc-

ess shown in Figure 1.4. Te process starts

by identiying the relevant stakeholders and

ensuring their active involvement, beginning

with dening the strategic objectives.

Te primary technical steps in strategy ormu-lation are, in a simplied orm, the ollowing:

Understand the broad housing situation inthe country, including basic housing qual-ity and related inrastructure conditions,the extent o preerence or home owner-ship versus renting

Determine eective housing demand byincome class and location (urban, rural) to

identiy and dene market segments andthe volume o nance required by eachsegment in the planning period (note thatbecause loan terms dier across householdswith dierent incomes, there is no simpleratio to apply to income to reach a housingnance demand estimate).

Inventory housing nance currentlyavailable - products and volumes, marketcoverage and lending policies (or example,treatment o dierent sources o incomein underwriting standards, physical access,registration requirements) by market seg-ment. Tis must include all sources - private,both ormal and micro, and government.Find out where lenders obtain their unds,and the elasticity o these sources

Determine the gap between potential de-mand and current supply or each market

segment

Prepare a plan or closing the gap that isinormed by the inormation developed inthe prior steps.

Sounds simple? As the ollowing chapters

document, it is not simple but it clearly is

doable. Te best technical analysis will be

unproductive i the nancial community,

broadly dened, and the relevant government

agencies are not ore square behind it.

Naturally, a primary question is how can the

gap be closed or the plan be ullled.

THE sTRaTEGY MUsTbE COMPREHEnsIVE

oo oten housing nance strategies address

only the requirements o ormal nance and

then only in a countrys principal cities. Such

strategies will oten omit a major share o all

households. Tis is a clear mistake.

Consider the ollowing observations over the

past decade o developing nations:

In Mexico, sel-built housing accounts orroughly hal o all new buildings.

In nine Asian countries, between 40 and95 percent o all households have no pos-sibility o acquiring a unit built by the

ormal sector, and thereore little chance oobtaining nance.

An estimated 70 percent o housing in-vestment in developing countries occurthrough progressive building and thereorewith little nance beyond householdsavings (UN-Habitat (2005), pp.70, 99,110).

Room- and unit-renters account or the

principal tenure orm or lower incomeamilies in the urban areas o a number oArican countries.


17/121

FIGURE 1.4 OVERVIEW OF THE HOUsInG FInanCE sTRaTEGY DEVElOPMEnT PROCEss

PREPaRaTIOn

Identiy leaders o the process

Identiy stakeholders and include them in all aspects o the process

Dene work objectives and outputs

Dene the analytic program and oversee it

Analysis

Identiy housing demand segments and estimate current and expected demand in each segment

Document the current supply o housing nance, the segments o the market each services and impedi-ments to expansion in volume and market coverage

Identiy gaps between demand and supply by segment

Strategy Formulation

Identiy options or closing the demand-supply gaps

Increasing the volume o lending

Increasing the unds available or housing lending

Determine the most easible and eective options

Develop the action plan to implement the selected options

Implementation Monitoring and Evaluation

Establish monitoring arrangements

Monitoring the household side: who isreceiving ormal and inormal loans

Monitoring the supply side

Loan and borrower proles

Delinquencies and deaults

erm distribution o liabilities

Dene eedback mechanism


18/121

Te situation is problematic in some transi-

tion countries:

In Kyrgyzstan, in the years ollowing thecollapse o the Soviet Union, over 200,000people migrated rom the countryside tothe capital, Bishkek, in search o work.Tey are living in basic dwellings mostlywithout inrastructure services but withincommunities with ormal layouts thanksto prompt action by the local government.Little nance is available or upgrading.

In some o the countries o SoutheasternEurope, the Caucuses, and Central Asia alarge share o rural households live with-out piped water or indoor toilets. Again,nancing or housing improvement isscarce.

Tese examples highlight the need or a na-

tional housing strategy to address all market

segments, where segments are dened by

household purchasing power, location, andtenure orm. o do this micro nance requires

a ull role as a source or housing lending

where needed.

Micro housing lending can and has allowed

amilies to accelerate the rate at which they

consolidate their dwellings. Micronance

Institutions make a succession o loans to

amilies constructing their homes incremen-

tally, beginning with small loans or short pe-

riods and gradually increasing the size and the

loan period as borrowers demonstrate their

ability to pay. Loan agreements are typically

simple and the emphasis is on ability to pay

rather than on pledged collateral.

echniques are available to reduce urther the

credit risks associated with such loans.

(For urther discussion, see Ferguson and

Navarrete (2003).) 2

One sometimes hears that ew Micronance

Institutions make loans or housing. In act,

estimates suggest that around 30 percent o

Micronance Institutions loans or business

purposes are diverted to incremental housing

improvements (e.g., Cain, 2007). In part the

task is to ormalse a type o lending that isalready occurring.

HOW WIll THE sTRaTEGYbE FUlFIllED?

Establish conditions so that private sector

lenders, ormal and micro nance, can extend

credit to most market segments without tak-ing undue risk.

Over the past twenty years there has been

a gradual but airly pervasive shit toward

market provision o housing. In some cases

this is due to the poor perormance o public

institutions charged with constructing hous-

ing or acting as lenders (see Chapter 5 or

more on this). But probably more important

is that in many countries governments have

helped set the necessary conditions that per-

mit private lenders to prosper and serve most

o the market.

2 In this Guide no distinction is made between licensed and unlicensed

MFIs. In part this is because o the variation in these organizations

ranging rom small, highly localized NGOs to commercial bank

subsidiariesthe diversity across countries in the legal environments

in which MFIs operate, and dierences in the degree o supervision to

which dierent classes o MFIs within a country are subject.


19/121

10

TablE 1.1 EXaMPlEs OF MORTGaGE FInanCE InHIbITORs anD POssIblE GOVERnMEnT

sTEPs TO REsOlVE THEMProblem Possible Solution

There is a mismatch between ormal lenders under-

writing standards and many borrowers qualications.

Banks are lending only to salaried employees and

have high minimum loan sizes.

The Central Bank and the Bankers Association agree

to press or lower loan sizes and or lending to those

with less easily documented incomes. It is recognised

that this will make lending more costly and should be

refected in the interest rate charged.

Micronance Institutions are short o lendable unds

or housing. One option is larger lines o credit rom

commercial banks. Banks are concerned that the risk

o such credit lines is signicant because Micronance

Institutions are not strongly supervised. o credits

to Micronance Institutions will be reduced and the

unds should fow.

The Government moves supervisory responsibility or

Micronance Institutions rom a social ministry to the

Central Bank. The Central Bank will develop alterna-

tive standards or Micronance Institutions compared

with commercial banks but supervision will denitely

be stronger or Micronance Institutions than it is

now. The risk level perceived by commercial banks in

extending lines

It is worth spending another moment on

the governments role. Governments are in a

unique position to execute two critical unc-

tions. Te rst is to set enabling conditions

in which private lenders are able to operate

successully. Tese are discussed urther in the

next section. It is oten not a matter o an ab-

solute shortage o unds but rather the terms

used by lenders in making lending decisions.

Banks in Arica, or example, are oten quite

liquid (Honohan and Beck, 2007, pp. 74-5),

but they do not lend because they employ

archaic underwriting standards or believe

the associated risks to be too great. able 1.1

gives two o many examples o actions that

government agencies could take to promote

increased lending by banks and micro lenders.

Second, governments are in the unique posi-

tion to acilitate market development through

three channels:

Legal ramework and related supervisionor Micronance Institutions: matureMicronance Institutions need the author-ity to attract unds as depositories and they

need strong supervision to give condenceto lenders who can provide mid-term und-ing to them that is suitable or multi-yearhousing loans.

Insurance: Risk sharing between govern-ment and private lenders, with governmenttaking the risk o exceptional events, e.g.loss rates beyond ex ante careul estimates.Examples include, mortgage deault insur-ance and insurance to commercial banks in

1.

2.


20/121

11

extending lines o credit or housing loansto micro lenders.

Creation o acilities critical to mortgagemarket development, which are too riskyor the private sector to undertake in thecurrent stage o market development.Tese might include a secondary mortgageacility or a credit rating agency. Tey arerisky in the short-run because o the largeup-ront investment required and the rela-tively low volume o initial activity. Whensuch organisations are more mature, theycan be privatised.

Concrete examples o these actions are pre-

sented in later chapters.

In summary, a housing nance strategy

requires a clear statement o objectives, an

understanding o local conditions, a sense

o how policy and programme eatures are

linked to outcomes, and a plan or generating

and applying the resources needed to imple-ment the strategy (Mayo et al., 1986, p.198).

Dening objectives at the start o the plan-

ning process is necessary to help guide it, but

the nal objectives will be inormed by the

results o the analysis undertaken in strategy

preparation. In short, objectives setting will

be an iterative process.

PRIOR COnDITIOns FORPREPaRInG a sTRaTEGY

Te process detailed below or developing

a realistic and actionable national housing

nance strategy presumes that some unda-

mental conditions exist. Simply stated, mar-

ket conditions must be reasonably conducive

to long term housing lending to make such

an exercise worthwhile. I basic conditions

3.

are not in place, improvements to the housing

nance system will be nearly impossible.

Conditions that should be in place to acilitate

micro and ormal lending include: 3

Support or both micro and ormal lending

reasonable macroeconomic stability - highor volatile ination produces large interestrate risks; stagnate growth restricts the abil-ity o borrowers to make payments

no interest rate controls on mortgage lend-ing; they typically make lending unprot-able

well-dened and protected property rightsthrough a unctioning registration system.

Support or micro lending

Legal support and supervision orMicronance Institution operations, in-

cluding multi-level licensing based on ca-pabilities and mission o the MicronanceInstitutions; at the highest level they canbe depositories and very strong nancialstability. Te supervisory authority musthave the same competence as the centralbank (and indeed could be this agency).

sUPPORT FOR FORMal lEnDInG

no directed credit requirements rom theCentral Bank to commercial banks orhousing lending; banks will lend i suchlending is protable; orced lending istypically ineective

The necessity or the various conditions in this list has been stated

by Buckley (994), Jaee and Renaud (997), Tomlinson (2007), UN

Habitat (2005), p.55.


21/121

1

no large government lending programsthat make long-term housing nance avail-able at below-market interest rates because

lenders will not want to try to competewith the cheap money and poor repaymentcollection rates

the basic legal system or lending in placeincluding a serviceable mortgage law anda unctioning, and at least moderately e-cient, oreclosure system to increase thecertainty and reduce the cost to lenders.

In the absence o these conditions, develop-

ment o a ull housing nance strategy is notadvisable. Rather, a strategic exercise ocused

squarely on these problems should be pre-

pared. When these impediments are in the

process o being resolved, a ull strategy can

be developed.

THEsE GUIDElInEs

Te ollowing presentation is ocused squarely

on increasing the volume o nance available

to all market segments in an economically

sustainable way. Tis ocus is maintained by

not explicitly covering several closely related

topics, particularly

the real side o the housing market, i.e.provision o sufciently serviced land or

housing, presence o reasonable housingconstruction standards, or an efcienthousing construction sector

broadly reviewing the current state ohousing nance development or identiy-ing which type o lender may be the most

efcient or eective in a given circumstance,i.e. savings association or commercial bank(more inormation on this topic can beound in UN-HABIA (2002) Chapter1 and UNHABIA (2005)

examining subsidies that operate outsidethe nancial system or low income house-holds to improve their housing conditions,e.g. the direct construction o rental hous-ing or this group.

Overall, the presentation consists o three

parts.Tere are two more chapters to this

rst part that give an overview o a national

housing nance strategy. Part IIs our chap-

ters detail how such a strategy is prepared on

a step-by-step basis. And Part III discusses

the development and implementation o the

Action Plan.


22/121

1

02WHaT a sTRaTEGY DOEs

A housing nance strategy is important as a

device or organising and ormalising think-

ing about how to more efciently employ a

countrys housing nance resources and toexpand them. It provides a ocal point or

such thinking and results in a realistic action

plan that can be implemented over the years.

Few strategies have been published. Tis is be-

cause there are working documents or single

countries, and not a generic document that is

ormally published. It does reect the reality

that not many have been prepared, presum-ably because this is a substantial undertaking.

By contrast, numerous examples have been

published about broad advice on mortgage

sector development and strategy statements

or improving the volume and access to hous-

ing nance or a particular market segment

(or example, Chan et al. (2003), Ikejioor

(1998), Mayo, Malpessi, and Gross (1986),

Pugh (1994), Ferguson and Navarrete (2003),

Siembieda and Moreno (1997), and Mints

(2004)).

an IllUsTRaTIOn

A good way to understand a strategys util-

ity is to examine one. anzania prepared a

strategy in 2003, although it was not ormally

labeled as such (Government o anzania andUN HABIA, 2003). At the time, the basic

conditions or the development o a housing

nance system were barely met in the country.

Te strategy appropriately devoted most o its

attentionespecially in the action plantosteps that would address these undamental

impediments to system development.

Some context or the strategy is provided in

the gures in ables 2.1 and 2.2 or Dar es

Salaam. In 1990, (the latest date or which

data were available in 2003),the majority o

dwellings acquired in both the ormal and

inormal sectors were constructed by simplelaborers (undis). Units constructed by own-

ers (sometimes with hired help) accounted

or 11 percent in the ormal and 32 percent

in the inormal sectors, respectively. Only

22 percent o occupants obtained their units

by inheritance or by purchasing an existing

unita tribute to the citys rapid growth.

Only 8 percent o ormal sector units were

built using construction loans rom nan-

cial institutions, especially the government

housing bank, which accounted or only 8

percent o ormal sector units. Savings were

overwhelmingly the most common source o

construction unds, accounting or 58 percent

o all units. In interpreting these gures it is

important to know that 56 percent o house-

holds in Dar were room renters at this time

and another 17 percent were unit renters.


23/121

1

TablE 2.1 HOUsE aCQUIsITIOn In DaR Es salaaM In 1990 (PERCEnT)

Mode o houe cquiitio Formal sector Inormal sector All ownersOwner constructed 2 24

Constructed by undis 56 5 54

Contractor-built --

Bought house rom previous owner 4 8

Inherited the house 9 9

Other means

Source: Government o Tanzania UN-Habitat (2003)

TablE 2.2 sOURCEs OF FUnDs FOR DWEllInG COnsTRUCTIOn (PERCEnT)

source o ud Form ector Uped ector a ower

Savings 6 56 58

Selling assets 9 5

Loan rom riends 7 4

Loan rom employer

Credit union 2

Tanzania Housing Bank 6 2 4

Unspecied 2 4 25

Source: Government o Tanzania UN-HABITAT (2003)

Loans rom the anzania Housing Bank, es-

sentially the only lender and now deunct, was

reported by only 6 percent o those construct-

ing ormal sector housing and 2 percent o

those constructing in the inormal sector.

Te strategy reports content is outlined in Box

2.1. It spent a good deal o time documenting

the existing problems in the housing sector as

well as in the provision o housing nance, and

it explains in some cases how these conditions

arose. Te report spent comparatively littleenergy on preparing estimates o potential

eective housing nance demand. Tis was

appropriate given the housing nance sectors

state o development.

Clearly the reports most important contribu-

tion has been the accurate identication o

the principal problems and the assignment o

responsibility to organisations or their reso-

lution - a real action plan. Te actions to be

taken are stated in the report as well.

Did the planning exercise make any dierence?

No ollow-up report documents the degree towhich the action plan was executed. But one

can compare the state o development o the


24/121

1

housing nance system in 2007 with that o

2002 and the extent to which legal changes

called or in the action plan have been ac-complished.

Te ollowing points provide an overview o

the mid-2006 status:

Critically, macroeconomic stability hadbeen restored. Ination had droppedrom 30 percent early in the decade to 4.2percent at the end o 2005. Tis stabilitymakes long-term lending much more at-

tractive to banks.

Formal sector housing lending remainedminimal, owing primarily to worries oexcessive credit risk.

Te micro nance sectors regulatorystructure had been notably strengthenedand major Micronance Institutions wereabout to become depositories. A clearobjective o the regulatory improvementswas to make banks more willing to providemicro lenders with term loans, althoughunds had yet to start owing. Whileexplicit housing loan products had yetto be introduced, a share o loans nomi-nally or other purposes was being used orhousing construction and improvement.(Inormation rom Merrill and omlinson

(2006).)

bOX 2.1 OUTlInE OF THE TanZanIa HOUsInG FInanCE sTRaTEGY

Review o current conditions and past initiativesMacroeconomic environment, real interest rates

Housing conditions, levels o unit construction by ormal vs inormal (compliance with building codes)

History o ailed GOT active housing interventions, particularly the Tanzania Housing Bank

Financial sector reorm process and current nancial sector conditions

Formal real estate sector, including the supply o serviced land

Inormal real estate sector

General assessment o eective housing demand and a comparison o the cost o alternative housingtypes.

Supply o housing nance: assessment o current sources

Impediments to development o the housing credit system

Real side

Land occupied without titleso no ormal nance available

Registration services are problematic

Financial side

Shortcomings o the Mortgage Law, Foreclosure Law

Shallowness and inexperience o the sector

Options to increase housing lending

Action plan statement and implementation responsibilities

.

2.

.

4.

5.

6.


25/121

1

able A.1 presents a more detailed status re-

view o selected legal changes called or in the

Action Plan by listing the changes called or in2003 and the relevant laws in 2007. (Tis sum-

mary draws on Rabenhorst and Butler (2007)

and inputs rom one o the authors.) Progress

has been mixed at best. Notable progress has

been achieved in three areas: (i) positioning

micro lenders to make housing loans and

obtain nance rom commercial banks; (ii)

implementation o a legal regularisation

programme or spontaneous settlements; and

(iii) making signicant progress in preparingthe way or a secondary mortgage acility. Te

rst two reorms are particularly important

because they may result in a dramatic expan-

sion o credit to households not qualiying or

bank housing nance.

While these three actions are encouraging,

one must also note that the other dozen called-

or actions listed in the table have seen littleprogress. Perhaps most damaging is the lack

o revision on key provisions in the mortgage

law. A mortgage market ask Force created

in 2006 is now working on these issues and

progress is expected.

lEssOns

Te anzania example makes three important

points:

a strategy can directly or indirectlyserve as an organising device in aconcerted eort to strengthen anddevelop a countrys housing nancesystem. Real improvements can anddo ollow. In most countries achiev-

i.

ing the kind o change called or ina strategy is challenging.

a country can develop a strategywith varying degrees o precisionthat requires varying level o re-sources. Much o the material in theollowing chapters addresses techni-cal analysis. Te limited range oborrowers served and the many othe key impediments to expandedlending will be evident with com-

paratively simple analysis. Beyondthis, better analysis is desirable;but more important is developingthe strategy collegially to attract allthe key stakeholders and obtain anagreed action plan that has a solidchance o being implemented.

strategies take time to come to reali-

sation. Key actions are oten thoserequiring changes in a nations lawor the creation o new institutions.Passing the necessary legislation istime consuming. In countries with aCivil Code legal structure, the Codeitsel may need to be modied whichis a difcult undertaking. Creationo new institutions, such as a credit

bureau or secondary market acili-ties, is a major undertaking requir-ing several years at best. Te timerequired to realize a plan shouldinorm action plan schedules andexpectations in the strategy prepara-tion process i disappointment is tobe avoided.

ii.

iii.


26/121

1

sTRaTEGY OVERVIEW

Successul housing nance strategy develop-

ment has two co-equal elements: technical

analysis and developing support in the nan-

cial community (broadly dened), relevantgovernment agencies, and consumer-oriented

non-governmental organisaions. Tis chapter

briey outlines the two components o a

ramework or a more detailed presentation.

THE TECHnICal ElEMEnTs

Tere are ve basic technical tasks or prepar-ing a housing nance strategy.

1. Dene concrete objectives. Although this is

listed rst, and some general goals should be

dened when the project is launched, realistic

objectives will be dened on the basis o the

inormation developed in steps 2 through 5.

Tis is an interative process.

2. Identiy housing nance demand segments.

Housing aordability is dened by savings

levels (downpayments), income with which

monthly loan repayments can be made, and

loan terms. Separate estimates are or owners

and renters and or principal locations, e.g.,

the capital, other cities, and rural areas. Te

result is an estimate or each market segment

03

o the annual volume o nance required to

meet demands.

3. Document the current supply o housing -nance. Tree sets o questions are addressed:

Who gets loans rom which lenders andon what terms? What lenders practicesmay be limiting access? (Lenders includebanks, ormal non-bank institutions, andMicronance Institutions.)

What are lenders sources o unds andhow elastic are they, i.e. could they ex-

pand to meet a larger demand or doesexpansion need to be included in thestrategy?

For each type o lender, what is the e-ciency in loan origination, servicing,overall operations, and risk manage-ment?

4. Identiy the gaps between demand and sup-

ply. Tis should be done by segment and one

should identiy impediments to a greater

volume o nance in segments where demand

ar outstrips supply.

a.

b.

c.


27/121

1

5. Develop creative responses to the gaps dened,

select an approach, and codiy it into an action

programme.

Tese topics are covered in Chapters 4-6.

THE POlITICal PROCEss

Without wide-based support in the nancial

community, including the Central Bank,

among consumers, and within government

agencies even the best technical analysis or ahousing nance strategy will have little eect.

Donor-led strategy eorts ace particular chal-

lenges in coalition building.

Te natural tendency will be to look rst and

oremost or leadership in government agen-

cies and the major commercial banks and

other major players in the ormal nancial

market. Tis is necessary and appropriate be-cause o the resources they command. But it is

ar rom sufcient. For a strategy to be eec-

tive it must comprehensively cover all market

segments, especially the micro-lenders serving

lower income amilies with smaller credits.

Tese lenders and consumers, represented by

Micronance Institutions and the relevant

non-governmental organisations, must as

well be ull partners in the process. Without

the ull range o participants, the strategy is

very likely to address the needs o only upper

income households in many developing

countries. Securing the kind o broad support

and buy-in essential or development and

implementation is a much greater challenge

than executing the technical analysis.

Below is a list o questions or those consider-

ing preparing a strategy. Questions that need

answers in order to structure the process andensure a successul outcome.

Who are the primary stakeholders andwhy?

Who has necessary motivation and prestigeto lead the exercise?

How can one engage the stakeholders?

How will they participate in the process?

Who prepares the strategy?

How is the strategys preparation nanced?

Ater agreement on the strategy or improve-ment, how will it be implemented? Whowill be responsible? With what resources?

Who will monitor the implementation?

Te process o motivating and organising the

strategys development and implementation isthe topic o chapters 7-9.


28/121

1

PaRT II

PREPaRInG THE sTRaTEGY


29/121

0


30/121

1

THE DEManD FOR HOUsInG FInanCE

Understanding the purchasing power - the

eective housing demand - o households is

undamental to valid housing strategy prepa-

ration. Preparing this part o the strategy ischallenging mostly because o the data require-

ments. But leaving these aside, demanding es-

timations can be mechanical exercise that alls

short o providing important insights into the

current situation and uture possibilities. Te

key step is to prepare an overview o current

conditions to orient the analysis.

CURREnT COnDITIOns

A modest number o acts can provide a clear

understanding o the basic housing conditions

in a country. Tese acts include

Te degree o urbanisation and the impor-tance o the largest 3-5 citiesthe share othe population living in ofcially dened

urban areas and the share living in the larg-est ew cities (usually there is a clear breakbetween the size o the largest 1-5 citiesand other smaller ones).

Te distribution o households in thelargest cities, other urban areas, and ruralareas among dierent tenure orms, rang-ing rom ormal home ownership to roomrenters in inormal dwellings.

Te share o dwellings in the three locationgroups that meet minimum standards interms o construction materials and provi-sion o public services.

Te growth in the number o householdsby location over the past 20 years and theorecast or the next 15-20 years.

Armed with this inormation, the results

o the demand estimates can be more intel-

ligently interpreted.

EsTIMaTInG HOUsInGDEManD

Te demand or housing nance is derivedrom households (owners and renters)

demand or housing services, and rom

the desire o owners to invest in housing as

an asset. Because owners and renters have

dierent housing objectives, the amount o

income they are willing to devote to housing

usually diers, with owners willing to spend

more. Similarly, amilies living in rural areas

generally have dierent housing requirements

and needs than their urban counterparts. Tis

leads to dierences in the share o households

who own their units and in the share o in-

come and the amount spent on expenses.

Tere is a good deal o experience in making

aordability calculations, where the loan

amount or households or dierent income

groups is based on their income and loan

terms. Tis inormation, combined with as-sumptions about downpayments, yields an

estimate o the dwelling price each particular

04


31/121

household group can aord. Te ollowing

list determines the number o loans or which

there is demand and then the amount onancing required to meet those demands.

Annex B contains a detailed exposition on

how to make such estimates.

Steps in computing housing nance demand,

with separate estimates prepared or house-

holds who will want to own their homes and

those who will be renters:

Determine the amount o money each in-come group is prepared to spend monthlyon housing.

Estimate the average loan size each incomegroup could take, using the expenditureinormation rom the previous step.

Estimate the annual number o loans thatwill be sought, based on inormation onhousehold ormation, complex householdsthat experience a amily moving out, thenumber o households trading up to betterunits, and the number o households whowill be continuing to develop their units orupgrade their completed units.

Te product o the loan size and number oloans sought gives the overall demand esti-mate and the target or the development ocorresponding housing nance supply.

A critical attribute o the estimates discussed

above is that they cover the whole market, not

just the market or ormal nance, which is

the ocus o most housing nance demand

estimates.

As stated earlier, these calculations ideally

should be done or owner-occupied and rental

housing in each type o location where there

are systematic dierences in the purchasing

power o the households (including bothhousehold incomes and the terms on which

1.

2.

3.

4.

loans are available) or in the price o dwellings

that would qualiy or ormal nance.

Te conclusion o this analysis is an estimate o

loan volume or 2-3 years in the uture, or sev-

eral estimates that vary with the assumptions

about the extent o institutional development

achieved, o the required volume o lending

by location, tenure, and income group. Tese

are the targets that policymakers will design a

strategy to achieve.

COnsTRUCTIOnPERIOD FInanCE

Tis type o nance provides interim unds

or building. Tese loans are paid o when

units are sold to owner-occupants, with the

developer using the sale proceeds to make

the payment, or when landlords obtain per-

manent nancing or the investment portionthey do not und with equity. In act, in many

countries loans rom ormal lenders or the

ull development o multi-amily or single

amily buildings constitute a small share o

total construction nance. In the inormal sec-

tor most construction is incremental, and its

unding has already been accounted or in the

computations presented above. Additionally,

developers o ormal dwellings oten obtain

much o their construction nance through

advanced payments rom the ultimate pur-

chasers.

In most transition and developing countries,

construction period loans are oten made only

on more prestigious projects and even then

only or a minority o the total cost. Lenders

have ound it difcult to underwrite these

loans and to control payments, leaving themvulnerable to credit risk. Market assessment


32/121

has proved challenging and many lenders have

suered signicant losses when they approved

loans at the top o a construction boom onlyto see the market collapse and loans turn sour.

Numerous lenders have released unds too

quickly compared to construction progress

and been unable to recover the unds when

the builder has decided to stop construction

or whatever reason.

It is evident that it is challenging to estimate

the volume o construction nance needed

and the amount that will actually be supplied.Many central banks do not gather separate

inormation on this kind o lending. Even

tracking recent lending levels is problematic.

In these circumstances, the best basis or de-

termining near-term uture loan levels and

what is needed to promote more o this kind

o lending, is discussions with both develop-

ers and lenders about their recent experiences,

problems o obtaining and making loans, anduture plans.

KnOWlEDGE OF THEMORTGaGE lOan

Te common assumption is that those in need

o housing nance understand housing loans,

particularly mortgages, and that they have

ready access to lenders oering these products.

Te truth is at odds with this assumption.

Large disparities have been observed between

the number o amilies who could borrow

and the number who actually do. o raise the

incidence o borrowing or home improve-

ments in many countries will require more

knowledge and better access to induce lendersto serve a wider market.

Tis section oers some inormation on

two aspects o the actual situationbasic

knowledge o mortgage loans and access by

lenders or both ormal and inormal housing

nance.

Te rst step to taking out a loan to purchase

a home to build a home is to know thatsuch loans exist and to have a rudimentary

knowledge o their main eatures. Tere is

little evidence o what consumers actually

know, particularly in developing countries.4

Te results o a recent survey conducted in

Cairo suggest that knowledge levels even

among the well-to-do are not very high (Box

4.1). Similar ndings were obtained in a

representative survey o amilies planning topurchase a dwelling in Indonesias seven larg-

est metropolitan areas (Struyk et al, 2008).5

Logically, i consumers have little knowledge

o mortgage products, they will be reluctant

to apply or a loan. It is in the interest o lend-

ers to mount an educational campaignss to

increase consumers awareness and boost loan

volumes.

4 There are marketing analyses or industrial countries o consumer

preerences or particular mortgage loan eatures, e.g., variable

rate versus xed rate mortgages. See LaCour-Little, M. (2007) The

Home Purchase Mortgage Preerences o Low- and Moderate-Income

Households, Real Estate Economics, 5, , 265-90, and other studies

cited therein.

5 No similar studies o the knowledge o housing micro nance products

have been located.


33/121

bOX 4.1 KnOWlEDGE OF THE MORTGaGE lOan In CaIRO

The use o mortgages to purchase a house is nascent in Egypt. Clearly, the rst step in activating consumers to

take out a mortgage is or them to understand this instrument. The results reported here are rom a February

2007 survey o 504 Cairo consumers rom the middle socio-economic bracket who stated that they planned

to purchase a unit in the next three years or had purchased a unit in the previous ve years. The survey was

designed to determine their understanding and knowledge o mortgage loans and their attitude towards them.

This is not a representative sample.The ndings give the upper limit on knowledge levels.

About 5 percent o respondents reported having heard or read something in the past year about taking out

a mortgage. This is a airly high penetration rate and indicates that the concerted marketing campaigns o the

prior years were reasonably successul. O those who had heard something on the topic, close to hal stated

that they had paid close attention to the inormation or had careully considered whether a mortgage would

be a good thing. Such a high rate o close attention indicates that many in the sample population have a real

interest in the subject, as one would expect rom recent and would-be home purchasers.

Respondents demonstrated only a low to modest understanding o mortgages, as ascertained rom responses

to an open-ended question asking them to say what a mortgage is. However, respondents scored higher on

their knowledge o specic knowledge loan eatures, such as the consequences o not making payments on

time, and the maximum share o income that could be used or monthly payments according to the law.

The results o regression analysis o actual knowledge o mortgages indicate that exposure to inormation on

this instrument, and the amount o attention consumers pay to it have been key in shaping knowledge and

attitudes. The respondents socio-economic class and age, or even participation in a registration, on the other

hand, has had little bearing. The ndings highlight the important role that education campaigns and media

coverage have played in inorming the population. They also indicate that experience with home purchase

mortgages and property registration is so limited that even among households rom higher socio-economic

groups, understanding and knowledge o these instruments are not common. Both points argue or the con-

tinuation o campaigns to inorm the public.

Source: Struyk (2007)


34/121

aCCEss TO HOUsInG FInanCE

Who can borrow and or what they can borrowis a moderately complex subject. Te discus-

sion is divided into two parts. Te rst (and

main) part covers general access to nance

in developing countries and then considers

evolving options that may well increase access

sharply. Te second part ocuses on access to

nance or rental housing or lower income

amilies, in part to emphasise this largely

neglected element among housing nance

instruments.

Overview. Because o the lack o studies on

access to housing nance due to geographic

isolation, the sel-imposed restrictions o

ormal sector lenders, and the comparatively

low incidence o micro lenders making hous-

ing loans, reliance here is on inormation on

general access to nance. Te basic picture is

available rom the gures in able 4.1 or 20transition and developing countries on the

percentage o households who took a loan

rom any source, ormal or inormal, in the

12 months preceding the survey that gathered

the inormation. Micro lenders are included

in the inormal group as are inormal money-

lenders, and rotating savings clubs. (See notes

to the table or details.)

Te unweighted mean or the percentageo households taking loans rom ormaland inormal lenders, respectively, are 6.5and 17.6 percent, i.e. inormal loans are2.7 times more common than ormal.

In seven countries, less that 5 percent ohouseholds took a loan rom a ormalsource. By contrast, only three countrieshad a low participation rate in inormalloans.

Five countries have very high rates (29 per-cent or higher) o households borrowing

rom inormal sources. Vietnam is alonein high borrowing rates rom both ormaland inormal sources.

Te overall picture is o low levels o consumer

lending by ormal lenders in these countries,

with inormal sources taking up most o the

slack.

Tree reasons stand out or low nancial

services accessibility rates, i.e. in this case, the

share o households who can readily borrow

money.

low service penetration rates, i.e.the extent to which lenders cater todierent sections o the population.

the geographic isolation o manyhouseholds rom lending branches,even in urban areas.

the unsuitability o products oered

to many amilies, including thehigh transactions costs o having asavings account.

A recent analysis examined the ormal lender

branch and AM penetration rates dened

on both geographic and demographic bases

in nearly 100 countries. (Geographic branch

penetration is the number o branches per

1,000 square kilometers; demographic pene-

tration is the number o branches per 100,000

population.) Analysis o these rates shows that

larger economies enjoy greater levels o out-

reach, suggesting scale economies in service

provision by larger nancial institutions. Te

quality o institutional support or the bank-

ing system and physical inrastructure explain

cross-country variation in outreach, e.g.

density o the road network, Finally, greater

importance o government banks is associatedwith less outreach (Beck et al. (2005), p.5).

i.

ii.

iii.


35/121

ransactions costs are a signicant impediment to households opening accounts with ormal

lendersoten the required rst step or loan eligibility.

TablE 4.1 sHaRE OF HOUsEHOlDs UsInG bORROWInG FUnDs In THE PasT 12 MOnTHs(PERCEnT)

Coutry Tht orrowed ipt 12 moth

Tht ued ormfci ititutio

Tht ue iormfce to orrow

Armenia - - 4.5

Bosnia & Herzegovina 2.8 6.2 5.8

Botswana - .7 29.

Bulgaria 5.4 5.4 -

China (Hebei & Liaoning 28. 5.2 24.7

Cote dIvoire 2.8 .2 2.

Guatemala .8 2.5 7.4

Guyana 4.7 . 2.5

Jamaica 0.5 .9 5.9

Kyrgyz Republic 6. 0. 5.

Namibia - 5. 5.2

Nepal 57.0 2.9 50.0

Nicaragua 22.5a 7.6 7.8

Pakistan 0. . 29.4

Panama .5 0.8 0.8

Peru 6.6 2. 4.5

Romania 5.9 6.0 .4

South Arica 44.8 4.9 42.6

Swaziland - 4. 6.

Viet Nam 49. 26. 0.4

Unweighted average 2.8 6.5 7.6

Source: Claessens (2006), Table 1.

a. Includes credit purchases.

-- Not available.

Note: Formal nance providers include banks (public or private), cooperatives, and credit un-

ions. For a ew countries other nancial institutions, such as security rms and postal savings,

are also included in ormal. Inormal includes others that provide nancial services, such as

Micronance Institutions and non-governmental organisations, rotating savings and credit as-

sociations, moneylenders, pawnshops and other country- or region-specic arrangements.


36/121

In this context, a notable success has been in

South Arica and its so-called Mzansi account,

with 3.3 million savers opening accounts inthe rst 18 months. Te account is a basic,

standardizsd debit card-based transactional

and savings account. All that is required to

open the account is a valid ID. o keep costs

down and to ensure that the account is easy

to use, transactions are limited to deposits,

withdrawals and debit card paymentsthe

account includes a debit card that can be

used at retail outlets. No management ees

are charged and one ree deposit per monthis permitted.

Mobile and cell phone banking holds the

promise o sharply increasing the share o

households with a deposit account with a

ormal nancial institution. Te need or

improved access is particularly prominent in

rural areas, but signicant areas o major cities

are underserved as well. One approach is theuse o mobile banking where the volume o

business does not justiy opening a branch o-

ce. Tere is notable positive experience with

this approach (Box 4.2).

Te promise o mobile banking or mortgage

origination has been realised by HDFC-India,

a major ormal sector lender, or a number o

years. Its agents visit villages on a regular basis

to collect deposits and take loan applications.

Tis technique has been an important element

in its overall lending strategy.

Lastly, cell phone banking is being utilised

in some places and shows promise. Certain

South Arican nancial institutions intro-

duced m-banking options rom about 2004

and marketed them strongly. An analysis o

this experience ound that by 2006 it had notresulted in a major increase in the number o

households with ormal banking relationships.

In part, this was due to substantial ignorance

about the product despite inormation

campaigns. On the other hand, the analysispoints out that since about one-third o the

unbanked have cell phones, there is reason to

anticipate expansion (Porteous, 2007).

Te uture role o inormation and commu-

nications technologies in increasing access

by the poor to banking services appears posi-

tive but a good deal o uncertainty remains.

Conclusions based on a recent survey o cur-

rent practices and the experience o 62 nan-cial institutions in developing countries with

inormation and communications technolo-

gies are that so ar not many new customers

are being attracted to these services, although

some poor households are using them. Te

protability o banks remain uncertain. 6

Rental housing. Te topic is the extent o ac-

cess o small investors to nance to developerental rooms or units. Where rentals are widely

accepted as an adequate and permanent hous-

ing solution, encouraging the development

o such housing through the availability o

nance will be an important element in a

housing nance strategy.

Tere is no inormation on the extent o either

the demand or or the supply o unds or this

purpose. Discussions about micro housing

nance lending ocus on unds or the devel-

opment or improvement o owner occupied

units. It may well be that a signicant share

o the micro small and medium enterprise

lending that is diverted to housing goes to the

construction o rental rooms or units. But the

inormation is lacking.

6 Analysis details and a list o the nancial institution respondents are in

Ivatury, G. (2006) Using Technology to Build Inclusive Finance Services.

Washington, DC: CGAP Focus Note no. 2 .


37/121

bOX 4.2 MObIlE banKInG REInVEnTED

Kenyas Equity Bank holds a prominent position among those interested in outreach to the poor. One reason

or this is the practical reintroduction o a relatively simple technology brought up to date: the mobile bankingunit. In 2006, the bank was doubling its number o vans and expected to have 00 in operation by the end o

year. Each van is equipped with laptops that have telecommunication links to a xed branch, allowing the van

to provide a wide range o banking services. The standard schedule is or a mobile unit to visit each location

once a week. By mid-200, two-thirds o the loans outstanding were to clients served through mobile banking

units.

Source: Honohan and Beck (2007), p. 155.


38/121

HOUsInG FInanCE sUPPlY

THE bIG PICTURE

A good starting point is to assess the market

or housing lending by understanding whichlenders are making housing loans available or

what purpose. able 5.1 illustrates this type o

classication. It is useul or identiying the

lenders rom whom inormation should be

acquired or dierent types o lending.

Te table lists six sources o nance and three

possible uses o nancepurchase o a ull

unit, incremental unit development, anddevelopment o rental housing. Financing o

rental housing is oten overlooked in housing

nance discussions, but rentals are an impor-

tant tenure orm in many countries, despite

the act that mass privatisation o ormer

State-owned housing in Eastern Europe and

the ormer Soviet Union in the 1990s created

a dozen or so super home owner states where

80 percent or more o all housing is owner-

occupied.7

7 Another source less requently employed are community unds. Such

a und is a nancial mechanism that encourages savings through

establishing and strengthening local savings groups with the und

typically used by communities or land purchase, inrastructure and

service investments, and, in some cases, housing construction. For

urther description, see UN Habitat. (2002) Financing Adequate

Shelter or All: Addressing the Housing Finance Problem in Developing

Countries. Nairobi: UN Habitat.

Te work outlined in the previous chapter will

provide a good idea o the volume o unds

required by dierent market segments over

the next several years. Here the work shits tothe supply side o the market. Te analysis in

this case is more critical than on the demand

side. In most countries it is obvious that the

unds needed to meet eective demand are

not available or, i lenders have unds, loans

are not being made available. In addition to

understanding the current lending patterns,

analysis must dig deeper to understand the

basis or the current pattern. It is quite pos-sible, or example, or lenders to believe that

they are open to serving a wide range o clients,

while not realising that their underwriting

standards or minimum loan sizes have the

eect o excluding most would-be borrowers.

Te analysis required consists o several related

activities. Overall, it is moderately complex,

and it is important or the reader to review all

the material and then to go back through it

again to be certain that he understands how

the pieces t together. A section at the end o

the chapter discusses the comparison o the

estimated demand or housing loans and the

supply o loanswhat is reerred to here as

the gap.

05


39/121

0

As noted in the last chapter, in Dar es Salaam

over hal o all households are room renters.

Tis pattern is airly widespread in Aricawhere rental housing is generally an important

housing source (more inormation on this can

be ound, or example, in Angel (2002), Angel

and Amtapunth (1989), Homan (1991),

and Ikejioor (1998).

Non-loan sources o nance are included in

the table or good reason: in most develop-

ing countries they are the primary source o

nance, as seen above or anzania. Whilecomprehensive data are lacking, that which

is available documents this act. In 2007, in

Cairo, 77 percent o those purchasing units in

the past ve years reported using cash, i.e. not

having a loan. A similar share o Jordanians re-

ported all-cash purchases in 1983 (Struyk and

Roman (2007)). Remittances are an impor-

tant source o savings or dwelling construc-

tion or purchaseor both owner occupancyand the development o rentals as an income

source. Housing investment is particularly

attractive in countries where savings products

do not carry competitive interest rates and

other investment opportunities to the small

investor are limited. Te low share o housing

nance provided by loans underscores again

the need or urther development o the hous-

ing nance sector in these countries.

In many developing countries, installment

sales by developers are a prominent source o

dwelling purchase nance. Purchasers must

make large up-ront payments and install-

ments thereater, which the developer uses

to nance construction. Oten, as in Egypt,

installment payments continue or several

years ater the unit is completed. Occupancy

may only be permitted when 80 percent ormore o all payments due have been made and

even then the title deed remains in some cases

with the developer until all payments are

made (Struyk and Roman (2007)). Similar

arrangements are common in Russia and

other countries o the Former Soviet Union.

In some cases there have been abuses by de-

velopers delaying construction while they use

the advanced payments or other purposes.

Some developers simply disappeared withdown-payments.

In some countries, government banks or

agencies have been an important source

o housing nance. Methods used include:

direct loans; down payment subsidies, such

as those in Chile, Hungary, and Russia; loans

originated by private banks or government

lenders at below market interest rates; and, acombination o these options. Te dierent

approaches have very dierent eects on the

development o private housing nance as

described on page 30.


40/121

1

TablE 5.1 POssIblE alTERnaTIVE FInanCInG sOURCEs bY HOUsInG InVEsTMEnT TYPE

source Ue o Fud

Purche ou uit

Icremet uitdeveopmet

Ret houigb

Market Loans

Mortgage loans rom

ormal institutions

Y N Unlikely

Micro housing nance loans Maybe Y Maybe

Developer installment sales Y N N

Loans rom employers Y Maybe N

Other sources

Savings rom overseas employment Y Y Y

Other sel-nance Y Y Y

Government programsa Y Y Y

a. Direct government lending or through private lenders

b. For construction or purchase o units or rental occupancy.

Tere may be important dierences in sourceso nance among large cities, smaller urban

areas, and rural areas. Separate analysis is typi-

cally appropriate by location type.

WHaT TO lEaRn abOUTEaCH TYPE OF lEnDInG

Te rst task is to understand loan volumes,

i.e. the volume o lending in each cell in the

upper part o able 5.1. While it is useul

to know total loans outstanding (stocks), it

is more important to receive ow data, i.e.

originations, or the most recent 1-2 years, to

understand what lender types are now most

active and which market segments are now

being served. Market segments are the

income-tenure-location groups o the type

shown in able 4.1. It is highly desirable touse the same denitions or the supply size

analysis as or the demand.

Some o the inormation needed to preparethese tables will be available rom ofcial

sources. Central banks track mortgage lend-

ing and other real estate lending o com-

mercial banks separately rom loans or other

purposes. Te regulator or micro nance

institutions may also have such data. However,

it is likely much o the inormation will have

to be gathered. Ideas or collecting this kind

o data are given below.

Te next task is to dene the products oered

by each lender or each o the three purposes

included in able 5.1. An illustration o the

inormation to be obtained or each product

or each lenderis in able 5.2. Note that the

let-hand column lists attributes o the loan

itsel, e.g. interest rates, the basis or comput-

ing the interest charge (simple interest or an

annuity loan), any additional charges or loanorigination, and so orth.


41/121

TablE 5.2 IllUsTRaTIVE TYPEs OF InFORMaTIOn TO COllECT On lEnDER REQUIREMEnTsFOR HOUsInG lOans TO InDIVIDUals

Fici term security d Uderwritig Coditio

Loan to value ratio: ocial and in-practice maximum Registered title to property required

Payment to income ratio: ocial and in-practice

maximum

Loan lien required to be registered

Loan term: ocial and in-practice maximum Guarantors required

Type o loan: xed rate, variable rate, price-level ad-

justed, dual rate

Income denitions employed: treatment o wies in-

come, remittances; level o documentation required

Interest rate Government workers subject to relaxed underwriting

Basis or interest calculations: annuity loan or simple

interest charge

Minimum loan amount applicable

Prepayment restrictions that apply Insurance requirements applicable: property, lie, title,

mortgage deault

Particularly critical in determining i the

lender will reach lower income borrowers are:

the security requirement o the borrowerhaving a registered title to the property

the income denitions employed andwhether certain types o income are in-cluded in the income used or loan under-writing, e.g. regularly received remittances;also, the degree o documentation requiredor each income source; and

minimum loan amounts.

Tese denitions can have a powerul eect on

restricting access to loans. For example, Egypts

National Housing Plan is providing low and

moderate-income households the chance to

purchase dwellings using the combination

o signicant downpayment subsidy, the

purchasers downpayment contribution, and a

mortgage loan at a near market interest rate.

It is important or the analyst to use the

terms in use by nancial institutions or the

loan-to-value ratio, payment-to-income ratio,etc. rather than rely on the maximums listed

in internal lending guidelines. Te analyst

should determine how these elements are

combined into the products most oten sold

by the lender.

Te right column lists certain security require-

ments the lender may impose, and underwrit-

ing requirements and other conditions that

may apply. For example, the Housing Finance

Company o Uganda requires land title, the

development where the dwelling is located be

in an urban area that has all services provided,

and the dwelling be constructed o durable

materials. In South Arica, some districts

are simply redlined because oreclosures have

been high there (UN-HABIA (2005), p.99

and 70, respectively).


42/121

o simpliy verication o income eligibil-

ity, the only acceptable applicants are those

whose income can be readily veried. Tisomits around 40 percent o all those believed

to be eligible or both the grants and loans

(Sims, 2007).

It would certainly be a mistake to think that

these restrictions are special to developing

economies. Stephens has documented a

host o restrictions among European Union

mortgage lenders, and red-lining is an ot-dis-

cussed phenomenon in the U.S. (see, respec-tively, Stepens (2004) and Avery et al. (1999)).

Further, it is likely that these restrictions will

multiply in the U.S. over the next ew years in

response to the sub-prime mortgage crisis.

As will be discussed in the next chapter, in-

ormation on the points listed in able 5.2

will be extremely valuable in thinking about

extending coverage to potential borrowerswho are now unserved.

What lending requirements should one obtain

rom lenders on loans made or rental housing?

Te nancial terms o interest are generally

the same as those or unit purchases, except

that the payment-to-income ratio will oten

not be applicable. However, the values will be

quite dierent or many items. For example,

loan-to-value ratios are much lowerseldom

over 60 percent; and, loan terms are shorter

than those or the longer home purchase

mortgages, oten only a ew years.

Te dierences o home purchase loans are

greater when it comes to loan underwriting.

One can list our distinctions that are impor-

tant to the underwriter:

Is the loan or a rental room being added tothe owners property, a room being added

1.

to a larger rental property, or or a singleamily unit?

Is the single amily unit or an elite higherincome market, including the expatriatemarket?

Is the loan or a multi-amily building ormiddle-income amilies?

Is the investor an individual or a com-pany?

I the loan is or something other thanroom construction, is it or the purchase

o an existing property or a new construc-tion?

Te oremost underwriting consideration will

always be the market analysisis their a ready

market or the unit? n other words, the key

question is whether there will be rental income

sufcient to cover the loan payments.For

small loans to individuals, the underwriting

is very similar to that or small and mediumenterprise loans. Security o title is another

constant requirement. But or new construc-

tion loans, with their extended timerames

and added uncertainties and larger multi

amily projects, other actors have to be care-

ully considered. (McCarthy (1996) discusses

residential property construction underwrit-

ing in the transition economy context.)

Te analyst should obtain inormation on

which o these various types o rental housing

a lender oers loans. For those or which loans

are made, a list o key underwriting actors

should be obtained. For those or which loans

are not made, inquiries should be made about

the reasons because these can eed into a list o

impediments that should be addressed when

drawing up a housing strategy.

2.

3.

4.

5.


43/121

Where can the analyst obtain inormation on

loan products, underwriting standards, and

other eligibility requirements? Te answerdiers depending on the type o lender.

For ormal sector lenders, the best optionis through a survey sponsored by theCentral Bank. Central banks are otenwilling to help gather inormation this way.For example, the Central Bank o Armeniacooperated in a survey eort with a teampreparing a housing nance project or theGerman KW agency. I this is not pos-

sible, then there are two other options: atrade association, such as the local BankersAssociation, which might be willing tosurvey its members; or interviews with asample o lenders, especially with the cur-rent major players.

For non-bank nancial institutions thatare not supervised by the Central Bank,a sample o these lenders will need to beinterviewed to obtain the inormation.

Since several regulators may be involvedand the number o institutions involvedsmall, it may be more efcient to conductthe survey without investing resources innegotiating with the agencies.

For Micro Finance Institutions the situa-tion may be more complex. Te rst choiceis to solicit sponsorship o a survey by theirregulating agency. I not, then interview asample o lenders. For those that are notregulated, a survey is the only option.

Finally, because installment sales by devel-opers are typically not regulated, interviewswith them will be necessary.

Inormation on which particular market seg-

ments are being served will only be available

rom lenders themselves, probably on an im-

pressionistic basis, and/or rom a household

survey that explicitly inquires about sourceso nance or dwelling purchase and incre-

mental unit development.

aDDREssInG lEnDEREFFICIEnCY anD RIsK

ManaGEMEnTHousing aordability depends substantially

on the price o the loan, i.e. the interest rate.

While lending rates on housing are driven

substantially by the lenders cost o unds,

other actors play major roles. In particular,

poor risk management, low overall efciency,

and weak practices in housing loan origina-

tion and servicing can lead to unnecessarily

large spreads.

Tese are important topics, but they are not

the central elements in a strategy. Tey are

discussed here because o their importance,

and because the prese

un habitat guide to housing finance strategy

Documents