Rising Household Debt: Implications for

Macroeconomic Stability

Presentation at Bank of Thailand Symposium 2014

October 17, 2014

Athiphat Muthitacharoen, PhD SCB Economic Intelligence Center

Phacharaphot Nuntramas, PhD SCB Economic Intelligence Center

Pasit Chotewattanakul Bank of Thailand

Disclaimers: Views and opinions expressed in this presentation belong to the authors and should not be interpreted as those of Siam Commercial Bank PCL. or Bank of Thailand.

Thailand’s household debt has been rising fast and is now at the level of rich countries

2

Thailand’s household debt has been rising fastest in the region during the last 3 years..

Household debt in % of GDP, Change in percentage point (2010vs.2013)

..and it is currently at level similar to the rich countries

Household debt in % of GDP, 2013

Source: CEIC

6 10

47 56 59 60

75 83 83 85 87

92 96

119

Jap

an

Ph

ilip

pin

es

Sin

gap

ore

Ge

rman

y

Ital

y

Sou

th A

fric

a

Ind

on

esi

a

New

Ze

alan

d

Thai

lan

d

UK

US

Au

stra

lia

Ko

rea

Mal

aysi

a

1.0

1.8

3.5

5.0

12.1

13.4

19.3

Philippines

Indonesia

Australia

Korea

Malaysia

Singapore

Thailand

At household level, the high indebtedness among low-income households is particularly worrying

3

Average DSR by income quintiles, 2009-2013

Unit: Percent

Notes: Dropping observations with unreasonably high DSR values (DSR > 1000%). (9 such observations in the 2013 dataset) Source: SES

24242428

49

28232325

29

47

28232325

29

49

27

Second quintile Third quintile First quintile All Fourth quintile Fifth quintile

2013 2009 2011

54% of Thai Households are Indebted

Debt Service Ratio (DSR)

Principal payment

Interest payment

Income

Poorest households

<9 9-14 14-21 21-33 >33 Income Brackets (Thousand baht)

Richest households

How could the large build-up of household debt be dangerous to the economy?

4

Short-term risk to financial stability

Rising Household Leverage

Restrain economic activity

Reduces shock absorber and increases economic volatility

Lower households’ disposable income and constrain credit access

The risks may not be equal across all lenders

Subjects banks to higher exposure towards household sector

Research Questions, Model Strategies and Data Construction

5

Debt at Risk Does the rising debt

increase short-term risk to financial stability?

Research Questions Model Strategies

1

Research Questions, Model Strategies and Data Construction

6

Debt at Risk Does the rising debt

increase short-term risk to financial stability?

Research Questions Model Strategies

1

Regression Analyses To what extent, does the debt surge hurt

consumption?

2

Research Questions, Model Strategies and Data Construction

7

Debt at Risk Does the rising debt

increase short-term risk to financial stability?

Research Questions Model Strategies

1

Regression Analyses To what extent, does the debt surge hurt

consumption?

2

Microsimulation How big is the threat associated with the potential rate hike?

3

Research Questions, Model Strategies and Data Construction

8

Debt at Risk Does the rising debt

increase short-term risk to financial stability?

Research Questions Model Strategies Data Construction

SES Surveys: 2009, 2011 and 2013

BOT’s supplemental survey to SES 2013Q1. It contains questions about household’s debt-servicing difficulties.

Best available resources on income and debt for Thailand’s middle- and low-income households

Allows focusing on the evolving structure of Thailand’s household debt instead of cyclical effects

1

Regression Analyses To what extent, does the debt surge hurt

consumption?

2

Microsimulation How big is the threat associated with the potential rate hike?

3

• Does the rising debt increase short-term risk to financial stability?

• To what extent, does the debt surge hurt consumption?

• Policy Implications

– How big is the threat associated with the potential rate hike?

– How do we live with highly leveraged households?

Agenda

9

US experience suggests that distribution of debt matters

10

US mortgage credit growth for house purchase by credit-score zip codes, 1999-2006 Unit: Index (1999 = 100)

Source: Mian and Sufi (2014)

There was an aggressive credit expansion among marginal borrowers in the years preceding the Great Recession

Debt growth has been higher among groups perceived as having lower risks

11

Debt growth (2013 vs. 2009) by education groups

Annualized growth rates (%)

Source: SES

Debt Share 64% 36%

6.16.2

3.2

5th quintile 4th quintile Bottom 3 quintiles

Debt growth (2013 vs. 2009) by income groups

Annualized growth rates (%)

Debt Share 17% 18%

2013 Income 0-21,000 21,000-33,000 >33,000

65%

11.4

3.0

College and above

No college

Higher debt growth among middle-income groups

The expansion of credit among middle-income groups also results in increasing shares of those with moderate debt-servicing burden

12

Distribution of debt by DSR groups Unit: %

Source: SES

35% 35% 34%

32% 36% 37%

15%16% 15%

6% 5% 7%11% 6%8%

100%

60%<DSR<=80%

DSR <= 20%

20%<DSR<=40%

40%<DSR<=60%

DSR>80%

2013 2011 2009

Rising share of debts among households with moderate DSR level

What determines households’ financial vulnerabilities?

13

Financial Vulnerabilities

Lifetime income

Occupation

DSR

Assets

Age

Probability of Debt-Servicing Difficulty

Probit Model

The “Debt at Risk” measure summarizes the health of our stock of household debt

14

Debt concentrated among high-risk households

Debt (baht)

Prob. Debt at Risk (baht)

A 80 50% 40

B 20 10% 2

Debt at Risk (% of total) 42%

1 Debt concentrated among low-risk households

Debt (baht)

Prob. Debt at Risk (baht)

C 20 50% 10

D 80 10% 8

Debt at Risk (% of total) 18%

2

Each household

Debt at Risk (DAR)

Outstanding debt

Probability of Debt-Servicing Difficulties

Summing DAR across households

Aggregate DAR

Aggregate Debt at Risk has been declining gradually since 2009

15 Source: SES

Recent surge in household debt has NOT been associated with an

increase in short-term risk to financial stability

1

Debt at Risk, 2009-2013 % of total debt

20.4

22.3

2011 2013 2009

23.4 Debt has been less

concentrated among risky households

28

18

28

1615

GHB&GSB Comm. Banks

Village Funds

Savings Co-Op

BAAC

Debt at Risk across different lenders, 2013

Unit % of Total Debt

Debt Share 22% 18% 14% 11% 11%

Not all lenders are equal in term of short-term risk to financial

stability

2

• Does the rising debt increase short-term risk to financial stability?

• To what extent, does the debt surge hurt consumption?

• Policy Implications

– How big is the threat associated with the potential rate hike?

– How do we live with highly leveraged households?

Agenda

16

Although the surge in debt does not increase short-term risk, it can restrain economic growth

17

Rising Household Leverage

Restraining Economic Activity

By Restricting

Consumption

Higher debt repayment burden

Less access to credit

More financially vulnerable

17

Thailand experienced a wide variation of debt growth from 2009 to 2011

18

Change in DSR by provinces from 2009 to 2011

Source: SES

27

-23

-1Medium debt growth

High debt growth

Low debt growth

Splitting Thai districts into 3 groups based on their DSR increase

Average change in DSR (Percentage point)

เชียงราย

เชียงใหม่

แม่ฮ่องสอน

กาญจนบุรี

ราชบุรี

พะเยา

น่าน ล าปาง

ล าพูน

ตาก

แพร่ อุตรดิตถ์

สุโขทัย

ก าแพงเพชร พิจิตร

พิษณุโลก

เพชรบูรณ์

นครสวรรค์ อุทัยธานี

ลพบุรี ชัยนาท

เลย

หนองคาย บึงกาฬ

นครพนม

สกลนคร อุดรธานี

หนองบัวล าภู

ขอนแก่น

ชัยภูมิ

นครราชสีมา

บุรีรัมย์ สุรินทร์ ศรีสะเกษ

ร้อยเอ็ด

ยโสธร

มุกดาหาร

อ านาจเจริญ

อุบลราชธานี

มหาสารคาม

ปราจีนบุรี

สระบุรี

กาฬสินธ์ุ

นครนายก

กรุงเทพมหานคร สมุทรปราการ

สมุทรสาคร สมุทรสงคราม

นครปฐม อ่างทอง

นนทบุรี สิงหบุรี

อยุธยา ปทุมธานี

สุพรรณบุรี

สระแก้ว ฉะเชิงเทรา

ชลบุรี

ระยอง จันทบุรี

ตราด

เพชรบุรี

ประจวบคีรีขันธ์

ชุมพร

ระนอง

พังงา

ภูเก็ต กระบี่

สุราษฏร์ธานี

นครศรีธรรมราช

พัทลุง ตรัง

สตูล สงขลา

ปัตตานี

ยะลา นราธิวาส

Greater DSR increase

We use this variation in DSR increases across 776 districts to examine the effect of debt acceleration on consumption during the

subsequent period.

The acceleration of debt burden represents the key impediment to consumption growth

Note: *The predicted mean is computed using average values of DSR increases in each district group.

Source: SES

Low debt growth

7.8

Medium debt growth

6.7

High debt growth

5.4

The income growth of 10% yields the consumption growth of:

Illustration of the regression results

Our analysis shows that recent weakness in consumption is closely related to lagged acceleration of household debt

The coefficient on lagged DSR increase is negative and statistically significant.

Regression Analysis Result Predicted mean of non-durable consumption growth*

19

When is household leverage financial burden to Thai households?

20

Predicted probability of having debt-servicing difficulty

The critical DSR threshold is 40%. Above this level, households in all occupations exhibit a significant increase in the probability of having difficulty paying their debt.

Source: SES

23

2423

21

27

2928

25

35

3735

3334

3534

32

37

3837

34

20

25

30

35

40

Agri bus Salaried Retired Non-agri bus

DSR<20% DSR>=80% 20%<=DSR<40% 60%<=DSR<80% 40%<=DSR<60%

Who are these financially vulnerable households?

21

Not necessarily poor people Concentrated in the Northeast

Have less liquid asset cushion but more fragile income

Source: SES

24%

20%

18%

19%

19% Q1

Q2

Q3

Q4

Q5

Poorest

Vulnerable households by income quintile

Unit % of Total

3% 13%

21%

54%

9% Bangkok

Central

North

Northeast

South

Vulnerable households by region

Unit % of Total

Vulnerable Not Vulnerable

Median liquid asset coverage

Share of fragile income

4.8 16.8

73% 58%

• Does the rising debt increase short-term risk to financial stability?

• To what extent, does the debt surge hurt consumption?

• Policy Implications

– How big is the threat associated with the potential rate hike?

– How do we live with highly leveraged households?

Agenda

22

We construct a microsimulation model to help assess the impact of the interest rate hike on households’ financial health.

Microsimulation Framework

Income Debt expense

Heterogeneous income growth across households Assume average income

growth consistent with BOT’s projected GDP growth

Allow variance of income growth to differ across income groups

Interest rate = MRR + Risk premium

Debt with fixed rate:

Consumption, Mortgage, Non-BACC Agriculture

Debt growth as function of income growth Construct a pseudo-panel

dataset with 48 household clusters based on education, age, occupation and location

23

24

Share of variable-rate debt

Q1 27%

Q5 11%

2.0%

Q4 Q5

2.1% 2.3%

2.9%

Q3

4.4%

Q1 Q2

Notes: 1. The microsimulation uses 2013Q1 SES observations as the base dataset. 2. Indebted households only Source: SES

Increase in average DSR by income quintiles Unit: Percentage point

Poorest households

Richest households

Our microsimulation indicates low-income households will be more burdened by the interest rate hike.

<9 9-14 14-21 21-33 >33 Income Brackets (Thousand baht)

25

Q2 Q3

3.3%

1.3%

Q4

3.1%

Q5

2.5%

3.7%

Q1

Notes: 1. The microsimulation uses 2013Q1 SES observations as the base dataset. 2. Indebted households only Source: SES

Increase in share of financially vulnerable households

Unit: % of indebted households

The rate hike will also threaten households who are “almost” financially vulnerable to begin with

Share of “almost” vulnerable households

Q1

4.8% Q5

3.0%

Poorest households

Richest households

<9 9-14 14-21 21-33 >33 Income Brackets (Thousand baht)

• Does the rising debt increase short-term risk to financial stability?

• To what extent, does the debt surge hurt consumption?

• Policy Implications

– How big is the threat associated with the potential rate hike?

– How do we live with highly leveraged households?

Agenda

26

Living with highly leveraged households

27

Falling prices and

wages

Debt repayment takes larger share of income

Spending falls

Stagnant economy

Be vigilant on debt deflation

Monetary Policy • Keep inflation expectation

steady.

• Deflation could be as

dangerous as high inflation.

27

Living with highly leveraged households

28

Reduce households’ downside risk

Access to insurance products

Health

Life

Accident

Disaster

Source: Dailynews

28

Living with highly leveraged households

29

Increase financial awareness

Technology can significantly simplify financial planning and budgeting

Source: CNET.com 29

Living with highly leveraged households

30

Manage the buffer

0

20

40

60

80

100

120

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

Crisis

Household debt

US household debt and public debt % of GDP

Public debt

0

20

40

60

80

100

120

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

Household debt

Thailand household debt and public debt % of GDP

Public debt

Fiscal space must be carefully managed as it is the ultimate shock absorber

Source: Federal Reserve Bank of St Louis, PDMO, Bank of Thailand, NESDB 30

Key Takeaways

31

- No. The debt has been rising primarily among lower-risk households

Does the rising debt increase short-term risk

to financial stability?

1

- Evidence shows the debt burden hinders consumption growth

- DSR>40% could hurt household’s consumption

To what extent, does the debt surge hurt

consumption?

2

- Poor households are likely to see disproportionate increase in DSR

- Impacts on consumption will not limited to poor households

How big is the threat associated with the potential rate hike?

3

32

End of document