Payment Hierarchy Analysis - A Study of Changes in Consumer … · 2015-12-04 · Payment Hierarchy Analysis: A Study of Changes in Consumer Payment Prioritization from 2007 through

TRANSUNION WHITE PAPER

Payment Hierarchy Analysis:A Study of Changes in Consumer Payment

Prioritization from 2007 through 2011

Matthew KomosSenior ConsultantAnalytic Services

Sean ReardonSenior Consultant

Business DevelopmentFinancial Services

Charlie WiseDirector

Research and ConsultingFinancial Services

Ezra BeckerVice President

Research and ConsultingFinancial Services

1 Overview

2 Analytical Detail

Data

Timeframe

Analytical Focus

3 Analytical Results

6 Subsequent Analysis–Auto Loan Hierarchy

10 Drivers of Performance

Traditional Hierarchy

Falling Home Values

Rising Unemployment

Deal Structure

Timing of Consequence

Number of Satisfied Accounts

Changing Purchasing Habits

Changes in Social Stigma

A First Step toward Quantifying These Drivers

15 Summary

© 2012 TransUnion LLC All Rights Reserved

No part of this publication may be reproduced or distributed in any form or by any means, electronic or otherwise,now known or hereafter developed, including, but not limited to, the Internet, without the explicit prior written consent from TransUnion LLC.

Requests for permission to reproduce or distribute any part of, or all of, this publication should be mailed to: Law DepartmentTransUnion555 West Adams Chicago, Illinois 60661

The “T” logo, TransUnion, and other trademarks, service marks, and logos (the “Trademarks”) used in this publication are registered or unregistered Trademarks of TransUnion LLC, or their respective owners. Trademarks may not be used for any purpose whatsoever without the express written permission of the Trademark owner.

transunion.com/business

TransUnion has published delinquency statistics for mortgages, creditcards and auto loans on a quarterly basis since before the beginning of the most recent recession. An analysis of those quarterly mortgage and credit card statistics over the course of the recession and beyond has revealed an interesting dynamic: namely, that serious mortgagedelinquency had increased over the course of the recession, and hasremained persistently at high levels since then, while serious credit card delinquency had declined and remained well controlled through the same time period, hitting record lows in 2011 (Figure 1). This findingseemed to conflict with the conventional wisdom which held that, whenforced to prioritize debt obligations, consumers would typically pay theirsecured obligations first. In other words, there were indirect indicationsthat the traditional payment hierarchy was no longer universal.

To gain insight into this apparent anomaly, TransUnion examined the comparative differences in delinquency rates for bankcards and mortgages across all risk segments over several years. The results of the initial study provided insights based on a number of factors, including geography and consumer risk profile, and indicated that there was a new payment-delinquency paradigm being observed. Ourupdate of the study through Q4 2011 has confirmed that this has been a sustained shift in consumer preference. In addition, we have providedadditional analysis to introduce auto loans into the payment dynamicalongside mortgages and bankcards, to provide a more complete understanding of consumer payment behavior and preferences through 2011.

Ser

ious

Del

inqu

ency

Per

cent

age

National Bankcard and Mortgage Serious Delinquency Rates

Source: TransUnion Trend Data

% OF CONSUMERS 60+ DPD ON MORTGAGES % OF CONSUMERS 90+ DPD ON BANKCARDS

2%

4%

8%

6%

Q1 2006

Q1 2007

Q1 2008

Q1 2009

Q1 2010

Q1 2011

Q4 2011

FIGURE 1

Recession

Data

The source of the underlying data usedfor this analysis was the TransUnionTrend Data® solution, a proprietary historical database consisting of approximately27 million anonymous consumer recordsrandomly sampled every quarter fromTransUnion's national consumer creditdatabase, with history dating back to1992. These randomly generated quarterlyfiles contain more than 200 credit variables, which were used to create custom attributes for this analysis. Using standard TransUnion definitions of bankcard and mortgage products, wewere able to create specific characteristicsto evaluate consumer delinquency statusacross products for the purpose of this study.

Each consumer’s record containsa calculated TransUnion AccountManagement Score 2.0 (a statisticallyderived credit risk score), a state identifier, and a Core-Based StatisticalArea (“CBSA”) identifier. The ability to examine consumer- and tradeline-specific attributes allows TransUnion to display account and balance distributions and performance by the various metrics mentioned above (e.g., it allows for the analysis of performanceon open bankcards by select states).

Given the nature of the data source andthe chosen sampling method, each timeperiod examined contains a unique set ofconsumers. Therefore, this analysis doesnot follow specific accounts over time,but rather compares the performance of accounts with similar definitions over the specified time periods.

Timeframe

The initial analysis examined quarterlydata from Q2 2007 through Q4 2009;TransUnion has subsequently updatedthe original findings to extend the analysis through Q4 2011 (Figure 2).This time period is particularly interestinggiven the onset of the recession, whichbegan in the sub-prime mortgage space.This overall view allows insight into the payment hierarchy of bankcard andmortgage consumers over an extendedtime frame, in order to determine ifthe recent conditions in the market havecontinued to affect how consumers paytheir bankcard and mortgage obligations.The incorporation of auto loan data intothe analysis began with Q1 2011 andwent through the end of the year.

Analytical Focus

The initial analysis focused on the following:

Consumers with at least one bankcard trade

and at least one mortgage trade;

Examining bankcard and mortgage delinquency

rates for this population each quarter;

The percentage of consumers delinquent on

a bankcard but not on a mortgage, compared

to the percentage of consumers delinquent

on a mortgage but not on a bankcard, and

the percentage of consumers delinquent on

both, over the period of study.

Analytical Detail

Payment Hierarchy Analysis: A Study of Changes in Consumer Payment Prioritization from 2007 through 2011

© 2012 TransUnion LLC All Rights Reserved2

National Bankcard and Mortgage 30+ DPD Delinquency Rates



6

4

2

8

10

FIGURE 2

30

+ D

PD

Del

inqu

ency

Per

cent

age

Q2 2007

Q3 2007

Q4 2007

Q1 2008

Q2 2008

Q3 2008

Q4 2008

Q1 2009

Q2 2009

Q3 2009

Q4 2009

Q1 2010

Q2 2010

Q3 2010

Q4 2010

Q1 2011

Q2 2011

Q3 2011

Q4 2011

Recession

07 0707 0707 07 0807 08 0808 8 0808 08 0808 08 0908 09 0909 9 0909 9 0909 09 1009 0 1010 0 1010 10 1010 10 1110 1 1111 1 1111 11 1111 111

3© 2012 TransUnion LLC All Rights Reserved


The results of the analysis confirm whatTransUnion has reported in terms ofdelinquency at a more general levelthrough our quarterly press releases.Specifically, mortgage 30+ days past due (DPD) delinquencies increased for ten consecutive quarters through

Q4 2009. Since then they have been on a slowly decreasing trend, with modestupticks in Q3 and Q4 2011 (Figure 3). At the same time, bankcard 30+ DPDdelinquencies have been on a decliningtrend since their recent peak in Q4 2007(at the onset of the recession).

This trend is similarly present in the subprime risk segment, albeit atmuch higher delinquency rate levels. The mortgage delinquency rate reached a peak of 42.4% among subprime consumers in Q4 2009 and has seen only modest improvement since then,while bankcard delinquency rates haveimproved since the pre-recession peak in Q4 2007.

The central dynamic explored in thisstudy is the observed preference amongconsumers with constrained liquidity topay their credit card debt service ratherthan their mortgage debt service. In Q1 2008, we see a reversal in preference priority: for the first time, the percentageof consumers current on bankcards butdelinquent on mortgages (a reversal of the traditional payment hierarchy) surpassed the percentage of consumerscurrent on mortgages but delinquent on bankcards (the traditional paymenthierarchy) (Figure 4). This “flip” is representative of the change in the conventional wisdom around the payment hierarchy, or which debt obligations consumers would choose to pay first.

The results following the official end of the recession in Q2 2009 show that,contrary to what some in the industrybelieved should have been a reversionback to the conventional payment hierarchy, the variance in payment preferences not only continued but actually increased. Analysis shows thepercentage of consumers delinquent on their mortgages but current on theirbankcards rising to a peak of 7.40% in Q2 2010 (from 4.26% in Q1 2008),

National Subprime Bankcard and Mortgage 30+ DPD Delinquency Rates



15

10

5

20

25

30

35

40

FIGURE 3

30

+ D

PD

Del

inqu

ency

Per

cent

age

Q2 2007

Q3 2007

Q4 2007

Q1 2008

Q2 2008

Q3 2008

Q4 2008

Q1 2009

Q2 2009

Q3 2009

Q4 2009

Q1 2010

Q2 2010

Q3 2010

Q4 2010

Q1 2011

Q2 2011

Q3 2011

Q4 2011

Recession

07 0707 0707 07 0807 08 0808 08 0808 08 0808 8 0908 09 0909 09 0909 09 0909 09 1009 0 1010 0 1010 10 1010 10 1110 1 1111 1 1111 11 1111 1111

Consumer Delinquency Preferences

Source: TransUnion Consumer Database

2%

4%

8%

6%

Q1 2006

Q1 2007

Q1 2008

Q1 2009

Q1 2010

Q1 2011

Q4 2011

FIGURE 4

Recession

% OF CONSUMERS 30+ DPD ON MORTGAGES BUT CURRENT ON BANKCARDS% OF CONSUMERS CURRENT ON MORTGAGES BUT 30+ DPD ON BANKCARDS% OF CONSUMERS 30+ DPD ON MORTGAGES AND 30+ DPD ON BANKCARDS

Perc

enta

ge o

f Con

sum

ers

Analytical Results

4 © 2012 TransUnion LLC All Rights Reserved

while the percentage of consumers delinquent on their bankcards but current on their mortgages decreased to a low of 2.61% in Q2 2011 (from4.16% in Q1 2008). These metrics are at 6.86% and 2.72%, respectively, as of the end of Q4 2011, demonstratingthat this payment hierarchy reversal persists 2½ years after the official end of the recession.

The same trend is evident within thesubprime risk segment (Figure 5).However, it should be noted that the flip in payment hierarchy occurred earlierin this segment, during Q4 2007 ascompared to Q1 2008 for the total market. In addition, the magnitude of delinquency in this segment is, notsurprisingly, significantly higher thanthe total market: subprime consumerswho were delinquent on a mortgage but current on all bankcards rose bymore than 60% from 19.10% in Q42007 to a peak of 31.1% in Q3 2010,and has remained above 30% throughQ4 2011. Over that same time period,the percentage of subprime consumersdelinquent on at least one bankcard but current on all mortgages decreasedby 32% from 18.10% in Q4 2007 to12.32% in Q3 2010, and has continuedto decline to 11.19% in Q4 2011.

Thus it is evident that, while the magnitude of this payment hierarchyreversal has lessened somewhat through2011, the dynamic clearly persists forconsumers overall as well as acrossconsumer risk score tiers.

When examining payment hierarchytrends at a state level for select states

such as California and Florida, the resultsare similar and pronounced. WithinCalifornia, the percentage of consumersdelinquent on their mortgages but current on their bankcards was 3.30% in Q2 2007, which was identical to the3.30% of consumers who were delinquenton their bankcards but current on theirmortgages in that quarter (Figure 6). As seen previously, the percentage of consumers delinquent on their mortgages

while current on their bankcardsincreased dramatically from that quarter,while the percentage of consumers delinquent on one or more bankcardswhile current on their mortgagesdeclined. The difference between thosedelinquency rates peaked in Q3 2010,and while it has contracted somewhatsince then, in Q4 2011 the percentage of consumers delinquent on their mortgage but current on their bankcards

Subprime Consumer Delinquency Preferences


10%

20%

30%

Q1 2006

Q1 2007

Q1 2008

Q1 2009

Q1 2010

Q1 2011

Q4 2011

FIGURE 5

Recession


Perc

enta

ge o

f Con

sum

ers

California Consumer Delinquency Preferences


2%

4%

6%

8%

10%

Q1 2006

Q1 2007

Q1 2008

Q1 2009

Q1 2010

Q1 2011

Q4 2011

FIGURE 6

Recession


Perc

enta

ge o

f Con

sum

ers


stood at 8.82% compared to 2.24%for consumers who were delinquent on their bankcards but current on their mortgages.

The results in Florida (Figure 7) show a similar pattern of consumer behavior,with the peak in the percentage of consumers delinquent on their mortgagesbut current on their bankcards at14.63% occurring in Q3 2010, andslowly improving since then to 13.84%in Q4 2011. Meanwhile, the percentageof Florida consumers delinquent on oneor more bankcards while current on theirmortgages stood at 2.88% in Q4 2011.

Within the subprime segments inCalifornia and Florida, the results reflectthe delinquency trends at the state leveloverall, although the magnitude ofdelinquency is significantly higher across the board. In addition, the trend of a gradual improvement in the percentage of consumers delinquenton their mortgages while current ontheir bankcards holds in the subprime segment in California: following a peak of 42.50% in Q1 2010, this statistic has dropped to 35.71% in Q4 2011. Yet even at that level, itremains nearly four times the 9.02% of subprime California consumers delinquent on one or more bankcardswhile current on their mortgages (Figure 8). In contrast, Florida has not seen a similar drop: since a peak of 43.02% in Q3 2010, the percentageof subprime Florida consumers delinquent on their mortgages while current on their bankcards has notdropped below 42% (Figure 9).


Florida Consumer Delinquency Preferences


2%4%6%8%

10%12%14%16%

Q1 2006

Q1 2007

Q1 2008

Q1 2009

Q1 2010

Q1 2011

Q4 2011

FIGURE 7

Recession


Perc

enta

ge o

f Con

sum

ers

Perc

enta

ge o

f Con

sum

ers

California Subprime Consumer Delinquency Preferences


Q1 2006

Q1 2007

Q1 2008

Q1 2009

Q1 2010

Q1 2011

Q4 2011

FIGURE 8


10%

20%

30%

40%

50%

Recession

Perc

enta

ge o

f Con

sum

ers

Florida Subprime Consumer Delinquency Preferences


Q1 2006

Q1 2007

Q1 2008

Q1 2009

Q1 2010

Q1 2011

Q4 2011

FIGURE 9

Recession


10%

20%

30%

40%

50%


As part of the update to the payment hierarchy study, TransUnionincorporated data related to auto payment preferences into the analysis of mortgages and bankcards. Traditional(pre-recession) consumer payment preferences would typically put the priority of paying an auto loan between that of credit card and mortgage payments. The purpose of the expanded analysis was to uncover the true payment preference in the current environment for consumers who have all three loan types. It should be noted that this initial analysisdid not include a separate study of consumers who held one or more auto loans and one or more bankcardsregardless of mortgage status; that analysis is under consideration for a separate follow-up study.

The analysis focused on the following:

Consumers with at least one bankcard trade,

at least one mortgage trade, and at least

one auto trade;

Examining bankcard, mortgage, and auto

loan delinquency rates for this population;

The percentage of consumers delinquent on

only one product, only two of those products,

or on all three of those products.

It is clear from Figure 10 that the percentage of consumers 30+ DPD on a specific trade was highest for mortgage trades and lowest for auto

trades, with bankcard trades in the middle, throughout 2011.

Another way to examine this dynamic is to explore the conditional delinquencyrates by product among those consumerswho had gone delinquent. In otherwords, given a group of consumers who were financially constrained and had to make a difficult choice, whatproducts did they choose to pay and

6 © 2012 TransUnion LLC All Rights Reserved

Subsequent Analysis–Auto Loan Hierarchy

Study of Consumers Holding All Three Loan Types–Q4 2011FIGURE 11

71.9% 43.3% 30.2%

MORTGAGE% of delinquent consumers 30+ DPD on mortgage

BANKCARD% of delinquent consumers 30+ DPD on bankcard

AUTO% of delinquent consumers 30+ DPD on auto loan


Perc

enta

ge o

f Con

sum

ers

Percentage of Consumers 30+ DPD on Mortgage, Bankcard, and Auto


Q1 2011

Q2 2011

Q3 2011

Q4 2011

FIGURE 10

% OF CONSUMERS 30+ DPD ON MORTGAGES % OF CONSUMERS 30+ DPD ON AUTO% OF CONSUMERS 30+ DPD ON BANKCARD

2%

4%

6%

8%

10%

12%



what products did they choose notto pay? Figure 11 (on page 6) showsclearly that most consumers who wentdelinquent in Q4 2011 on at least one of the three primary products chose to do so on their mortgages (71.9%), followed by credit cards (43.3%) andleast likely on their auto loans (30.2%).

Digging deeper into this dynamic, it is useful to look at the first choice ofdelinquency. In other words, when a consumer begins to run out of funds,what is the first payment he chooses

most often to miss? Now, the structure of the study does not allow for panelanalysis, i.e., following specific consumersover time. However, we can evaluate thedelinquency rates for those consumerswho are delinquent on only one of thethree products to get some sense of thisfirst choice dynamic. Figure 12 showsthat among delinquent consumers in Q4 2011:

Only 9.5% were delinquent on an auto

loan while current on their credit cards

and mortgages

17.3% were delinquent on a credit card while

current on their auto loans and mortgages

39.1% of all consumers in the study were

delinquent on a mortgage while current on

their auto loans and credit cards

This would indicate that, of those consumers that have a bankcard, auto,and mortgage trade, consumers are farmore likely to default first on a mortgagethan on any of the other two trade types,and almost twice as likely to default on a bankcard, rather than on an auto loan.


Study of Consumers Holding All Three Loan Types–Q4 2011FIGURE 12

MORTGAGE ONLY% of delinquent consumers 30+ DPD ona mortgage while current on bankcards and auto

MORTGAGE AND AUTO% of delinquent consumers 30+ DPD on both mortgage and auto while current onbankcards

AUTO ONLY% of delinquent consumers 30+ DPD on an auto loan while current on mortgage and bankcards

MORTGAGE, BANKCARD AND AUTO% of delinquent consumers 30+ DPD on mortgage, bankcard and auto

AUTO AND BANKCARD% of delinquent consumers 30+ DPD on both auto and bankcard while current on mortgage

BANKCARD ONLY% of delinquent consumers 30+ DPD on a bankcard while current on their mortgage and auto

MORTGAGE AND BANKCARD% of delinquent consumers 30+ DPD on both mortgage and bankcard while current on auto

39.1%

11.1%

17.3%

9.5%

8.2%

13.5%

1.4%



Clearly, the payment preference for consumers with all three loan types is to prioritize auto loans ahead of theirmortgages and bankcards. This can befurther examined when looking at the percentage of consumers who are delinquent on two of the loan typeswhile remaining current on the thirdloan type, to identify consumers’ preferences for the last loan type onwhich they are likely to go delinquent.

13.5% were delinquent on a credit card and

mortgage while current on their auto loans

8.2% were delinquent on a mortgage

and auto loan while current on their

credit cards

1.4% were delinquent on a credit card and

auto loan while current on their mortgage

The payment behavior demonstrated bythose consumers in 2011 who have allthree loan types is to prioritize their autoloan payments ahead of either bankcardsor mortgages. This trend is evident notonly at the national level but across all 50states. Figures 13 and 14 show two

examples at the state level: Figure 13shows the distribution of choice amongdelinquent consumers in California, oneof the states hardest hit by the recessionand its aftermath, while Figure 14 (onpage 9) shows the statistics forMinnesota, one of the most stable statesfrom a delinquency perspective and onethat was relatively less affected by thedown economy.


Study of California Consumers Holding All Three Loan Types–Q4 2011FIGURE 13








47.0%

13.4%

11.6%

6.4%

6.7%

13.8%

1.1%


© 2012 TransUnion LLC All Rights Reserved 9

A few likely reasons why auto loans havebecome the preferred payment includethe need for a car to get to work or lookfor employment, and the fact that anauto loan is not a revolving loan. Theimpact of repossession is greater than the loss of a credit card, and the utility of an auto loan (access to a vehicle andtransportation) is much more difficult to replace than the utility of a credit card (access to a purchasing tool and

short-term credit). In addition, consumers may have equity in their vehicles after several years of payments that they are looking to preserve.Meanwhile, negative equity has becomeincreasingly common for homes, whichhas contributed to the shift in paymentpreference away from mortgages. Theseand other drivers of the payment hierarchyreversal shall be more fully discussed inthe next section.


Study of Minnesota Consumers Holding All Three Loan Types–Q4 2011FIGURE 14








38.0%

10.7%

18.9%

8.2%

7.0%

15.9%

1.3%



It is clear that the traditional hierarchy of payment and delinquency haschanged. The next question is to try to understand why this occurred, and to evaluate how long this reversal willremain in effect—if, in fact, it is not apermanent change. To gain insight intothe reversal of the payment hierarchy, we first provide a brief discussion of the drivers of the traditional paymentand default hierarchy. It is through an understanding of how those forces havechanged that we understand why thepayment hierarchy itself changed.

Traditional Hierarchy

The traditional default hierarchy is commonly described as follows: if a consumer ran into liquidity problems,the first debt he would stop paying is his credit card. If liquidity constraintscontinued, the consumer would nextstop paying his auto loan. Only in themost dire liquidity crisis would the consumer stop paying his mortgage. The drivers of this hierarchy included the following.

Competition in the credit card space was fierce,

and credit cards were readily available. Having

one credit card closed due to non-payment

might make access to further credit card credit

more expensive, but would not generally

reduce that access materially.

Alternate payment methods such as cash,

checks or debit cards could act effectively

as substitutes for credit cards; they could not

act effectively in place of a car as a means of

transportation or a home as a means of shelter.

There was generally no physical consequence

of credit card default. In other words, a negative

mark on one’s credit report and perhaps more

expensive credit card access for a period of

time was the price of card default. In contrast,

auto loan default leading to repossession

would force the consumer to use public

transportation or not be able to get to work,

grocery stores or other required destinations,

or use a car to seek employment. As well,

foreclosure would lead to the necessity of

finding an alternative place to live, with the

consequent uprooting of the family, changing

schools, etc.

The traditional down payment for a home

was 20%. Most consumers had to work hard

and save for years to amass a sufficient down

payment to purchase a home. Thus, home

ownership was both a financial achievement

and a psychological achievement.

Home values generally increased over time.

Homes were considered an appreciating asset

and, taking into consideration the large down

payment required to buy the home, a significant

investment. In contrast, credit cards were usually

considered a convenience at best, or even a

status symbol in some cases.

The social stigma of foreclosure was significant.

One could hide from one’s neighbors the fact

of credit card default; it was more difficult to

hide the fact of automobile repossession, and

virtually impossible to hide the fact of foreclosure.

In short, there were many compellingreasons the average consumer wouldchoose to default first on a credit cardand only as a last resort on a mortgage.Let us now consider what changes tothese drivers might contribute to achange in payment hierarchy.

Falling Home Values

Perhaps the single biggest contributorto the change in payment hierarchy andthe shift in the mortgage from the firstpriority to the last priority is the drop in housing prices. Once considered anappreciating asset, in most cases homevalues have dropped significantly fromtheir peak in 2006. In fact, it can beargued that home value depreciation was a key cause of the recession. In manycases, homeowners owe more on theirmortgages than their homes are worth, a situation often referred to as being“underwater.” In the most severe cases, it is doubtful that home values will riseto the levels they hit prior to the collapseof the housing bubble in any reasonabletime horizon.

Thus, for many consumers, homes are no longer perceived as appreciating, significant assets and investments, butrather as liabilities with no near-termfinancial value. For these consumers,there is no longer an equity-based motivation for paying a mortgage beforeother debt obligations. In fact, in manycases, it became financially advantageousfor consumers with homes significantlyunderwater to stop payment on theirhomes and simply walk away and make a fresh start, rather than pay their lender the sizable differencebetween their loan amount and what they would receive from a sale.

In contrast, the used auto market is as strong or stronger today than it hasever been. Consumers find themselveswith significant auto equity, an assetposition once held by the home. It maytherefore be asserted that consumers

Drivers of Performance



have shifted their payment priorities to maximize equity protection whereverit may be found.

Rising Unemployment

Not surprisingly, unemployment is theprimary driver of credit delinquency.Consumers without a steady income arehard-pressed to meet their debt serviceobligations, particularly once any savingsor investment cushion is exhausted. The unemployment rate reached doubledigits nationally during this recession,with many areas of the country facingsevere job losses. In Q4 2010, the unemployment rates in Florida andCalifornia hit 11.9% and 12.5%, respectively, and in Michigan it peaked in Q3 2009 at 14.1%.1 At the nationallevel, unemployment remains stubbornlyhigh at 8.2% as of the printing of this paper.

In addition, the duration of unemployment has more than doubledthrough the recession and beyond, froman average of 16.2 weeks in Q1 2008 to an average of 40.8 weeks as of Q42011.2 The consensus among economistsis that job recovery will continue to besluggish through 2012. Faced withunemployment, or uncertainty in one’semployment status, it is natural toattempt to preserve available credit card credit as a cushion for difficulttimes. The fear of unemployment, andthe harsh reality of being unemployed,has forced many consumers to focus onthe immediate need for liquidity to purchase groceries, clothing, medication,and other necessities. As well, the

protection of the automobile has becomeparamount, as employed consumers wishto maintain their ability to commute towork, and unemployed consumers usetheir cars to seek employment andremain more attractive to potentialemployers.

Deal Structure

Before the recession, the advent andincreasing ubiquity of low-interest loansrequiring little or no down payment had made home ownership much more widely available.3 As well, the (erroneous)anticipation that home values would continue to rise, and competitive pressures among lenders, led to theacceptance of increasingly higher loan-to-value ratios for loans booked.

Consumers no longer had to amass aconsiderable down payment to becomehomeowners. This fact, in conjunctionwith instant loan approval processes, low-documentation or no-documentationrequirements, generous closing costallowances, and expansive underwritingpolicies meant it took very little effort to obtain a mortgage. As well, the optionto pay only the interest generated by amortgage, or indeed in some cases evenless than that, made managing the debtservice for these loans much easier.Consumers could purchase much morehouse than they would normally be ableto afford under full amortization and rising interest rates.

Unfortunately, the prevalence of thesedeal structures had unintended negativeconsequences. Without having to put

forth a great deal of effort over severalyears to amass a large down payment,many borrowers were neither financiallyinvested in their homes to any significantextent, nor psychologically invested in home ownership—in some sense,achieving home ownership became tooeasy, which in turn made it much easierto forsake the mortgage obligation infavor of maintaining other credit relationships.

Timing of Consequence

Naturally, failure to make on-time, regular payments carries consequences.For credit cards, the consequence is theloss of the use of the card; for auto loans,the consequence is repossession; and for mortgages, the consequence of non-payment is foreclosure. Yet the timing of consequence is far different for the three products. In most cases,credit card lenders will close an accountand bar access to the credit line after a90 days of delinquency, and frequentlysooner than that. As well, auto repossessionusually occurs soon after 90-day delinquency. In contrast, the foreclosureprocess is usually quite lengthy, often taking more than a year from the initialdelinquency—during which time, thehomeowner still has the use of the home.Thus the impact of non-payment ismuch more immediate for auto loans and credit cards. Since the choice ofwhich debt relationship on which todefault is often driven by short-term liquidity constraints, the timing of consequence can play a significant role in the prioritization of default.


Number of Satisfied Accounts

Generally, a single mortgage payment is significantly larger than the minimumpayment required on a given credit card or auto loan. Indeed, for most consumers, the mortgage payment is the largest single debt service of themonth, often several times larger thanthe next biggest bill. Hence, a borrowercould use a single mortgage payment to satisfy the auto loan and several creditcard minimum payments. In this case,the borrower only has to deal with a single default relationship instead ofmany, and the consequent volume ofdunning calls and letters is much lower.

Changing Purchasing Habits

For a long time, credit cards were primarily a convenience, a means of easing the disparity between income and expense, and hence a way of makingsignificant purchases more facile. Usually,any merchant that accepted a credit cardwould also gladly accept cash. For many,credit cards were a symbol of status: thegold card versus the standard card, theplatinum card versus the gold card, and so forth. Yet the advent of Internet shopping, and the manner in which consumers have embraced that channel,has made credit cards a necessity toaccess certain purchasing channels. Aswell, credit and debit cards are becomingthe preferred payment vehicles for somemerchants. For example, in June 2009American Airlines began phasing out theacceptance of cash for in-flight purchases,and today the “cashless cabin” is the standard on all American Airlines flights,a policy that other major airlines havealso adopted.4

In addition, it is exceedingly difficult to rent a car without a major credit card. Indeed, cash is no longer king.Thus, credit cards have gone from a convenience to a purchasing necessity.

Changes in Social Stigma

Simply stated, the recession has reduced the social stigma associated with foreclosure. There is abundant literature that discusses this change, both scholarly and in more mainstreamgeneral media. In fact, the phrase “strategic default” was coined during this past recession to describe the conscious choice of a consumer todefault on a mortgage. While we wouldargue that every default is, in a sense,strategic, the important point here is thatit is no longer considered a sign of poorethics or loose morals to default ona mortgage—on the contrary, in manycases it is seen as a savvy move for managing personal finances.

A First Step toward Quantifying

These Drivers

Clearly, a great deal of analysis could be performed to explore the relationshipsbetween the drivers of payment hierarchyreversal described above and the prevalenceof mortgage delinquency in lieu of carddelinquency, or card delinquency beforeauto loan delinquency, by geographic ordemographic sub-segment. We certainlyencourage these explorations, as theycould provide a material benefit to strategies designed to maintain customerloyalty while minimizing delinquency.

As an initial verification of causality, at the state level we explored the

correlation between the percentage of consumers current on their creditcards but delinquent on their mortgagesand unemployment rates, as well as thecorrelation of the former to housing pricedepreciation.1 In the initial study, wefound that the percentage of consumerscurrent on their credit cards but delinquenton their mortgages by state is correlatedto unemployment with a Pearson correlation coefficient ρ = 0.57, whichindicates a strong correlation. With theupdated data, we were able to see thatthe percentage of consumers current ontheir credit cards but delinquent on theirmortgages, by state, is now more stronglycorrelated to unemployment with aPearson correlation coefficient ρ = 0.68.This is likely due to the persistent andlonger duration of unemployment evensince the end of the recession in Q22009. In both instances, the correlationcoefficient is statistically significant at the α = 0.01 testing level.

In addition, that same delinquency statistic was correlated to the change in Housing Price Index (HPI) betweenQ2 2006, the peak of recent U.S. housing prices according to theS&P/Case-Shiller Home Price Index, and subsequent points in time, includingQ4 2007 (the start of the recent recession,according to the National Bureau ofEconomic Research), Q2 2009 (the official end of the recession), and Q42011, the most recent data period. A correlation of state-level delinquencyrates to the Q2 2006 – Q4 2007 state-level HPI changes has a correlationcoefficient ρ = -0.50, showing a moderate correlation that is statisticallysignificant at the α = 0.01 testing level.



Comparisons to the subsequent timeperiods – when even more dramatichome price drops were observed, are even more significant. A correlation ofstate-level delinquencies to the Q2 2006– Q2 2008 state-level HPI changes has a correlation coefficient ρ = -0.76, and a comparison to the Q2 2006 – Q4

2011 HPI changes has a correlationcoefficient ρ = -0.84, with both periodsshowing very strong correlation levels.

This is actually encouraging news. Tobegin with, it implies that changes incustomer preferences and performancecan be explained and, more importantlyfrom a strategy perspective, quantified.

As well, it implies that the reversal ofpayment hierarchy is not permanent.Rather, once the drivers of the hierarchyreversal abate, the payment hierarchy is likely to revert to the traditional prioritization. In other words, onceunemployment eases and home valuesstabilize and begin to rise, we would


% Current on BCbut 30+DPD on Mtg

Unemployment Rate Housing Price Index Housing Price Index Housing Price Index

State 2011 Q4 2011 Q42Q 2006-4Q 2011

% Change2Q 2006-2Q 2009

% Change2Q 2009-4Q 2011

% Change

Nevada 14.36% 12.84% -55.19% -35.10% -30.96%

Florida 13.84% 10.51% -39.30% -27.01% -16.85%

Arizona 10.96% 8.98% -42.87% -25.36% -23.46%

California 8.82% 11.58% -37.89% -31.95% -8.73%

Georgia 7.82% 9.99% -10.64% 0.78% -11.33%

New Jersey 7.81% 8.88% -13.09% -9.61% -3.84%

Mississippi 7.30% 10.04% -20.51% -13.93% -7.64%

Delaware 7.22% 7.33% -15.06% -4.14% -11.39%

South Carolina 7.17% 9.95% -2.65% 6.59% -8.67%

Maryland 7.03% 6.86% -23.19% -11.69% -13.03%

Illinois 6.87% 9.75% -13.83% -5.47% -8.84%

New York 6.86% 7.98% -8.53% -4.26% -4.46%

Rhode Island 6.49% 10.33% -26.26% -15.88% -12.34%

Michigan 6.48% 11.42% -21.14% -13.93% -8.38%

Idaho 6.38% 8.98% -18.12% 0.48% -18.51%

Connecticut 6.23% 8.14% -13.69% -7.15% -7.05%

North Carolina 6.12% 10.00% 1.83% 7.44% -5.22%

Washington 6.07% 9.20% -13.40% -1.37% -12.20%

Louisiana 5.99% 7.59% 4.87% 7.93% -2.83%

Oregon 5.94% 9.68% -16.54% -3.46% -13.55%

Texas 5.84% 8.24% 7.64% 10.02% -2.17%

Utah 5.81% 7.58% -5.78% 6.96% -11.91%

Alabama 5.78% 9.39% 0.24% 7.89% -7.09%

Tennessee 5.73% 9.28% 0.48% 6.27% -5.45%

Washington, DC 5.64% 10.92% -5.62% -7.31% 1.81%

FIGURE 15

Housing Price Index and Unemployment Rate


expect the reversal of payment hierarchyto largely disappear. In the interim,lenders can quantify this dynamic and incorporate the information intocustomer treatment strategies.

It is now evident why the reversal of the traditional payment hierarchy wouldbe more pervasive in California andFlorida, and why subprime borrowers

would show this effect to a larger extent.It is because the former are regions withmore severe home value depreciation andmore extensive unemployment, while thelatter generally have less liquidity, lessaccess to credit card credit and are moreimmediately impacted by unemployment(fewer financial resources to managethrough tough times, etc.).



% Current on BCbut 30+DPD on Mtg

Unemployment Rate Housing Price Index Housing Price Index Housing Price Index

State 2011 Q4 2011 Q42Q 2006-4Q 2011

% Change2Q 2006-2Q 2009

% Change2Q 2009-4Q 2011

% Change

New Mexico 5.63% 7.47% -5.63% 3.49% -8.81%

Maine 5.62% 7.00% -7.00% -1.35% -5.73%

West Virginia 5.54% 9.46% 1.02% 3.88% -2.75%

Ohio 5.49% 9.06% -7.91% -3.33% -4.74%

Indiana 5.42% 8.82% -0.82% 1.44% -2.22%

Hawaii 5.36% 5.82% -13.37% -8.71% -5.10%

Pennsylvania 5.27% 8.27% -1.76% 1.84% -3.54%

Arkansas 5.27% 8.40% 0.65% 3.38% -2.64%

Massachusetts 5.27% 7.41% -10.83% -10.68% -0.17%

New Hampshire 5.14% 5.52% -12.58% -9.55% -3.35%

Oklahoma 5.05% 6.18% 6.60% 8.44% -1.69%

Missouri 5.03% 8.42% -4.07% 0.59% -4.64%

Colorado 4.94% 7.90% -3.08% 0.92% -3.96%

Kentucky 4.89% 9.33% 3.17% 4.22% -1.01%

Virginia 4.83% 6.31% -13.97% -5.67% -8.80%

Minnesota 4.72% 6.89% -15.95% -7.03% -9.59%

Montana 4.18% 7.76% 0.26% 9.25% -8.23%

Kansas 4.17% 6.34% 2.54% 4.18% -1.57%

Wyoming 4.08% 6.09% 7.88% 13.41% -4.88%

Wisconsin 4.02% 7.80% -5.55% -0.55% -5.03%

Vermont 3.97% 5.69% -0.27% 1.50% -1.74%

Iowa 3.77% 5.99% 4.10% 4.14% -0.04%

Nebraska 3.42% 4.56% 1.59% 1.74% -0.15%

South Dakota 3.11% 4.56% 7.14% 8.92% -1.63%

Alaska 3.08% 7.94% 4.25% 3.89% 0.35%

North Dakota 2.21% 3.62% 17.74% 10.44% 6.60%

FIGURE 15

Housing Price Index and Unemployment Rate

The conventional wisdom has always been that consumers will pay their secured obligations first, particularly their mortgages, when faced with a financial crisis. The results of this analysis show that the traditional payment hierarchy has changed. The reasons for that changeare several. As we continue through a gradual (and often fragile) recovery, high mortgage delinquency rates persist, while bankcard and auto loan delinquency rates are well controlled. The reversal of the traditional payment hierarchy that began in Q1 2008 will continueuntil the drivers of that reversal, resulting from the particular dynamics of the recession, abate.

In addition, recent analysis of the consumer behavior that examinesauto loans as well as mortgages and bankcards reveals that consumersare placing a higher payment priority on auto loans ahead of those otherloan types. The key takeaway from these observed shifts in consumerpayment preferences and priorities is the simple reality that consumerbehavior is constantly evolving. Lenders who are using historical paymentmodels to set strategy or, even worse, relying on experience and intuitionrather than data-driven analytics, risk experiencing significant losses ontheir loan portfolios, or missing significant acquisition opportunities tocompetitors, when consumer preferences inevitably shift. By betterunderstanding how bankcard, auto and mortgage consumers within the total market are paying their obligations, lenders can better assessstrengths, weaknesses and specific areas of risk and opportunity whenreviewing current strategies, as well as when considering revised acquisition and account management strategies going forward.

1 The source of unemployment and housing price index data was Eforecasting.com.2 U.S. Bureau of Labor Statistics: http://www.bls.gov/news.release/empsit.t12.htm.3 The rising prevalence of adjustable-rate mortgages (ARMs) after the 2001 recession is a critical

example of the changing nature of deal and loan structures for mortgages. This dynamic is analyzed more thoroughly in a separate study.

4 http://www.aa.com/i18n/travelInformation/duringFlight/dining/cashless.jsp as of 4/25/2010.

© 2012 TransUnion LLC

All Rights Reserved

555 West Adams Street

Chicago, Illinois 60661

USA

transunion.com/business

Documents

Payment Hierarchy Analysis - A Study of Changes in Consumer … · 2015-12-04 · Payment Hierarchy Analysis: A Study of Changes in Consumer Payment Prioritization from 2007 through