Uncertainty and Consumer Credit Decisions

Transcription

1 USC FBE FINANCE SEMINAR presented by Rodney Ramcharan WEDNESDAY, Oct. 19, :15 pm 1:30 pm, Room: ACC-205 Uncertainty and Consumer Credit Decisions BY MARCO DI MAGGIO, AMIR KERMANI, RODNEY RAMCHARAN AND EDISON YU 1 Abstract This paper shows that the effects of uncertainty on consumer credit decisions can be large, especially in the period around a financial crisis. These effects also vary considerably by borrower credit-risk. Among high credit-risk borrowers, increased uncertainty is associated with an increase in credit card balances, but a decrease in the size of their credit card lines. In contrast, low credit-risk borrowers reduce their balances in response to increased uncertainty, but benefit from an increase in their borrowing capacity. A similar pattern emerges in the mortgage market. This evidence suggests that economic and policy-related uncertainty could independently affect economic activity, in part by shaping financial constraints across the business cycle for some borrowers. 1 Di Maggio: Harvard Business School and NBER (mdimaggio@hbs.edu); Kermani: University of California, Berkeley, Haas School of Business and NBER (kermani@berkeley.edu);ramcharan: University of Southern of California, Price School of Public Policy (rodney.ramcharan@gmail.com); Yu: Federal Reserve Bank of Philadelphia (Edison.Yu@phil.frb.org). The views in this paper are those of the authors and do not necessarily reflect those of the Federal Reserve Bank of Philadelphia or the Federal Reserve System. We thank Steve Davis, Matt Kahn and Jose Fillat, as well as seminar participants at the Bank of Canada, Bank of International Settlements, BYU (Marriott School of Business), CEPR Household Finance Conference, Federal Reserve Bank of Atlanta, Stanford and USC for helpful comments.

2 I. Introduction Uncertainty abounds, affecting household and firm decision making through a number of different mechanisms. Well known arguments observe that economic uncertainty can increase the real option value of delaying difficult-to-reverse investment and hiring decisions (Bernanke (1983), Bloom (2009)). Uncertainty can also increase the precautionary saving motive among consumers, with these delays affecting the economy. In addition, uncertainty can operate through credit markets. Higher microeconomic uncertainty can affect collateral values and increase credit spreads in the presence of financial frictions, limiting the supply of credit to entrepreneurs and consumers (Christiano, Motto and Rostagno (2014)). Consistent with these theoretical arguments, the mostly aggregate evidence suggests a powerful role for uncertainty in driving economic fluctuations. 2 Anecdotal discussions echo a similar refrain. 3,4 However, while suggestive, the mostly aggregate evidence on uncertainty is difficult to interpret causally. And far less is known about the effects of uncertainty on households behavior, and whether heighted uncertainty usually associated with financial crises and their aftermaths might directly affect consumer credit decisions. 5 Yet these credit decisions are of enormous economic importance: the stock of mortgage and unsecured consumer credit in the US economy was around 13 trillion dollars as of 2 The aggregate VAR evidence in Bloom (2009) and Caldera et. al (2016) show for example that volatility shocks might be associated with significant declines in output and employment. Bloom, Baker and Davis (2015) provide further evidence, showing that firms most exposed to the public sector might be most sensitive to political uncertainty, while Kelly, Pastor and Veronesi (2015) show that political uncertainty also affects asset prices. 3 The British vote to leave the European Union created substantial uncertainty about future regulatory and economic policy in Britain and the European Union, potentially weighing on economic growth in those countries. The Fed tapering has also ignited a debate about the optimal time to reverse the unconventional measures adopted in the aftermath of the crisis due to the uncertain effects on the recovery path. 4 Criticisms of the New Deal activism during the Great Depression also mainly centered around the harmful effects of policy uncertainty on business investment (Shales (2008)). The head of DuPont chemicals observed in 1938: there is uncertainty about the future burden of taxation, the cost of labor, the spending policies of the Government, the legal restrictions applicable to industry all matters affecting computations of profit and loss. It is this uncertainty rather than any deep-seated antagonism to governmental policies that explains the momentary paralysis of industry. It is that which causes some people to question whether the recuperative powers of industry will work as effectively to bring recovery from the current depression as they have heretofore. excerpted from Akerlof and Shiller (2009), pg There is already substantial evidence that consumer credit outcomes, reflecting both supply and demand forces, shaped economic activity during and after the financial crisis ((Mian, Rao, & Sufi, 2013), (Ramcharan, Verani, & van den Heuvel, 2016), Benmelech, Meisenzahl and Ramcharan (forthcoming)). 2

3 2013. There are however at least two principal challenges to estimate causally the effects of uncertainty on households behavior. First, uncertainty is usually measured in the aggregate. Indexes such as the VIX, which are useful to characterize the economy-wide response during turbulent times, do not provide enough micro-variation to identify a households response to uncertainty. Second, several arguments have observed that uncertainty might endogenously co-moves with first moment shocks (Benhabib, Lu and Wang (2016)). For instance, policy-related uncertainty usually increases after a period of weak economic activity, as governments experiment with new policies. Thus, both measuring uncertainty with sufficient microeconomic variation and disentangling its effects from the direct response to first moment negative shocks is difficult. This paper develops a number of empirical tests to help overcome these challenges. We find evidence that individuals credit decisions are indeed significantly impacted by uncertainty. We use two proprietary datasets that span the period before the crisis ( ), the financial crisis, and up through 2015 periods of remarkable quiescence and unprecedented monetary and regulatory uncertainty. These datasets contains information on major credit card decisions and a rich set of observables such as credit scores, age and zip code of residence. For a subset of individuals, one of these datasets also link information on liabilities to detailed information on mortgage contracts. We also have separate data that comprehensively cover the mortgage market over a similar time period. We then exploit the spatial granularity available in the consumer credit data, constructing new measures of microeconomic uncertainty uncertainty specific to counties and even finer geographic units. These measures are derived from the excess returns of public firms and are constructed to be free of aggregate first moment shocks. They are then aggregated up to the 4-digit NAIC sector level and mapped into counties using quarterly sectoral employment data. Intuitively, this micro-uncertainty series captures in part the spatial and temporal variation in 3

4 uncertainty due to local labor market risk emanating from idiosyncratic sectoral demand and technological shocks. We uncover evidence that uncertainty can drive consumer credit outcomes. We also document significant heterogeneity in the response to uncertainty across individuals. In the case of the mortgage market, we find significant evidence that increased uncertainty is associated with a precautionary contraction in the demand for credit among higher income borrowers. In contrast, the amount of credit demanded from lower income borrowers appear less sensitive to uncertainty. This heterogeneity likely reflects the fact that higher income borrowers generally face higher default costs and are less likely to engage in risk-shifting behavior when uncertainty increases ({SatyajitChatterjee:2007uh}. In response to differing risk-shifting incentives across borrowers, we also find that increased uncertainty is associated with a significant reduction in the supply of mortgage credit, primarily to lower income borrowers. The cost of mortgage credit at the intensive margin also increases sharply in counties where the median credit score is at or below the 660 subprime threshold; the coefficient is insignificant in counties with less risky borrowers. These effects are economically large. The implied decline in mortgage credit demand is around $28 billion a year. Also, the increase in the average 30 year fixed rate mortgage interest rate in response to a one standard deviation increase in uncertainty within a county suggests an additional $15,000 in mortgage interest payments over the life of a typical $250,000 mortgage. The unsecured consumer credit market operates differently from the mortgage market, but the basic results are strikingly similar. Among less credit-worthy borrowers, increased micro-uncertainty is associated with a significant increase in credit card balances, and a decline in the size of credit lines: Their credit utilization increases. In contrast, as with the mortgage market, more credit-worthy borrowers decrease credit card balances when uncertainty increases, but at the same time, their access to credit actually improves, when measured in terms of the size of credit card lines and the number of cards. While this pattern holds even in 4

5 the sample period, the effects of uncertainty are especially large during the financial crisis and its aftermath ( ). Although these results are similar across very different credit markets, data collection methods and controls, they might still be driven by unobserved heterogeneity or be specific to our micro-uncertainty measure. Therefore, to facilitate better causal inference and gauge the generalizability of these results, we build on Di Maggio et. al (2015). In particular, we exploit the plausibly exogenous timing of exposure to interest rate risk in adjustable rate mortgages (ARMs) to identify the impact of uncertainty on consumer behavior. In these ARMs, the mortgage interest rate is fixed for the first 5 years, but then adjusts to the prevailing LIBOR or Treasury rate after this period. Thus, after the reset date, borrowers monthly payments are determined by the prevailing short-term interest rate. As a result, we can take advantage of the variation in the timing of exposure to interest rate risk across individuals, which is predetermined five years in advance, by comparing the credit card balances of individuals with the same type of contract and similar characteristics, who experience the rate reset at different point in time. Also, this institutional setting suggests that monetary policy uncertainty will likely be the most relevant source of uncertainty. This is indeed what we find: Monetary policy uncertainty has a powerful impact on spending decisions as exposure to interest rate risk nears. A one standard deviation increase in the monetary policy uncertainty index developed by Baker, Bloom and Davis (2016) is associated with a 1.1 percent drop in credit card balances two months prior to the date of the interest rate reset; a 2.3 percent decline one month prior to the reset; and a 1.3 percent drop one month after reset. And as before, these reductions in balances are also more pronounced among the low credit-risk borrowers; they are also robust to most plausible controls. In this setting, labor market risk might also be relevant for credit decisions, and we show that micro-uncertainty can also affect credit card balances around the reset period. Interestingly, these results do not hold when we perform a similar analysis with a measure of uncertainty about fiscal policy, which highlights that the sources of 5

6 concern for the borrowers are risks associated with monetary policy, and to a lesser extent employment. Taken together, the evidence in this paper suggests that economic uncertainty might significantly affect consumption and consumer credit decisions. These findings also suggest that the increase in economic and policy-related uncertainty commonly observed during financial crises and their aftermath could independently impede the supply of credit, reducing consumption and economic activity more generally. The heterogeneity across credit-risk types also suggests uncertainty could drive financial constraints across the business cycle for some kinds of borrowers. In section 2 of the paper we discuss some of the underlying theories and data; Section 3 presents the main results and Section 4 concludes. II. Theories and Data II.A Theories There are several channels through which uncertainty might affect consumer credit decisions. Mortgages, and to a lesser extent other forms of consumer credit such as credit cards, are long-term obligations that are difficult to abrogate. And arguments based on irreversible investment and non-convex adjustment costs suggest that the real-option value of waiting to enter into these types of contracts might be higher during periods of increased economic uncertainty (Bernanke (1983) and Titman (1985). These arguments also predict that increased uncertainty can cause delays in irreversible investment and hiring decisions at firms (Bloom (2009)). This in turn can lead to increased employment risk or labor market uncertainty, potentially reducing the demand for some kinds of credit among consumers. The labor market is also a key channel in models that use idiosyncratic uncertainty or risk to explain aggregate fluctuations. In the presence of financial frictions, an increase in idiosyncratic uncertainty the variance of productivity 6

7 shocks to firm capital increases credit spreads for firms (Christiano, Motto, & Rostagno, 2014). When idiosyncratic uncertainty or risk is high, the basic intuition in these models is simple. Credit spreads increase, and credit extended to firms decline. With fewer financial resources, firms acquire less raw capital. Because investment is a key input into the production of capital, investment falls. With this decline in the purchase of goods, output, consumption, and employment fall. Declining employment can then impact the demand for some kinds of credit among consumers. Similarly, models of frictional unemployment also note that an increase in the variance of idiosyncratic shocks--demand or technological--can increase job destruction, reallocation and possibly the unemployment rate (Mortensen & Pissarides, 1994). This prediction rests on the idea that while greater idiosyncratic volatility might imply an improvement in the productivity of some existing jobs, other existing jobs might become less profitable, possibly increasing the rate of job destruction and reallocation, which again could impact consumer credit decisions. Beyond the labor market, uncertainty can also affect credit decisions through asset prices and an individual s financial net-worth (Kelly, Pastor and Veronesi (2015),(Pástor & Veronesi, 2012). For example, during periods of high stock market volatility, households, especially those with a higher fraction of their wealth denominated in stocks, might face greater uncertainty about the value of their financial wealth; the present discounted value of their financial assets might also decline in response to increased uncertainty and higher discount rates. Rather than committing to a contract requiring a series of payments extending far into the future, these households might then find it optimal to reduce or altogether postpone these commitments until uncertainty abates (Hahm and Steigerwald (1999). These arguments all suggest that economic uncertainty can have a sizeable impact on credit decisions, but its impact might also vary across individuals (Satyajit Chatterjee, 2007). There is for example substantial heterogeneity in the 7

8 option value of default across individuals, as borrowers with low credit scores generally have substantially more expensive and limited access to credit (Morse, 2011). For these borrowers, since the default option is cheaper, greater uncertainty can increase their incentives to engage in risk shifting, increasing their demand for mortgage and other consumer debt when risk increases. In contrast, because of their ready access to cheap and plentiful sources of external finance, default is significantly more expensive for borrowers with good credit scores, and risk shifting incentives are less likely to feature in their credit decisions when uncertainty increases. In addition, the impact of economic uncertainty on equilibrium credit outcomes jointly reflects lender decisions, and their decisions might reinforce the heterogeneity in responses across individuals. In anticipation of risk shifting incentives or greater employment risk, lenders might be unwilling to enter into longer term debt contracts with some types of borrowers during periods of increased uncertainty. Instead, lenders may increase credit access to those perceived to be more able to repay when risk increases (Ramcharan et al., 2016). These various theories posit a rich and powerful role for economic uncertainty in shaping consumer credit decisions. But credibly identifying this relationship is difficult, principally because second moment shocks to economic processes often coincide with first moment shocks. This co-movement makes it difficult to disentangle the effects of uncertainty on behavior from a first moment shock that might also independently shape debt decisions. For example, uncertainty might rise during recessions because a decline in economic activity might lead to a decline in information production; at the same time, recessions independently affect spending and credit decisions (Van Nieuwerburgh and Veldkamp (2006), Fajgelbaum, Schaal, Taschereau-Dumouchel (2013)). A number of other mechanisms can also generate endogenous countercyclical fluctuations in uncertainty over the business cycle (see Benhabib, Lu and Wang (2016); Ludvigson, Ma and Ng (2016); and the discussion in Kozeniauskas, Orlik and Veldkamp (2016)). 8

9 This identification challenge also raises measurement concerns. Most existing studies rely on aggregate time-varying uncertainty measures. These include the VIX the implied volatility of the S&P 500 stock market index--along with the policy-related economic uncertainty index developed by Baker, Bloom and Davis (2016) (BBD Index). 6 However, these measures of aggregate uncertainty tend to move in tandem with credit and other aggregate first moment shocks, reinforcing the inference problem. In addition, the theoretical literature emphasizes employment risk and exposure to financial assets as key channels through which economic uncertainty might affect individual credit decisions. But individuals exposure to these channels vary considerably across space, and aggregate uncertainty measures might be less informative when trying to understand individual-level credit decisions. We first address this measurement challenge. Because local or microuncertainty can be key to individual credit decisions, we develop a time varying county-level measure of economic uncertainty that is constructed to be free of aggregate credit market and other first moment shocks henceforth referred to as micro-uncertainty. This micro-uncertainty measure reflects instead the idiosyncratic volatility or risk that likely affects local labor markets and individual portfolios. Direct evidence on the latter is difficult, but we provide correlations suggestive of a robust link between this equity market based micro-uncertainty measure and county and sector level employment outcomes. The empirical strategy then studies the relationship between micro-uncertainty and credit decisions in both the mortgage market and the unsecured consumer credit market. These markets operate very differently and are subject to very different laws and regulations. They also collectively represent about 90 percent of the overall US consumer credit market. In both markets, we also have access to 6 The BBD Index is built on components that quantify newspaper coverage of policy-related economic uncertainty; reflects the number of federal tax code provisions set to expire in future years; and uses disagreement among economic forecasters as a proxy for uncertainty. A related index from policyuncertainty.com is based on categories of economic policy uncertainty culled solely from newspapers. The subcategories include: monetary policy; taxes; health care; national security; entitlement programs; regulation; financial regulation; trade policy; and sovereign debt crises. More details can be found at policyuncertainty.com. 9

10 comprehensive datasets that span the financial crisis as well as the periods before and after. Therefore, this empirical setting allows us to study the impact of microuncertainty across very different credit markets; with very different and detailed controls, many of which are available at the individual-level; and across very different time periods. Because of this level of detail, we can control for myriad aggregate and local economic conditions first moment shocks and establish baseline associations between uncertainty and credit decisions that are robust across very different data generating processes. That said, this approach may be insufficient for causal inference. We therefore turn to proprietary data from Black Box Logic merged with Equifax (BBL) for direct causal evidence of the impact of uncertainty on consumer behavior. The BBL dataset consists of borrowers with adjustable rate mortgages (ARMs) originated between 2005 and These contracts have a fixed interest rate for the first 5 years. After this initial 5 year period, borrowers become directly exposed to interest rate uncertainty: The ARM resets to the prevailing short term interest rate index on the first month of the 6th year, and then continues to adjust either every 6 months or every 12 months thereafter. We use this data generating process to study the response of the individual's monthly credit card balances to micro-uncertainty in the period around the interest rate reset (Di Maggio et. al (2016)). Because the variation in the timing of exposure to interest rate uncertainty across individuals is predetermined some five years prior, these responses are likely to reflect the causal impact of uncertainty on credit decisions. This identification strategy the focus on the change in interest rate exposure also suggests very specific sources of uncertainty and it allows us to gauge the generalizability of these findings to other measures of uncertainty. In particular, monetary policy uncertainty is likely to be most relevant for consumer decision making when interest rate exposure is imminent; conversely, health care uncertainty should be largely irrelevant for spending decisions in this context. We next describe the various datasets before turning to these specific tests. 10

11 II.B Data Measuring Uncertainty Because labor market risk and exposure to financial assets the key channels through which economic uncertainty might affect credit decisions varies substantially across space, this subsection develops a time varying county-level measure of economic uncertainty that is likely free of aggregate credit market and other first moment shocks henceforth referred to as micro-uncertainty. The measure captures the variance in idiosyncratic demand or technological shocks within local labor markets. For each public firm, we first remove the systematic component in daily excess returns by regressing excess stock returns on an augmented three factor model: returns of the S&P 500 index, the book to market ratio, and relative market capitalization (Fama and French (1992)); because we are especially concerned about mismeasurement due to first moment aggregate credit shocks, which might influence individual credit outcomes, we also include the TED spread and the spread between BBB and AAA corporate bonds. The TED spread the difference between the interbank rate and the 3-month Treasury Bill is a common measure of aggregate banking sector distress, while the corporate bond spreads proxy for distress in bond markets. The residuals from these regressions are unlikely to include aggregate first moment shocks, such as time-varying shocks to financing constraints, but instead contain firm-level idiosyncratic demand or technological shocks. The second step computes the daily industry portfolio residual returns by weighting the daily residual returns of firms by their relative size among firms in the same 4 digit sectoral industrial classification code (NAIC) code the firm s relative market capitalization. The third step calculates the quarterly sectorspecific standard deviation of these daily idiosyncratic returns ((Gilchrist, Sim, & Zakrajšek, 2014)). This produces a sector specific index of volatility. The final step draws upon the quarterly sectoral employment data from the Quarterly 11

12 Census of Employment and Wages (QCEW), which lists employment in each county by the 4 digit NAIC. In this final step, we use the QCEW data to create an employment weighted index of economic volatility by county: the 4 digit NAIC sector specific index of volatility is weighted by the county s employment share in that sector with a one-year lag. The use of a one-year lag in the employment share mitigates the potential contemporaneous endogenous response of employment to uncertainty. Along with this second moment index, we also construct the first moment analog: The weighted mean idiosyncratic stock returns at the county level henceforth referred to as micro returns. For each sector, we compute the sectoral daily weighted residual returns by weighting each firm s residual returns by its relative market capitalization within the sector at a daily frequency. We then take the average of the sectoral returns over a quarter to obtain the quarterly mean residual returns for the sector. As before, we map these sector level weighted idiosyncratic returns into the local economy by weighting the sectoral returns by the lagged employment shares at the county level. Figure 1 illustrates the variation in both the aggregate VIX and the microuncertainty index. It plots the time variation in the micro-uncertainty index at different points in its distribution the 10 th, 50 th and 90 th percentiles in each quarter along with the VIX. While the crisis is associated with a significant increase in uncertainty and a concomitant spike in the VIX, countyquarter observations at the 10 th percentile of the local index experienced a far smaller increase in the index. The 90 th -10 th percentile spread in the micro index also increased by a factor of three, suggesting that because of differences in employment patterns and other factors, some counties were far more exposed to the crisis and fluctuations in economic uncertainty than others. The simple correlations in the Table 1 also reveal more of this distributional heterogeneity across space. Movements in the VIX are correlated positively with all three series, especially during the crisis period. But restricting the sample to the post 2009 period, movements in the micro-uncertainty index at the 10 th 12

13 percentile are actually negatively correlated with the VIX and the BBD index. That is, for some counties, the micro-uncertainty index does not mirror mechanically aggregate uncertainty, but likely contains information about economic uncertainty relevant for the local area. That said, the micro-uncertainty series is subject to considerable measurement error. Sectoral idiosyncratic volatility is derived solely from public firms, but mapped into the county-quarter dimension using QCEW employment data derived from both public and private firms. If private and public firms differ in the idiosyncratic shocks that they face, the micro-uncertainty index may poorly measure sectoral and county-level economic uncertainty. Similarly, if the microuncertainty series is driven by firm-specific rather than sector specific shocks, the series may also mis-measure sectoral uncertainty. More fundamentally, because financial markets can be excessively volatile, the micro-uncertainty measure might contain little relevant information for individual credit outcomes. Therefore, before examining the impact of micro-uncertainty on consumer credit decisions, we first show that the empirical relationship between the microuncertainty measure and employment outcomes is broadly consistent with the predictions from the theoretical literature. 7 In column 1 of Table 2A, the dependent variable is the log number of employees in each sector in each quarter, beginning 2000 Q1 through 2015 Q4, for both public and private firms the data are from the QCEW. There are 313 sectors at the NAIC four digit level of disaggregation. The regressor of interest is the sector specific uncertainty series: The standard deviation of the weighted daily residuals for public firms operating in the same 4-digit NAIC sector; the weighting factor is a firm's relative market capitalization within the sector. The other controls include the weighted mean returns within the quarter, sector fixed effects, along with year and quarter fixed effects. Firm employment decisions might respond with some lag to uncertainty, and in column 1, both the sectoral volatility and weighted mean returns enter with 7 See more detailed evidence in Davis, Faberman, Haltiwanger, Jarmin, and Miranda, (2010) linking business variability to direct measures of job creation, destruction and unemployment. 13

14 lags up to four quarters. Although measurement error can arise because the sector uncertainty series uses only public firms and is derived from possibly excessively volatile equity market returns, the sector uncertainty point estimates are consistently negative and statistically significant at the third and fourth quarter lags. These coefficients suggest that a one standard deviation increase in sectoral volatility is associated with a 1.4 percent decline in the level of employment three quarters later, and up to a 2.1 percent drop one year later. Column 2 examines this relationship at annual frequency. A one standard deviation increase in sectoral uncertainty is associated with a 3 percent decline in sectoral employment one year later. All this suggests that notwithstanding measurement error at the sectoral level of the computation exercise, an equity market derived measure of uncertainty might be related to broader labor market outcomes. We next examine the relationship between the micro-uncertainty series and employment outcomes at the county level. The dependent variable in column 1 of Table 2B is the quarterly growth in total QCEW employment in the county, and the regressor of interest is the county-level micro-uncertainty variable, along with the first moment analog based on weighted micro-returns. Year and quarter fixedeffects along with county fixed effects are also included, and standard errors are clustered at the state-level. At the county-level, increased uncertainty is associated with an immediate and sizeable decline in employment growth, as firms likely suspend hiring decisions. This is followed by a rebound in employment growth, beginning three quarters after the initial increase in micro-uncertainty. The cumulative effect is however negative. Over the four quarters, a one standard deviation increase in the index is associated with a 0.4 percentage point decline in employment growth; the mean employment growth rate in the sample is 0.6 percent. Increased uncertainty within a county might also be associated with increased labor market flux: greater labor reallocation and dispersion in employment across sectors within a county. To measure this relationship, we create the weighted 14

15 standard deviation in employment growth across sectors within a county-quarter observation. Let denote the growth rate in employment within sector i in county j between period t and t-1. And let equal sector i s employment share in county j in period t. The variable = is the weighted average growth rate in employment within the county, computed over all sectors i; the dispersion measure in employment growth across sectors within a county is =.. The evidence in column 2 suggests that increased uncertainty is associated with greater dispersion in employment growth rates across sectors inside a county. This positive effect is most noticeable in the second and third quarters after an increase in micro-uncertainty. And over the four quarters, a one standard deviation increase in micro-uncertainty is associated with a 1.25 percent increase in the dispersion in employment growth within a county. These basic correlations suggest that the micro-uncertainty measure might be related to labor market fluctuations a key source of risk that can influence the credit decisions of individuals and intermediaries. We next describe the data on credit decisions. Credit Decisions The analyses focus on mortgage and consumer credit decisions. According to the Federal Reserve s Flow of Funds data, these two sources of credit account for approximately 13 trillion dollars or about 90 percent of total consumer liabilities in Our various data sources are representative of these two very different credit markets, and together comprehensively cover the US consumer credit market. Mortgage Credit: Loan Processing Service (LPS) and Home Mortgage Disclosure Act (HMDA) 8 The Flow of Funds data can be found here: 15

16 Data from HMDA record the universe of mortgage credit applications and outcomes for non-rural Metropolitan Statistical Areas in the United States. Data on applications as well as loan origination outcomes can help gauge the impact of uncertainty both on the demand for mortgage credit as well as the supply response of lenders. These data include key borrower characteristics like income, race, census tract of the property and loan amount; the loan application is linked to the bank in many cases. We collected these data annually from , yielding some 72 million mortgage credit applications. Unfortunately, while HMDA provides information on quantities, it does not consistently record interest rates. We thus turn to county-level quarterly data from LPS a proprietary source of mortgage data derived from seven of the largest mortgage loan processers. We use these data to construct the average interest rate, weighted by loan shares, for newly originated mortgages. The panel in Figure 2 presents denial rates and median applicant income over time (HMDA), and mortgage interest rate spreads (LPS) over Consumer Credit: NY Federal Reserve s Equifax Consumer Credit Panel and Black Box Logic We draw a two percent sample from the New York Federal Reserve s Equifax Consumer Credit Panel (Equifax). This is a proprietary consumer credit dataset, and the sample results in a balanced panel of about 220,000 individuals. It includes comprehensive quarterly information on key dimensions of debt usage: credit card balances, as well as credit limits from The panel also includes relevant individual-level information on age; census tract of the primary residence; and the Equifax Risk Score --an important credit scoring index commonly used in credit decisions; higher values suggest less credit risk. In what follows, we primarily use data on credit card balances and lines to measure consumer credit. We supplement Equifax with proprietary data from Black Box 16

17 Logic (BBL) panel. The BBL data links consumer credit usage with mortgage contract terms at the monthly frequency. The structure of the dataset allows us to make further progress in causally identifying the impact of uncertainty on consumer credit outcomes. Table 3 reports basic summary statistics for some of the individual variables, observed in 2008 Q1 from the Equifax and BBL. The Equifax panel is more representative of the general credit-using population, and contains information on non-homeowners and homeowners alike. The average credit card limit in Equifax is around $13,500 while the average credit card balance is a little less than half that number. The average utilization rate, the ratio of balances to limits, is around 70 percent. The average age, around 48, is higher than the US average; and the typical risk score is just under 700 well above the traditional subprime cutoff of 660 for mortgage credit. Unlike Equifax, Black Box Logic contains a richer set of data but for homeowners with prime credit. Vantage scores similar to but distinct from Equifax Risk Scores are significantly higher, with the average around 740. The mean credit card limit and balance are also much higher than the more general population surveyed in Equifax, but utilization rates are much lower. Mortgage balances are also much higher among the BBL ARM sample. Unlike Equifax, BBL also contains mortgage contract loan terms. These loans were contracted during and the mean interest rate is around 5.8 percent, with LTV ratios averaging 77 percent. The panel in Figure 3 plots the median outcomes for these variables over the crisis and post crisis sample period (2008 Q1-2013Q4) among the set of individuals with positive balances for both the more general Equifax dataset and the BBL data. There are differences across the two samples, likely reflecting the different economic circumstances of the median individual across the two datasets ((Di Maggio et. al (2016)). In both datasets for example, utilization rates decline sharply with the crisis, but this rate recovers after the recession in the Equifax 17

18 data, but it continues to decline in the BBL dataset, potentially due to the mortgage debt overhang after the housing crisis. III. Main Results IIIA.Basic Associations: Micro-uncertainty and Mortgage Credit This subsection presents the basic associations between the micro-uncertainty index and mortgage credit. Table 4 uses the HMDA applications data to study the relationship between the index and indicators of mortgage credit demand. To proxy for demand, the dependent variable in column 1 the log volume of mortgage credit demanded in these applications as the dependent variable; column 2 uses the log number of applications inside a county within a calendar year. The sample period extends from Controls include standard demographic and income variables from the American Community Survey, including the log of population, all observed between , along with year and state fixed effects; standard errors are clustered at the state-level and all county-level regressions are weighted by population. For the full sample period, there is no evidence of a robust statistical relationship between micro-uncertainty and these proxies for mortgage credit demand. We divide the sample into two equal 5 year panels, with columns 3-4 focusing on the crisis and its aftermath: , while columns 4-5 consist primarily of the pre-crisis period, During the period, a one standard deviation increase in the micro-uncertainty index is associated with a 5.4 percent drop in the amount of mortgage credit demanded in loan applications. The impact on the number of applications is nearly identical, but this relationship is not statistically significant (p-value=0.14). In both cases, during the relatively tranquil period, the micro-uncertainty variable enters with the opposite sign 18

19 and is imprecisely estimated. 9 The estimates in column 3 appear economically important. Using the mico-uncertainty index coefficient in column 3, we use the variation in the micro-uncertainty index to compute the predicted drop in the volume of mortgage credit demanded. Over the sample period, this point estimate suggests a $141 billion decline or about a $28.4 billion per annum drop in the volume of mortgage credit sought by potential borrowers. Mortgage loan applications are an imperfect proxy for loan demand: These results could reflect an anticipation among borrowers of a decrease in credit supply instead of a precautionary reduction in demand among borrowers in response to increased uncertainty. However, the variation in borrower income available in HMDA can help in understanding better this negative relationship. Because higher income borrowers are less likely to face a decline in credit supply, any negative relationship between the micro-uncertainty series and loan applications for this subsample is more likely to reflect precautionary behavior in response to uncertainty. Contrasting forces might drive outcomes among lower income borrowers. These borrowers are more likely to face credit constraints when uncertainty increases, leading to a pre-emptive drop in applications. But given their lower default cost, lower income borrowers may also be more inclined to engage in risk shifting, increasing their demand for mortgage debt when risk increases, or evincing less sensitivity to increased risk. Using income heterogeneity, the dependent variable in column 5 is the log volume of credit demanded by borrowers with above median income in the county. In column 6, the dependent variable is the log of credit demanded in the county, but computed only for those borrowers with below median incomes. Among the above median-income sample of borrowers (column 5) those borrowers less likely to face binding financing constraints the impact of microuncertainty is large, negative and highly statistically significant in the These differences across the two sample periods are statistically significant: estimating the full sample and allowing the coefficient on micro-uncertainty and weighted local returns to differ during the time period yields an interaction term with a coefficient of (p-value=0.03). 19

20 subsample. A one standard deviation increase in the micro-uncertainty index is associated with a 6 percent drop in loan volumes within the year among this sample of borrowers. But among the below median-income borrowers, the index coefficient is about 25 percent smaller and not statistically significant. A similar pattern emerges when using the log number of applications across the two income subsamples (columns 7-8). This evidence suggests that an increase in risk appears to be associated with a relative decline in the demand for mortgage credit among higher income borrowers, skewing the pool of applicants towards lower income individuals. Table 5 uses the individual application level data to study the supply response to micro-uncertainty. We first focus on the extensive margin. The dependent variable in column 1 is the probability that a loan application is denied. We control for borrower composition using the log of borrower income; the log of the requested loan amount; race and gender. We also use county-fixed effects and cluster standard errors at the state-level. Consistent with the previous demand evidence reported in column 1 of Table 4, column 1 of Table 5 shows no significant relationship between micro-uncertainty and a lender s decision to deny a loan when using the full sample of borrowers over the entire sample period: As before, the impact of micro-uncertainty on mortgage credit decisions become economically and statistically significant in the period (column 2). The point estimate suggests that a one standard deviation increase in micro-uncertainty is associated with a 0.2 percentage point increase in the probability that a loan is denied; the mean unconditional probability of denial is 11 percent in the sample period. The positive impact of uncertainty on denials appear concentrated among borrowers with below median incomes (column 4). In this subsample, the impact of uncertainty on denial rates are nearly double that estimated in the full sample. Moreover, from column 5, denial rates do not significantly increase in response to uncertainty among the above-median-income borrowers subsample. 20

21 An individual s decision to seek a mortgage from a specific lender could be correlated with potentially relevant lender unobservables that might bias these results. For example, poorer or riskier borrowers might seek out lenders believed to be less sensitive to risk when making loan decisions. This endogenous matching could in turn bias downwards the mico-uncertainty estimates. To partially address this issue, we use bank-fixed effects for the subsample of data for which we know the bank involved in the application process. These results are presented in columns 7 and 8. The point estimates are larger than before, and they continue to suggest that lenders react most forcefully to micro-uncertainty when dealing with mortgage credit applications from below-median-income borrowers. We next focus on the intensive margin in Table 6. Columns 1 and 2 model the intensive margin by focusing on loan sizes for those loans that were originated. There is some indication that originated loans shrink in response to uncertainty, especially among below-median-income borrowers, but these estimates are imprecise. When taken together, these basic associations on quantities suggest that increased micro-uncertainty might alter the composition of borrowers, as high quality borrowers reduce their demand for mortgage credit. Increased microuncertainty is also associated with more selective lending, as rejection rates rise disproportionately for lower income borrowers. The remaining columns of Table 6 use data from LPS to investigate the relationship between micro-uncertainty and the average price of newly originated mortgage credit using the county-quarter unit of observation. The HMDA quantity based evidence suggests that the impact of micro-uncertainty on average interest rates within a county-quarter might be ambiguous. The reduced demand for credit among higher quality borrowers could lead to higher average interest rates on loans originated in equilibrium in a county-quarter. But because lenders appear to raise lending standards in response to uncertainty, the equilibrium average interest rates inside a county might be decline. The dependent variable in column 3 is the average interest rate inside a county in a given quarter; this average is weighted by loan size. The specification includes county-fixed 21

22 effects along with year-by-quarter fixed effects to absorb aggregate shocks and standard errors are clustered by state; we weight the county-level regressions by population. From column 3, over the sample period, there is a significant relationship between uncertainty and the average cost of mortgage credit one quarter later. A one standard deviation increase in micro-uncertainty is associated with a 4 basis point increase in the average interest rate originated in the county in the subsequent quarter. Building on the HMDA-level results that borrower heterogeneity features in the supply and demand responses to micro-uncertainty, we split the sample into counties where the median FICO score is below 660 the subprime cutoff and those above this threshold. This FICO score is computed in 2006 to avoid endogenous changes in credit ratings. In those counties where potential borrowers are more likely to engage in risk-shifting behavior (column 4), the impact of increased micro-uncertainty on the average cost mortgage credit is positive and significant one quarter ahead. A one standard deviation increase in micro-uncertainty is associated with a 17 basis point increase in the average cost of mortgage credit the next quarter inside the county. Among the sample of county-year observations with higher median FICO score borrowers (column 5), the point estimates are statistically and economically insignificant. Both the individual and county-level associations drawn from different data sources and collection methods suggest that increased uncertainty might affect mortgage credit at both the extensive and intensive margins. However, the precise channel underlying this relationship remains unclear. For example, beyond shaping individual risk-shifting incentives and default probabilities through labor market and asset pricing risk, the micro-uncertainty measure could also affect local house prices a key local asset. For example, liquidation values for homes could decline in response to increased uncertainty within a county. This effect could help explain bank and borrower behavior at the different margins. Also, because HMDA offers a limited set of individual-level controls, we cannot be certain whether these results are driven by risk-shifting behavior, or 22

23 unobserved first moment shocks emanating from the housing market that might also be correlated with the micro-uncertainty measure. Therefore, before taking up the question of causality, we study the impact of micro-uncertainty on credit decisions made in the unsecured consumer credit market; this market operates very differently from the mortgage market, helping us to gauge the extent to which these results might generalize. The data on unsecured consumer credit transactions also offer a richer set of controls that can help address biases from unobserved heterogeneity. IIIB.Basic Associations: Micro-uncertainty and Consumer Credit Table 7 examines the impact of micro uncertainty on unsecured consumer debt decisions using individual-level data from Equifax. The data are quarterly and the sample period, 2002Q1-2015Q4, encompasses the financial crisis, as well as the periods before and after. All specifications control for individual-level observables such as age, and the previous year s average Equifax Risk score, along with individual fixed effects and year-by-quarter fixed effects; individual fixed effects absorbs possibly time invariant individual level factors such as risk aversion, while year- by-quarter effects captures aggregate first moment and other shocks. As before, we also control for local weighted returns at the county-level the first moment analog to the 4-digit NAIC based micro-uncertainty index, and standard errors are clustered at the state level. In column 1 of Table 7, the dependent variable is the log of the individual s credit card balance in the quarter. We also control for the individual s debt capacity using the log of the credit line in that quarter as a regressor. The coefficient on the micro-uncertainty variable is negative but not statistically different from zero. The coefficient itself suggests that a one standard deviation increase in uncertainty is association with a 1 percent drop in credit card balances. We have already seen evidence in the mortgage market that the response to uncertainty can vary across borrower types. In the context of the Equifax, high Risk score borrowers are unlikely to be credit constrained, and changes in their 23