Hey, If you are searching Loans with bad credit on Google or other search engines or you want to know more about this topic, then you’ve came to the right place. Here you will get all precious knowledge about this topic like Inherent Credit Risk from Inefficient Lending and what are Best Practice Loan Performance and the Efficiency of Lending, how to get urgent loans for bad credit. By which method bad credit loans guaranteed approval you will get. So Read the full article in order to get updated and precious knowledge..

Loans with bad credit: How Bad Is a Bad Loan? Distinguishing Inherent Credit Risk from Inefficient Lending (Does the Capital Market Price This Difference?)

Introduction – Loans with bad credit

The ratio of nonperforming loans to total loans a bank experiences reflects both the inherent credit risk the bank targets and the bank’s proficiency at evaluating credit risk and monitoring the loans it has made. In our 2013 data on U.S. banks, we find that small community banks and the largest financial institutions on average experience the highest ratios of nonperforming loans. How much of this nonperformance is due to inherent credit risk and how much, to their proficiency at loan making? To answer this question, we develop a novel technique based on stochastic frontier estimation and apply it to data on large as well as small banks to compare their lending performance. The ratio of banks’ nonperforming loans is decomposed into three components: first, a minimum ratio that represents best-practice lending given the volume and composition of a bank’s loans, the average contractual interest rate charged on these loans, and market conditions such as the average GDP growth rate and market concentration; second, luck or statistical noise, which should be removed from the bank’s observed ratio of nonperforming; and, third, the bank’s excess nonperforming loan ratio, which equals the difference between the bank’s observed nonperforming loan ratio adjusted for luck and its best-practice minimum ratio and which represents the bank’s proficiency at loan making. The best-practice ratio of nonperforming loans represents the inherent credit risk of the loan portfolio – the nonperforming loan ratio a bank would experience if it were fully efficient at credit evaluation and loan monitoring. This best- practice minimum is obtained by estimating a stochastic lower envelope of nonperforming loan ratios conditioned on variables associated with inherent credit risk. The stochastic frontier estimation eliminates the influence of luck, statistical noise, and gauges systematic failure to achieve the best-practice minimum.

Stochastic frontier techniques have been applied in numerous studies of cost, production, and profit. Greene (2012) offers a textbook description while Kumbharkar and Lovell (2000) and

Coelli, Rao, and Battese (1998) give an encyclopedic treatment. Hughes and Mester (2010, 2015) and Berger and Mester (1997) discuss how these techniques are applied to analyze the performance of financial institutions. Hughes, Jagtiani, Mester, and Moon (2018) apply our technique to commercial and industrial loans and to commercial real estate loans made by community banks. 1 They find that large community banks are more efficient at lending for these two types of loans than small community banks. Hughes, Jagtiani, and Moon are engaged in research that applies our technique to compare consumer lending by commercial banks and consumer lending by fintech organizations.

We estimate the frontier over 710 banks of all sizes – privately as well as publicly traded – that have a focus on lending. The size of these banks ranges from $92.7 million to $2.4 trillion in consolidated assets. While the largest banks exhibit the largest ratio of nonperforming loans, we find that nonperformance at these institutions results from inherently more risky lending, not a lack of proficiency at credit evaluation and loan monitoring.

The publicly traded banks in our sample allow us to investigate whether capital market pricing distinguishes between inherent credit risk and lending inefficiency. We find that Tobin’s q ratio is negatively related to the nonperforming loan ratio for all but the largest 9 banks. For 8 of the largest 9 banks, the relationship is positive and statistically significant. The dichotomy in the treatment of nonperforming loans by the capital market suggests that credit-risk strategies of the largest banks differ from those of smaller banks. When the nonperforming loan ratio, adjusted for statistical noise, is decomposed into the inherent credit risk (the best-practice minimum ratio) and the lending inefficiency ratio, a statistically significant positive relationship is found between Tobin’s q ratio and inherent credit risk at many large banks. While the relationship is negative at many smaller banks, it is not statistically significant. The evidence that the capital market rewards higher credit risk at the largest banks is consistent with their higher ratio of nonperformance and

1 An earlier version of this paper, Hughes, Jagtiani, and Mester (2016), has been superceded by Hughes, Jagtiani, Mester, and Moon (2018).

suggests that market discipline may not enhance financial stability through the lending channel at these large banks. On the other hand, lending inefficiency undermines value at banks of all sizes.2

The Data (Loans with bad credit)

To estimate the frontier, we start with data on 807 top- tier holding companies at year-end 2013 whose balance-sheet and income statement information is reported on the Y9-C report. The 807 companies are obtained after dropping one company with no nonperforming loans and all companies without data on small business loans (commercial and industrial loans with an origination amount under $ 1 million). The data on small business loans are obtained by summing the loan amounts of subsidiaries of the top tier company from the Call Report data.3 In this section, we provide details on trimming the full sample to achieve a sample of banks whose focus on lending is sufficient to allow the estimation of a stochastic best-practice loan performance frontier.

The average ratio of total loan volume to consolidated assets is 0.636, bounded by a minimum of 0.055 and a maximum of 0.962. Table 1 sorts the data by the ratio of total loan volume to consolidated assets for those companies with ratios less than 0.40. The 4 companies with the smallest loan ratios, less than 0.15, are not focused on the loan-making function of commercial banks and, thus, are trimmed from the data used to estimate best-practice loan making.

In addition, some companies exhibit unusually large ratios of nonperforming loans to total loans. Nonperforming loans include loans past due less than and more than 90 days plus nonaccruing loans, lease financing receivables, placements, and other assets (BHCK525+BHCK5524 +BHCK5526)4; gross charge-offs (BHCK4635); and other real estate owned (BHCK2150). Since some banks are more aggressive in charging off past-due loans and, consequently, would appear to have a lower ratio of past-due loans, gross charge-offs are added to past-due loans to eliminate distortions caused by differences among banks in charge-off strategies. Table 2 summarizes loan

2 Hughes, Mester, and Moon (2016) find a similar relationship between the equity capital ratio and financial performance based on market value measures. At the margin, the largest financial institutions improve financial performance by increasing financial leverage while smaller institutions, by reducing leverage.

3 These data are used in Hughes, Jagtiani, Mester, and Moon (2017) and constructed by Quinn Maingi of the Federal Reserve Bank of Philadelphia.

4 The BHCK numbers refer to categories in the Y9-C regulatory reports filed by bank holding companies.

performance of banks with the highest nonperforming loan ratios. Those whose ratio exceeds 0.15 ⁵ are trimmed to estimate the best- practice performance frontier. Table 3 summarizes loan performance of banks with the lowest nonperforming loan ratios. Those whose nonperforming loans constitute a proportion less than 0.01 of total loans are also trimmed to estimate the best-practice performance frontier. The log transformation of the volume of nonperforming loans is plotted in Figure 1 against the log transformation of the volume of total loans for the trimmed sample of 710 top-tier holding companies. In Figure 2, where the ratio of nonperforming loans to total loans is plotted against the log transformation of total loans, a higher ratio of nonperformance for the largest banks is evident. In addition, the smallest banks also exhibit higher ratios of nonperformance.

Table 4 shows that banks in all but the largest category have similar mean and median values of the ratio of loans to assets. Banks in the largest group, whose consolidated assets exceed $250 billion, allocate on average 51.74 percent of their assets to loans while banks in the other size groups allocate on average 63.22 to 66.78 percent of their assets to loans.

As suggested in Figure 2, the mean ratio of nonperforming loans to total loans is higher in the smallest size group, community banks with assets less than $1 billion, and in the largest size group, banks whose consolidated assets exceed $250 billion. On average 4.61 percent of loans at these small community banks are nonperforming while loans at the largest banks exhibit a mean nonperformance rate of 6.10 percent. In contrast, loans in the next largest group (assets between $50 billion and $250 billion) default at 3.07 percent. Table 5 shows the details of nonperformance for the banks whose consolidated assets exceed $50 billion.

The strikingly higher average ratio of nonperformance among the largest banks and, to some degree, among the small community banks, raises the question of whether these banks are on average less efficient at credit evaluation and loan monitoring or whether they may be lending to riskier borrowers who have a higher expected rate of default.

Best-Practice Loan Performance and the Efficiency of Lending

We use stochastic frontier techniques to distinguish between nonperformance due to less effective credit evaluation and loan monitoring and nonperformance due to the bank’s choice of the overall credit risk of its loan portfolio. The frontier is a stochastic lower envelope of the nonperforming loan ratios conditioned on the volume of loans and the composition of the loan portfolio by types of loans it holds. In addition, we include variables that characterize inherent credit risk, such as the average contractual loan rate, an index of market concentration, and the GDP growth rate in banks’ local lending markets. The stochastic lower envelope of nonperforming loan ratios takes into account the observed loan performance of all banks in the sample and eliminates the influence of luck, statistical noise, on loan performance. It estimates observed Best Practice Loan, the minimum ratio of nonperforming loans a bank could obtain if it were fully efficient at credit evaluation and loan monitoring given its loan volume, the composition of its loan portfolio, its average contractual lending rate and the economic conditions in its local lending markets.

The control variables characterize banks’ exposure to credit risk. Default probabilities differ by the type of loan, and so it is important to include variables that characterize the composition of the loan portfolio. In addition, the contractual interest rate charged on a loan includes a credit risk premium and, itself, influences the quality of loan applicants through adverse selection.5 In addition, conditions in the markets in which a bank lends, such as the macroeconomic growth rate and the bank’s market power, influence loan performance. Petersen and Rajan (1995) find that a bank with market power is able to price a loan to a young business at a lower- than-competitive rate to reduce the probability of default. As the firm gains experience and continues to borrow from the bank, the bank recovers its implicit subsidy by lowering the loan rate but not as much as would occur in a competitive market. Thus, the degree of market concentration can influence loan performance.

The choice of specifying the frontier in terms of the ratio of nonperforming loans to total loans is motivated by the need to obtain an unbiased estimate of the best-practice minimum ratio of

5 Morgan and Ashcraft (2003) find that the interest rates charged by banks on business loans predict future loan performance.

urgent loans for bad credit, bad credit loans guaranteed approval, legit personal loans for bad credit, Best Practice Loans, urgent loans for bad credit, bad credit loans guaranteed approval, legit personal loans for bad credit, Best Practice Loans

nonperforming loans to total loans, which will be used to gauge inherent credit risk. If the log
transformation of the amount of nonperforming loans or the log transformation of the ratio were
used as the dependent variable in the frontier estimation, an unbiased estimate of the degree of
inefficiency can be obtained; however, the estimate of the minimum ratio of nonperforming loans to
total loans computed from the log transformation is not unbiased. Thus, we avoid the log
transformation.
Defining the frontier in terms of the ratio of nonperforming loans provides two important
advantages. First, it gives an unbiased estimate of inherent credit risk; and, second, it decomposes
the nonperforming loan ratio, adjusted for statistical noise, into two ratios capturing inherent credit
risk and lending inefficiency.
We use maximum likelihood to estimate a best-practice l an performance frontier that
determines the minimum nonperforming loan ratio conditional on the total loan volume (expressed
in 100 billions), the average contractual interest rate, macroeconomic conditions and market
concentration in the bank’s markets, and the composition of the loan portfolio. That is,		(1)
NPi = a0 + a1	(Total loansi (100 billions)) + X•β + εi.
where NPi = observed ratio of nonperforming loans to total loans at bank i,
and X is a vector of other control variables:
x1	= Contractual lending ratei ,
x2	= Herfindahl index of market concentration across banki’s markets,
x3	= GDP growth rate across banki’s markets,
x4	= Small business loan volumei / Total loansi,
x5	= Total business loan volumei / Total Loansi,
x6	= Consumer loan volumei / Total Loans i,
x7	= Residential real estate loan volumei / Total Loansi,
x8	= Commerical real estate loan volumei / Total Loansi ,
and εi = νi +	µi is a composite error term.

The Herfindahl index of market concentration is a weighted average of banking market

concentration in each state in which the bank operates. The weights are the proportions of deposits

located in each state. The GDP growth rate is a 10-year weighted average state GDP growth rate in

the states in which the bank operates. The weights are the same as those used to compute the

Herfindahl index. The composite error term, εi

= νi

+ µ , is the sum of a two- sided, normally

distributed error term,

~ iid

(0,

σν2

statistical

noise, and a term,

, which is a

), that captures

positive, half-normally distributed error term,

μi (≥0) ~ iid N(0,σμ2), that gauges systematic excess

nonperformance.

The best-practice (minimum) nonperforming loan ratio is given by the deterministic kernel

of the frontier:

= FNPi = a0 + a1 (Total loansi (100 billions)) + X•β.

(2)

best-practice NPi

The bank’s excess nonperforming loan ratio, µI, cannot be directly measured so, following

Jondrow, Lovell, Materov, and Schmidt (1982), we define bank-specific excess nonperformance or

lending inefficiency by the expectation of µi conditional on εi, the two-sided error term,

excess nonperforming lo n ratioi

E µi

^εsided

(3)

( |

error term,

while luck is measured by statistical noise, the two-

(4)

luck

E ν εi

εi

E µi εi

frontier estimation

Thus, the

=(|)=

−(|).

decomposes the observed nonperforming loan ratio into three

components: the best-practice (minimum) nonperforming loan ratio, the excess nonperforming

loan ratio (over best-practice), and luck:

+ E(µi|εi) + E(νi|εi)

NPi = a0 + a1 (Total loansi (100 billions)) + X•β

(5)

= best-practice NPi

+ Excess NPi

+ lucki.

The ‘expected’ best-practice nonperforming loan ratio, conditional on the values of the control

variables is the ratio that would be achieved were the bank totally efficient at credit evaluation and loan monitoring. As such, it represents the inherent credit risk of the bank’s loan portfolio. The excess nonperforming loan ratio, which measures the effectiveness of the bank’s credit evaluation

Figure 2. Best Practice Loans with bad credit

and loan monitoring, can be expressed as the difference between the observed nonperforming loan
ratio adjusted for noise, (NPi − νi), and the frontier value of the nonperforming loan ratio:	(6)
*Lending Inefficiency* = *Excess NPi = E(µi\|εi) = [ NPi − E(νi \|εi) ] − FNP*i.
Figure 3 illustrates the frontier and these components of the decomposition of the

nonperforming loan ratio. The hypothetical bank whose loan performance is illustrated in this figure has total loans of 0.08 (expressed in 100 billions6) and experiences a nonperforming loan ratio adjusted for statistical noise, NP − ν , of 0.025. The ‘expected’ best-practice minimum ratio equals 0.01. Its noise-adjusted ratio, 0i.025,i exceeds the best-practice minimum, 0.01, by 0.015. Thus, its lending inefficiency is 0.015.

Figure 4 shows these estimated values for a portion of the sample. For illustration, the four banks with the largest volume of loans, Bank of America, Wells Fargo, JP Morgan Chase, and Citibank, are highlighted. For a given volume of loans, the noise-adjusted observed nonperforming loan ratio is indicated in blue while the best-practice (minimum) ratio, in red. The difference between these ratios is the excess nonperforming loan ratio, an indication of the proficiency of the bank in lending.

3.1 The Estimation of the Best-Practice Loan Performance Frontier – Best Practice Loans with bad credit

The sample is restricted to holding companies with a loan volume that exceeds 15 percent of consolidated assets and nonperforming loans, broadly defined, that are at least 1 percent and no more than 15 percent of the total loan volume. In addition, data on small business lending obtained from the Call Reports of subsidiaries of the holding company must be available. These restrictions yield a sample of 710 bank holding companies whose nonperforming loan ratios are plotted in Table 6 shows the estimated parameters of the frontier specified in equation (1). We find that for any given volume of loans, a higher contractual interest rate is associated with a higher best- practice nonperforming loan ratio, while a higher GDP growth rate and a higher proportion of 6 The Y9-C data report amounts in 1000s, and so 0.08 (in 100 billions here) corresponds to 8,000,000 (in thousands) in the Y9-C data report.

small business loans are associated with a lower best-practice ratio of nonperforming loans. The significantly positive coefficients on the proportions of total loans made up of residential real estate loans and consumer loans imply a positive correlation of residential real estate loans and consumer loans with the nonperforming loans ratio.7 The coefficients on the ratios of total business loans and commercial real estate loans are all insignificantly positive. The negative coefficient on the Herfindhal index of market concentration is not significant; however, the sign is consistent with the hypothesis of Petersen and Rajan (1995).

In Table 7, the sample is partitioned into five size groups. Summary statistics for each group are reported for the observed ratio of nonperforming loans, which is divided into the best-practice minimum ratio and the difference between the observed ratio, with statistical noise eliminated, and the best -practice ratio – the measure of lending inefficiency.

3.2 Credit Risk and Lending Inefficiency at Community Banks ( Loans with bad credit )

In Table 7 both the mean and median values of the ratios of observed nonperforming loans to total loans is larger at small community banks (under $1 billion in assets) than at the groups of other banks with assets less than $250 billion. For example, the mean 0.0461 at these small banks exceeds the value 0.0383 at large community banks, 0.312 at banks with assets between $10 billion and $50 billion, and 0.0307 at banks with assets between $ 50 billion and $250 billion. The contrast of loan performance at small community banks with that at larger banks under $250 billion raises the question of whether these banks are lending to riskier borrowers or whether they are less efficient at lending. The mean average contractual lending rate at small community banks is the highest of the groups of banks with assets less than $250 billion, which suggests either that these community banks are lending to riskier borrowers or that they are pricing the loans to cover their relatively high default rate that could result from a lack of proficiency at lending.

7 Since the proportion of assets allocated to lending is one of the control variables, a variation in any category of loans implies an offsetting variation in the omitted categories of loans. The exception is a variation in small business loans since the proportion of total business loans is a control variable. These omitted categories include leases, agricultural loans, loans to nondepository institutions, and other loans.

The inherent credit risk of their loans can be gauged by the value of the nonperforming loan

ratio on the deterministic frontier – the best practice ratio of nonperforming loans given the volume

of loans, their composition, their average contractual interest rate, and the GDP growth rate and concentration in banks’ local markets. The mean value of inherent, best-practice credit risk at small community banks, 0.0123, is higher than the mean value, 0.0109, at large community banks and 0.0096 at banks in the range $10 billion to $50 billion. This pattern across size groups is similar to that of the average contractual lending rate.

The distance of the observed noise-adjusted ratio of nonperformance from the minimum, best-practice ratio on the deterministic frontier constitutes lending inefficiency. The mean inefficiency, 0.0338, of these small community banks exceeds 0.0274 at large community banks, 0.0217 at banks in the range $10 billion to $50 billion, 0.0179 at banks from $50 billion to $250 billion, and 0.0186 at banks whose assets exceed $250 billion. Thus, small community banks appear less efficient at lending.

3.3 Credit Risk and Lending Efficiency at assests the Largest Banks

The seven largest banks with consolidated exceeding $ 250 billion experience the highest mean ratio of nonperformance, 0.0610, among the five groups. Moreover, the mean average contractual lending rate, 0.0538, and the mean inherent credit risk of their lending, 0.0424, are the highest among the five groups. On the other hand, their mean inefficiency at lending, 0.0186, is the second lowest among the five. Banks in the range $50 billion to $250 billion exhibit an inefficiency ratio of 0.0179; in the range $ 10 billion to $ 50 billion, 0.0217; in the range $ 1 billion to $ 10 billion, 0.0274; and less than $1 billion, 0.0338. Thus, the high mean ratio of nonperformance of the largest financial institutions appears to result from lending to riskier borrowers rather than inefficiency at lending.

Table 8 shows the values of lending inefficiency, the best- practice ratio of nonperforming loans measuring inherent credit risk, and the observed ratio of nonperforming loans for large banks with consolidated assets greater than $50 billion. Nonperforming loans as a percentage of total loans range from 7.98 percent for Wells Fargo to 1.46 percent for Comerica. The best-practice percentage spans from 5.89 percent for Bank of America to 0.17 percent for Northern Trust. The ¹² lending inefficiency ranges from 3.07 percent for PNC to 0.19 percent for Discover Financial Services.

3.4 Factors that Influence Lending Inefficiency (Loans with bad credit)

Factors that potentially explain a bank’s lending inefficiency are explored in Table 9 in regressions specified by the general- to-specific modeling strategy, which is a well -grounded method to search for the best model specification. 8 It is statistically consistent in reaching the true model when the true model is included in the set of candidate models. The general specification generously includes many regressors to reflect the existing literature and our presumptions. From the pool of the general specification and all possible restrictive specifications, we then select the one with the smallest AIC – the specific specification. Maddala (2001, p. 483) aptly describes this approach as “intended overparametrization with data-based simplification.”

In the general specification we account for the asset size of the bank, which influences the diversification of credit risk and the techniques used to evaluate and monitor credit risk. We characterize size, first, by the log transformation of consolidated assets and its square and, second, by the number of states in which the bank operates and by the interaction of the number of states and the total number of branches. The economic conditions in a bank’s lending markets are described by the 10-year weighted average growth rate of state GDP and by a weighted average of banking market concentration in each state in which the bank operates where the weights are the proportions of deposits located in each state. Additionally, we account for the proportion of total assets allocated to loans, which accounts in part for a bank’s focus on lending and the lending experience it is likely to have developed. We include the ratio of Tier 1 equity capital to consolidated assets to account for the cushion protecting a bank from loan losses and, hence, a component of the opportunity cost of loan losses. Since the credit analysis and monitoring of some types of loans are more difficult than other types, we control for important categories of loans as

8 Hendry (1983) provides the first complete application, and Campos, Ericsson, and Hendry (2005) survey this technique.

proportions of total loans in the general specification. Moreover, we include the average contractual interest rate on loans to account for a bank’s choice of credit risk. We find that it is important to allow for size effects and flexibility in the specification of the average contractual rate so we include the rate, the rate squared, and the interaction of the rate and the log transformation of assets, the interaction of the rate squared and the log transformation of assets, the interaction of the rate and the squared log transformation of assets, and the interaction of the squared rate and the squared log transformation of assets. The parameter estimates of all regressors in the general specification are reported in the second column of Table 9. The general specification is narrowed with an AIC- based search that estimates all possible specifications with all combinations of the above-mentioned regressors. The fourth column shows the parameter estimates of the regressors that survive the AIC-based search and define the specific specification.

The best specific specification reported in the fourth column of Table 9 provides evidence that higher ratios of loans to assets and equity capital to assets as well as a higher GDP growth rate are associated with lower lending inefficiency. The magnitude of the coefficient estimate on the capital ratio is notable. The median capital ratio of the full sample is 0.0960, and the median inefficiency, 0.0216. Hence, an increase of 0.01 in the capital ratio is associated with a decrease of

0.0014365 in the inefficiency ratio, which is a 6.7 percent decline in inefficiency. On the other hand, lending inefficiency is positively related to the number of states in which a bank operates. While operating in more states may tend to diversify credit risk, it also increases organizational complexity and makes managerial control more difficult. An increase by 1.0 in the number of states on average increases lending inefficiency by 0.0023, which is an increase of 10.7 percent over the median value of inefficiency, 0.0216. Increasing the number of branches, controlling for the number of states, is associated with reduced lending inefficiency, but the effect is quite small in magnitude and does not seem economically significant. While an increase in total business loans as a proportion of total loans is associated with increased lending inefficiency, an increase in the proportion of small business loans, holding the proportion of total business loans constant, is associated with reduced lending inefficiency – this is a substitution of small business loans for

larger business loans. The magnitude of these loan composition effects is small. For example, increase of 0.01 in the proportion of business loans implies an increase of 0.0004096 in lending inefficiency while a like increase in small business loans, a decrease of 0.0009910 in lending inefficiency. Given the median value of lending inefficiency for the full sample, 0.0216, the 0.01 increase in the business loan ratio implies an increase of 1.9 percent of the median inefficiency and the 0.01 increase in the small business loan ratio, a 4.6 percent decrease in the median inefficiency. The association of the average contractual interest rate with lending inefficiency is given by the derivative of inefficiency with respect to the interest rate.9 This derivative is positive for 704 of the 710 banks and statistically significant for 701; and, it is negative for 6 banks and statistically significant for 4 of the 6 banks. Thus, the higher credit risk associated with a higher contractual interest rate is also associated with higher lending inefficiency, which suggests these loans involving higher credit risk are more difficult to make and to monitor.

Do Capital Markets Distinguish between Inherent Credit Risk and Lending Efficiency

To obtain evidence on whether capital markets price inherent credit risk and lending inefficiency, in Table 10 we compare lending characteristics and financial performance between banks in the halves of the sample with lower and higher observed ratios of nonperforming loans. The half that experiences lower ratios of nonperformance holds on average less total assets, makes more loans as a proportion of assets, charges a lower average contractual rate on loans, takes on lower inherent credit risk, and is more efficient at lending. The higher mean Tobin’s q at these banks suggests capital markets price these lower ratios of nonperformance.

In Panels A and B of Table 11 these two groups composed of banks with lower and higher ratios of nonperforming loans are each further divided into the more and less efficient at lending. In both groups, the more efficient experience significantly lower ratios of overall nonperformance;

	From	, the derivative is	given by ∂(lending inefficiency)/∂(contractual interest rate) = 2.73758
+2[-1.02265][contractual interest rate][log(Book Value Assets(1,000s)].
9		Table 9

however, strikingly, there is little or no difference in the mean inherent credit risk and the contractual lending rate between the more and less efficient subsamples in each group although, as reported in Table 10, the group with a higher mean ratio of nonperformance is exposed to higher inherent credit risk and charges on average a higher contractual lending rate. While, in Table 10, publicly traded banks in the half of the sample with lower ratios of nonperformance on average experience a higher Tobin’s q ratio, in Table 11, banks in the more efficient partitions on average obtain a higher Tobin’s q ratio than those in the less efficient partitions.

These differences in the mean Tobin’s q ratio raise the question of whether capital markets simply penalize differences in nonperformance or whether these markets distinguish and price differently inherent credit risk and lending proficiency. Before investigating this issue, we compare the efficient halves of the banks with lower and higher ratios of nonperformance in Panel C of _efficientTable11. The efficient banks with a higher ratio of nonperformance are on average larger than the

banks with a lower ratio of nonperformance. Moreover, they assume higher inherent credit risk and charge a higher average contractual lending rate, and, while belonging to the more efficient half of the group with higher overall nonperformance, their mean lending inefficiency is significantly higher than that of the efficient half of the group with lower overall nonperformance. The lower q ratio in the efficient half of the group with a higher ratio of nonperformance suggests that nonperformance is penalized by capital markets.

4.1 Relationship of Financial Perform_capitalnce to Nonp_pricesrforming Loans (Loans with bad credit)

To obtain evidence on whether the market inherent credit risk and lending inefficiency, the sample is restricted to 244 publicly traded, top-tier holding companies at year-end 2013. Financial performance is measured by the market value of assets, expressed by Tobin’s q ratio. Tobin’s q ratio is given by the ratio of the market value of assets to the replacement cost of assets, whose commonly used proxy is the sum of the market value of equity and the book value of liabilities to the book value of assets.10 Market value has the advantage over accounting measures of

10 See Hughes and Mester (2010, 2015) for a review of the finance literature that uses Tobin’s ratio to

measure performance. q

performance in capturing the market’s expectation of the discounted value of the firm’s current and

future cash flow where the discount rate reflects the market’s assessment of the relevant risk

attached to the cash flow. In addition, market values permit investigation of investment incentives

provided by the capital market.

We first investigate the relationship of Tobin’s q ratio to the nonperforming loan ratio. From Table 7, we learned that small community banks and the largest banks on average charge relatively high contractual loan rates and exhibit relatively high ratios of nonperformance compared to the other three size groups. Are these high average ratios of nonperformance associated with better financial performance, which is to ask if the higher average contractual interest rate associated with the riskier lending results in a higher discounted cash flow that more than compensates for higher expected loan losses. After answering this question, we investigate the relationship of financial performance to the decomposition of the nonperforming loan ratio into inherent credit risk and lending inefficiency. We are particularly interested in answering the question, is the higher inherent credit risk of the largest banks associated with higher market value.

The relationship of Tobin’s q ratio to the nonperforming loan ratio is reported in Table 12. In the first column, the regressors of the general specification are listed and their parameter estimates and p-values, in the second and third columns. We include among the regressors the log transformation of the book value of assets and the square of this variable to control for differences among banks related to size, such as the potential for economies of scale due to better diversification, network economies, and spreading overhead costs. In addition, we account for the importance of lending in the business model of the bank by controlling for the ratio of loans to assets and the square of this ratio. We include the ratio of nonperforming loans to total loans and its square. And, we add its interaction with the log transformation of assets since size -related diversification may influence the risk associated with any given nonperforming loans ratio.

The general specification is narrowed with an AIC-based search that estimates all possible specifications with all combinations of the regressors in the general specification and declares the specification with minimum AIC the best specification. The fourth column shows the parameter

estimates of the regressors that survive the AIC-based search and define the specific specification.¹⁷ The key parameter estimates are those of the nonperforming loan ratio, −2.8247, and its interaction with the log transformation of assets, 0.1553. The sign of the derivative of the ratio with respect to the nonperforming loan ratio switches from negative to positive at assets equal to $79.3 billion. Thus, the derivative is positive for the 14 largest banks and negative for the 230 smaller banks; however, none of the 14 positive values of the derivative is statistically significant while 183 of the 230 negative values are statistically significant. Hence, financial performance and the nonperforming loans ratio are negatively related at 183 of the 244 banks.

The statistically insignificant positive relationship between the q ratio and the

nonperforming loan ratio for the largest banks as well as their high average observed ratio of nonperformance suggests there may be dichotomous credit risk strategies for maximizing value that differ between larger and smaller banks, which are not captured by the imprecisely estimated derivative for the large banks. We seek to improve the precision of these estimates by adding an indicator variable for very large banks to the set of regressors of the general specification. To define this indicator variable, we see from Table 4 that banks in the largest size group with assets exceeding $ 250 billion exhibit a much higher average ratio of nonperformance than smaller banks in the other four size groups. When banks are sorted by asset size, the largest 9 banks range in size from $ 175.4 billion to $2.4 trillion. The next largest bank is $129.7 billion. Thus, we set the lower bound on the indicator variable at $170 billion, and we interact it with variables involving the nonperforming loan ratio and add these variables to the general specification. The general specification is described in the first column of Table 13 and its parameter estimates, in the second column. The AIC based search over all possible specifications with combinations of variables taken from the general specification leads to the specific model in the fourth column. Based on the parameter estimates associated with the specific specification in the fourth column, the derivative of Tobin’s q ratio with respect to the nonperforming loan ratio is given by

∂Tobin’s /∂(nonperforming loan ratio) = −0.0391[ (Book Value Assets (1,000s))]

−14.3103(Indicatorq Variable Assets > $170 Billion) ln

+ 0.8325(Indicator Variable Assets > $170 Billion)[ln (Book Value Assets (1,000s))]. (7)

The third term in the derivative is positive when banks hold assets in excess of $170 billion, and it increases with the volume of assets. For the nine largest banks, the derivative is positive, and for 8 of the 9 banks, the derivative is statistically significant at stricter than 10 percent. Thus, the relatively high average observed ratio of nonperforming loans among these banks is associated with a higher market value. In the case of these very large banks, market discipline appears to encourage greater risk-taking in lending and, in this respect, tends to work against financial stability. For the remaining 235 smaller banks, the statistically significant negative sign of the derivative suggests that a higher ratio of nonperformance is associated with poorer financial performance. In the case of these smaller banks, market discipline appears to discourage risky lending.

4.2 Relationship of Financial Performance to Inherent Credit Risk and Lending Inefficiency

The regressions in Tables 12 and 13 provide evidence that the capital market rewards a lower nonperforming loan ratio at most banks and a higher ratio at the very largest banks. In regressions reported in Table 14, we investigate the degree to which the capital market distinguishes nonperformance due to inherent credit risk from that due to lending inefficiency. In the general specification, we substitute variables involving inherent credit risk and lending inefficiency for those involving the nonperforming loan ratio. The specific specification in Table 14 results from an AIC-based search that estimates all possible specifications with all combinations of regressors in the general specification and declares the specification with minimum AIC the best specification. The relationship of market value to inherent credit risk, the best-practice value of the nonperforming loans ratio, is given by the derivative of Tobin’s q ratio,

∂Tobin’s q/∂(inherent credit risk) = −6.8334 + 0.4566[ln(Book Value Assets (1,000s))]. (8)

Panel A of Table 15 reports that the sign of the derivative is negative at 133 smaller banks

but the derivative is not statistically significant at stricter than 0.10 at any of them. The sign reverses and the derivative becomes positive at 111 banks larger than $3.17 billion; however, it is statistically significant only at the largest 39 of these 111 banks. At these 39 banks, market discipline appears to encourage riskier lending.

Panel B of Table 15 lists these 39 banks. They are the largest banks. At these banks, making riskier loans is associated with improved financial performance. To evaluate the economic significance of these derivatives, consider the largest value, 3.032, obtained by JPMorgan Chase. From Table 8, the value of JP Morgan’s inherent credit risk is 0.0496. An increase of 0.005 in this value, about 10 percent, is associated with an increase of 0.01516, about 1.48 percent, in its q ratio, 1.02368. This incentive to make riskier loans is strongest for the four largest banks.

The value of the derivative of Zions, the smallest of the 17 banks larger than $50 billion that are subject to enhanced supervision under the Dodd-Frank Act, is 1.313, and for the smallest of the 39 banks exhibiting a positive, significant derivative, 0.766. Thus, the incentive to make riskier loans diminishes at smaller banks.

On the other hand, the statistically significant negative coefficient on lending inefficiency, −0.6188, in the specific specification reported in Table 14 indicates that the capital market penalizes a lack of proficiency in loan making at all banks.

The decomposition of the noise-adjusted nonperforming loan ratio into inherent credit risk and lending inefficiency shows that the credit market rewards inherent credit risk at the largest banks and penalizes lending inefficiency at all banks. A bank’s ratio of nonperforming loans reflects both the inherent credit risk it has assumed, its proficiency at evaluating credit and monitoring loans, and the statistical noise; hence, the relationship of its q ratio to its nonperforming loan ratio reflects the net influence of these three components. The evidence obtained on this net effect from the specific regression in Table 13 shows a statistically significant negative derivative with respect to the nonperforming loan ratio for 235 out of 244 banks in the publicly traded sample. From the

evidence of the specific regression in Tablethese14, it would appear that the influence of lending

inefficiency dominates and accounts for negative net value effects. For the 9 largest banks (Mega Bank = 1 in Table 13) with positive values of the derivative, the influence on financial performance of inherent credit risk dominates. When the influence of these two effects is modeled separately in the regressions in Table 14, the negative value effect of lending inefficiency is obtained by all banks. The positive value effect of making riskier loans appears at 111 banks and is statistically significant at 39 banks – all large banks. Hence, the decomposition uncovers evidence of a risk-taking incentive for lending at many more banks than the incentive revealed by the aggregate nonperforming loan ratio.

In sum, the net value effect of the nonperforming loan ratio suggests a dichotomous lending strategy to maximize value that differs between the largest banks and smaller banks. The nonperforming loan ratio is positively related to value at the largest banks and negatively related at smaller banks. However, the positive net effect at the largest banks is associated with making riskier loans while the negative net effect at smaller banks is related to reducing lending inefficiency. The evidence of a dichotomous lending strategy that differs between the largest financial institutions and smaller institutions is similar to evidence of dichotomous capital strategies found by Hughes, Mester, and Moon (2016), which also differs between large banks that at the margin improve value by reducing their capital ratio and smaller banks that at the margin improve value by increasing their capital ratio. Marcus (1984) demonstrates that pursuing a low-risk investment strategy to minimize the probability of financial distress maximizes value at banks with high-valued investment opportunities while adopting a high-risk investment strategy to exploit the option value of explicit and implicit deposit insurance maximizes value at banks with low valued investment opportunities. Hughes, Mester, and Moon (2016) show that smaller banks experience higher valued investment opportunities than larger banks. McConnell and Servaes (1995) find similar results for nonfinancial firms which they interpret as capital structure that addresses the overinvestment problem at firms with low valued investment opportunities and the underinvestment problem at firms with high valued investment opportunities.

Summary and Conclusions

We develop a novel technique to analyze nonperforming loans and apply it to data on top-tier bank holding companies at year-end 2013. The ratio of banks’ nonperforming loans to their total loans is decomposed into three components: first, a minimum ratio that represents the best-practice nonperforming loans ratio given the volume and composition of a bank’s loans, the average contractual interest rate charged on these loans, and market conditions such as the average GDP growth rate and market concentration; second, a ratio, which is the difference between the bank’s observed ratio of nonperforming loans and the best-practice minimum ratio, that represents the bank’s inefficiency at lending; third, a statistical noise. The best- practice ratio of nonperforming loans represents the inherent credit risk of the loan portfolio.

Bank holding companies are divided into five size groups. The largest banks with consolidated assets exceeding $250 billion experience the highest ratio of nonperformance among the five groups. Moreover, the inherent credit risk of their lending is the highest among the five groups. On the other hand, their inefficiency at lending is one of the two lowest among the five. Thus, the high ratio of nonperformance of the largest financial institutions appears to result from lending to riskier borrowers, not from inefficiency at lending. Small community banks also exhibit higher inherent credit risk than all other size groups except the largest banks; however, their lending inefficiency is the highest among the five size groups.

When the sample is restricted to publicly traded bank holding companies, market values can be used to gauge how financial performance is related to nonperforming loans. In contrast to accounting measures of performance, market value reveals the market’s expectation of future as well as current cash flows discounted by market-priced risk. The ratio of nonperforming loans to total loans is negatively related to financial performance except at the largest institutions, which suggests a dichotomous credit risk strategy for maximizing value that differs between the largest banks and smaller banks. The decomposition of the noise-adjusted nonperforming loan ratio into

inherent credit risk and lending inefficiency shows that market discipline rewards riskier lending at

large banks and discourages inefficient lending at all banks.

References
Berger, Allen N., and Loretta J. Mester (1997), “Inside the Black Box: What Explains Differences in
the Efficiencies of Financial Institutions,”			Journal of Banking and Fi ance			21, 895-947.
Campos, J., N.R. Ericsson, and D.F. Hendry (2005), “Introduction,” in J. Campos, N.R. Ericsson, and
D.F. Hendry (eds.),	General-to -Specific Modelling			, Edward Elgar Publishing, Cheltenham, 1-18.
Coelli, Tim, D. S. Prasada Rao, and George E. Battese (1998),
	, Kluwer Academic Publishers, Boston.				An Introduction to Efficiency and
Productivity Analysis
				, 7 edition, Prentice Hall, Upper Saddle River, 435-
Greene, William H. (2012),
436 and 839-845.		Econometric Analysis th				Scottish

Hendry, D.F. (1983), “Econometric Modelling: The ‘Consumption Function’ in Retrospect,”
Journal of Political Economy		30, 193-220.
Hughes, Joseph P., Julapa Jagtiani, and Loretta J. Mester (2017), “Is Bigger Necessarily Better in
Community Banking?” (supersedes, Federal Reserve Bank of Cleveland, Working Paper 1615).
Hughes, Joseph P., and Loretta J. Mester (2010), “Efficiency in Banking: Theory, Practice, and
Evidence,” in			edited by A.N. Berger, P. Molyneux, and J. Wilson,
Oxford University Press, Oxford, 463–485.
	The Oxford Handbook of Banking,
Hughes, Joseph P., and Loretta J. Mester (2015), “Measuring the Performance of Banks: Theory,
Practice, Evidence, and Some Policy Implications,” in							second
edition, edited by Allen N. Berger, Philip Molyneux, and John Wilson, Oxford University Press,
Oxford, 247–270.				The Oxford Handbo k of Banking,
						“Market Discipline Working For
Hughes, Joseph P., Loretta J. Mester, and Choon-Geol Moon (2016),
and Against Financial Stability: The Two Faces of Equity Capital in U.S. Commercial Banking,”
Department of Economics, Rutgers University, Working Paper 201611.
Jondrow, J., C.A.K. Lovell, I.S. Materov, and P. Schmidt (1982). “On the Estimation of Technical
Efficiency in the Stochastic Frontier Production Function Model,”								19, 233–
238.					Journal of Econometrics

Kumbharkar, Subal C., and C. A. Knox Lovell (2000), Stochastic Frontier Analysis, Cambridge
University Press, Cambridge.
Marcus, A.J. (1984), “Deregulation and Bank Financial Policy,” Journal of Banking and Finance 8,
557-565.				Journal of
McConnell, J.J. and H. Servaes (1995). “Equity Qwnership and the Two Faces of Debt,”
Financial Economics	39, 131-157.
Morgan, Donald P., and Adam B. Ashcraft (2003), “Using Loan Rates to Measure and Regulate Bank
Risk: Findings and an Immodest Proposal,”		Journal of Financial Services Research	24:2/3, 181–200.

Petersen, Mitchell A., and Raghuram G. Rajan (1995), “The Effect of Credit Market Competition on Lending Relationships,” Quarterly Journal of Economics 110:2, 407–443.

Figure 1 (Loans with bad credit)

The log transformation of nonperforming loans is plotted against the log transformation of total loan volume for 710 top-tier bank holding companies at year-end 2013. These 710 companies represent banks whose loan volume exceeds 15 percent of consolidated assets and whose nonperforming loans (including gross charge-offs and foreclosed real estate) amount at least to 1 percent and no more than 15 percent of total loans. While it appears that, for any given volume of loans, the degree of nonperformance is wide, it is important to remember that some of this wide variation in nonperformance is due to differences in the average contractual interest rate, the composition of the loan portfolio, the GDP growth rate, and market concentration.

LOESS curve (the thick curve in the above figure) and its 95% confidence interval (the shaded band) are included. LOESS is short for locally estimated scatterplot smoothing, which fits local polynomial regression model to scatter points. LOESS is one of the most popular local smoothing methods and is robust to a long-tailed error distribution while it is highly efficient when the error distribution is normal.

Figure 2 (Loans with bad credit)

The ratio of nonperforming loans as a proportion of total loans is plotted against the log transformation of total loan volume for 710 top-tier bank holding companies at year-end 2013. These 710 companies represent banks whose loan volume exceeds 15 percent of consolidated assets and whose nonperforming loans (including gross charge-offs and foreclosed real estate) amount at least to 1 percent and no more than 15 percent of total loans. While it appears that, for any given volume of loans, the degree of nonperformance is wide, it is important to remember that some of this wide variation in nonperformance is due to differences in the average contractual interest rate, the composition of the loan portfolio, the GDP growth rate, and market concentration.

Figure 3 (Loans with bad credit)

Best-Pra

Loan Nonperforma

ce Frontier11

This figure illustrates the best- practice minimum ratio of nonperforming loans to total loans that is

obtained by stochastic frontier estimation of the relationship between the nonperforming loan ratio

and total loans (expressed in 100 billions), controlling for the loan portfolio composition, the

average contractual lending rate, and the GDP growth rate and market concentration in the bank’s

market.₂ The error term,

i =

+ , is a composite term used to distinguish statistical noise,₂

i ~ iid

(0,

), from the term,

i, which is a positive, half-normal error term,

( 0) ~ iid N(0,σμ ), that

measures the systematic excess nonperformance from bank ’s best-practice minimum

σν

minimum

≥

nonperforming loan ratio. The ‘expected’ best-practice

nonperforming loan ratio is given

by the deterministic kernel of the estimated function.

) and experiences a nonperforming loan

In this example, bank has total loans of 0.08 ($8 billions

ratio adjusted for statistical noise,

i −

i, of 0.025, which is an excess of 0.015 over the ‘expected’

best-practice minimum,

i, of 0.010. Thus, its

inefficiency is 0.015.

FNP

lending

11 Adapted from Hughes, Jagtiani, and Mester (2016)

12 0.08 (in 100 billons, here) = 8,000,000 (in thousands in the Y9-C data report) = 8 billions.

	Figure 4 (Loans with bad credit)
	Best-Practice and Observed Nonperforming Loan Ratios
This figure illustrates for banks with total loans between $ 8.9 billion $967.3 billion. The estimated
best-practice minimum ratio of nonperforming loans to total loans is obtained by stochastic frontier
estimation of the relationship between the nonperforming loan ratio and total loans, controlling for
the loan portfolio composition, the average contractual lending rate, and the GDP growth rate and
market concentration in the bank’s market. The ‘expected’ best- practice minimum nonperforming
loan ratio is given by the deterministic kernel of the estimated function, which gauges inherent
credit risk. The excess nonperforming loan ratio, a measure of a bank’s proficiency at credit risk
assessment and loan monitoring, is given by the difference between the noise-adjusted observed
ratio (	) and the best-practice minimum (	). These values are indicated for the four
largest financial institutions.		red +
blue +

					Ln(Total Loans)
NAME	ln(Total	Noise-Adjusted	Best-Practice	Excess NP Loan Ratio
	Loans)	Observed NP Ratio NP Loan Ratio (Lending Inefficiency)

Bank Of America Corporation	20.69	0.072	0.059	0.013
Wells Fargo & Company
JP Morgan Chase & Co.	20.55	0.080	0.054	0.026
Citigroup Inc.	20.46	0.056	0.050	0.006
U.S. Bancorp	20.33	0.057	0.053	0.004
	19.28	0.048	0.022	0.026

Table 1 (Loans with bad credit)

The initial data set includes 807 top-tier bank holding companies at the end of 2013. They are sorted in
	Banks with the Lowest Ratio of Total Loans to Consolidated Assets at Ye -End 2013
ascending order by the ratio of total loans to total assets up to 0.40. To estimate the best-practice loan
performance frontier, the sample is trimmed of companies with a ratio less than 0.15, which eliminates 4
companies that are not focused on the typical functions of commercial banking.
	Name of Bank Holding Company		Book Value of Total Assets		Total Loans/Total Assets
			(1000s)
2	State Street Corporation		243,028,090	0.05549
1	Goldman Sachs		911,595,000		0.08951

3
	Morgan Stanley	832,702,000		0.09235
4
	Bank Of New York Mellon		374,310,000		0.13771
5
6	BRISCOE RANCH	1,246,790		0.16534
7	VILLAGES BC	1,579,895		0.17964
8	INDUSTRY BSHRS	2,480,503		0.18925
9	Stifel Financial Corp.	9,008,870		0.22830
10	FIRST CMNTY BSHRS	1,368,359		0.27399
11	FIRST TX BC	910,068		0.28382
12	Northern Trust Corporation	102,947,333		0.28544
13	FIRST BC	578,509		0.29945
14	INDEPENDENT BKR FC	2,188,247		0.30389
15	FARMERS ENT	694,668		0.30855
16	SECURITY BC TN	450,765		0.31137
17	JPMorgan Chase & Co.	2,415,689,000		0.31672
18	MIDWEST INDEP BSHRS	360,387		0.32309
19	EAST TX BSHRS	537,204		0.32400
20	FIRST AMER BK CORP	3,379,295		0.32498
21	COCONUT GROVE BSHRS	591,480		0.32878
22	SNBNY HOLD	6,672,456		0.33274
23	DICKINSON FC II	2,100,898		0.34222
24	N W SVC CORP	354,765		0.34324
25	Citigroup Inc.	1,880,382,000		0.36010
26	FIRST NM FC	427,134		0.36063
27	MIDLAND FC	653,403		0.36307
28	CU BK SHARES	568,135		0.36339
29	OTTAWA BSHRS	488,233		0.36534
30	Century Bancorp, Inc.	3,431,154		0.36861
31	PACIFIC COAST BKR BSHRS	545,753		0.37425
32	Westamerica Bancorporation	4,848,866		0.37694
33	STURM FNCL GROUP INC	2,059,318		0.37805
34	National Bank Holdings Corpora	4,914,115		0.37848
35	BLACKHAWK BC	1,101,428		0.38149
36	AMERICAN BC	1,000,668		0.38369
37	Umb Financial Corporation	16,911,852		0.38564
38	Cullen/Frost Bankers, Inc.	24,388,272		0.39021
39	RED RIVER BC	707,400		0.39058
40	CBS BANC CORP	1,482,676		0.39072
41	Southside Bancshares, Incorpor	3,445,663		0.39221
	ROCKHOLD BANCORP	536,342		0.39785

Table 2 (Loans with bad credit)

Nonperformance Banks with is measured the Highest by the Proportion following data of from Non performing the Y9-Creport Loans at year-at end Year 2013-End.Non performing 2013 Loans = past due loans less than and more than 90 days plus nonaccruing loans, lease financing receivables, placements, and other assets (BHCK525+BHCK5524 +BHCK5526); Nonperforming loans + Gross Charge-offs (BHCK4635); Nonperforming Loans + Gross Charge- offs + Other Real Estate Owned (BHCK2150); Total Loans (BHCK2122). Banks with a ratio (Nonperforming Loans + Gross Charge-offs + Other Real Estate Owned) / (Total Loans) > 0.15 are trimmed. The list below exhibits the 50 banks with the highest ratios.

					Book Value				(Nonperforming		(Nonperforming
					Book Value		(Nonperforming		(Nonperforming		Loans + Gross
		Name of Bank			of Total		(Nonperforming		Loans + Gross		Loans + Gross
		Name of Bank			of Total		Loans) / (Total		Loans + Gross		Charge-offs + Other
		Holding Company			Assets		Loans) / (Total		Charge-offs) /		Charge-offs + Other
		Holding Company			Assets		Loans)		Charge-offs) /		Real Estate Owned)
		BEACH CMNTY			(1000s)		Loans)		(Total Loans)		Real Estate Owned)
					(1000s)				(Total Loans)		/ (Total Loans)
											/ (Total Loans)
	1			562,450		0.36595		0.37205		0.55477
		BSHRS
2		BSHRS
		BUILDERS FC		242,192		0.04101		0.04731		0.41082
	3	BUILDERS FC		242,192		0.04101		0.04731		0.41082
		_OFNORTHWEST_IL	BC	280,762		0.02752		0.03207		0.40509
4		_OFNORTHWEST_IL	BC	280,762		0.02752		0.03207		0.40509
		GEORGIA BSHRS		312,641		0.23661		0.24598		0.34973
5		HOMEBANCORP		706,902		0.30363		0.30389		0.30399
	6	METROPOLITAN		2,430,812		0.25052		0.25058		0.30245
		BK GRP
7		BK GRP
		SEAWAY BSHRS		556,584		0.24470		0.26986		0.29428
	8	CAPITOL CITY		286,974		0.15994		0.17402		0.27172
		BSHRS
	9	BSHRS
		_FCPUTNAM-GREENE		504,882		0.19133		0.20483		0.25619
10		_FCPUTNAM-GREENE		504,882		0.19133		0.20483		0.25619
		Porter Bancorp, Inc.		1,076,121		0.15994		0.20590		0.24944
	11	NORTHERN ST FC		392,412		0.11239		0.18098		0.24471
12		PERSONS BKG CO		361,270		0.12118		0.12901		0.22417
13		UNITED NAT CORP		2,344,689		0.04787		0.20891		0.20913
	14	Doral Financial		8,493,455		0.17458		0.18714		0.20884
		Corporation
15		Corporation
		CAPITOL BC		915,264		0.11639		0.14365		0.20831
16		FMB BSHRS		556,134		0.07232		0.08801		0.20669
	17	N W SVC CORP		354,765		0.10618		0.11956		0.20562
	18	DIAMOND		726,618		0.10409		0.12281		0.20537
		BANCORP
19		BANCORP
		_BSHRHAMILTON ST		1,596,351		0.13725		0.14508		0.20074
20		_BSHRHAMILTON ST		1,596,351		0.13725		0.14508		0.20074
		ALIKAT INV		292,174		0.05005		0.10611		0.19372
21		HCSB FC		434,819		0.06310		0.09225		0.18964
	22	LIBERTY SHARES		593,103		0.12969		0.14816		0.18950
23		CERTUSHOLDINGS		1,655,555		0.09777		0.11354		0.18216


									(Nonperforming
			Book Value				(Nonperforming		(Nonperforming
			Book Value		(Nonperforming		(Nonperforming		Loans + Gross
	Name of Bank		of Total		(Nonperforming		Loans + Gross		Loans + Gross
	Name of Bank		of Total		Loans) / (Total		Loans + Gross		Charge-offs + Other
	Holding Company		Assets		Loans) / (Total		Charge-offs) /		Charge-offs + Other
	Holding Company		Assets		Loans)		Charge-offs) /		Real Estate Owned)
	TENNESSEE ST		(1000s)		Loans)		(Total Loans)		Real Estate Owned)
			(1000s)				(Total Loans)		/ (Total Loans)
									/ (Total Loans)
24		658,104		0.09991		0.11401			0.17705
	BSHRS	762,264		0.11288		0.13985		0.16550
25	BSHRS
	Peoples Financial
	Corporation	12,656,925		0.09496		0.13891		0.15543
26	Corporation
	First Bancorp
27	COMMUNITY FIRST	449,275		0.07813		0.08564		0.15255
28	DICKINSON FC II	2,100,898		0.02702		0.07734		0.14922
29	Village Bank And	444,091		0.07253		0.09172		0.14849
	Trust Financial	1,008,007		0.05767		0.08269		0.14701
30	Trust Financial
	BRIDGEVIEW BC
31	OXFORD FC	446,836		0.08085		0.08831		0.14648
32	EDUCATIONAL SVC	2,665,828		0.14390		0.14466		0.14493
	OF AMERICA	671,885		0.11074		0.11899		0.14060
33	OF AMERICA
	FIRST NAT BSHRS
34	LINCO BSHRS	654,483		0.05586		0.07048		0.14026
35	SAINT LOUIS	401,683		0.04763		0.05787		0.13994
	BSHRS	495,791		0.10117		0.11801		0.13902
36	BSHRS
	F&M FC
37	Popular, Inc.	35,749,000		0.09074		0.12530		0.13793
38	SOUTHERN BSHRS	2,269,581		0.10391		0.11409		0.13621
39	CHAMBERS BSHRS	751,058		0.03865		0.05735		0.13438
40	AMBOY BC	2,148,004		0.05589		0.05836		0.13324
41	CITIZENS NBC	504,803		0.08539		0.09090		0.12721
43	BSHRS
42	NEW PEOPLES	684,711		0.08262		0.09488		0.12703
	MESABA BSHRS	642,830		0.07810		0.08537		0.12608
	MESABA BSHRS	642,830		0.07810		0.08537		0.12608
44	United Security	569,801		0.05266		0.09497		0.12498
	Bancshares, Inc
45	Bancshares, Inc
	SECURITY CAPITAL	500,191		0.04085		0.04499		0.12447
46	FLORIDA CAP GRP	410,618		0.06221		0.07526		0.12427
47	SEQUATCHIE	600,935		0.08466		0.09264		0.12352
	VALLEY BSHRS	1,030,607		0.07774		0.10454		0.12168
48	VALLEY BSHRS
	INLAND BC
49	BARABOO BC	632,567		0.06882		0.10252		0.12167
50	American Founders	276,706		0.05572		0.06602		0.12134

Table 3 (Loans with bad credit)

Non performance Banks with is measured the Lowest by the Proportion following data of from Non performing theY9-Creport Loa.Non performing sat Year-End Loans 2013=past due loans less than and more than 90 days plus nonaccruing loans, lease financing receivables, placements, and other assets (BHCK525+BHCK5524 +BHCK5526); Nonperforming loans + Gross Charge-offs (BHCK4635); Nonperforming Loans + Gross Charge- offs + Other Real Estate Owned (BHCK2150); Total Loans (BHCK2122). Banks with a ratio (Nonperforming Loans + Gross Charge-offs + Other Real Estate Owned) / (Total Loans) < 0.01 are trimmed. The list below exhibits the 70 banks with the lowest ratios.

								(Nonperforming		(Nonperforming
				Book Value						(Nonperforming
		Name of Bank		Book Value		(Nonperforming				Loans + Gross
		Name of Bank		of Total		(Nonperforming		Loans + Gross		Loans + Gross
		Holding		of Total		Loans) / (Total		Loans + Gross		Charge-offs + Other
		Holding		Assets		Loans) / (Total		Charge-offs)		Charge-offs + Other
		Company		Assets		Loans)		Charge-offs)		Real Estate Owned)
		Company		(1000s)		Loans)		/ (Total Loans)		Real Estate Owned)
										/ (Total Loans)


		TOLLESON	429,133		0.000510			0.000510		0.000510
	1	TOLLESON	429,133		0.000510			0.000510		0.000510
		WEALTH MGMT
2		WEALTH MGMT
		FIDELITY HC	403,660		0.000000		0.000134		0.000596
3		FIDELITY HC	403,660		0.000000		0.000134		0.000596
		AMERICAN BK	528,961		0.000724		0.000730		0.000730
		INC
4		INC
		MERCHANTS BC	1,179,418		0.000517		0.000869		0.001177
5		MERCHANTS BC	1,179,418		0.000517		0.000869		0.001177
		Cardinal Financial	2,894,230		0.001169		0.001260		0.001260
		Corporation	1,730,936		0.000945		0.001297		0.001389
6		Corporation
		Merchants
		Bancshares, Inc.
7		Bancshares, Inc.
		LEADER BC	658,722		0.002205		0.002571		0.002571
8		LEADER BC	658,722		0.002205		0.002571		0.002571
		SNBNY HOLD	6,672,456		0.002503		0.002813		0.002813
9		SNBNY HOLD	6,672,456		0.002503		0.002813		0.002813
		CBX CORP	1,159,027		0.001387		0.003037		0.003037
10		CBX CORP	1,159,027		0.001387		0.003037		0.003037
		FIRSTPERRYTON	963,592		0.002691		0.003063		0.003063
		BC
11		BC
		FINANCIAL CORP	631,803		0.002865		0.003328		0.003328
		OF LA	2,399,892		0.003165		0.003795		0.003795
12		OF LA
		First Of Long
		Island Corp
13		Island Corp
		SIGNATURE BC	646,087		0.002682		0.004044		0.004044
14		SIGNATURE BC	646,087		0.002682		0.004044		0.004044
		CAMBRIDGE BC	1,533,711		0.004045		0.004110		0.004110
15		CAMBRIDGE BC	1,533,711		0.004045		0.004110		0.004110
		HOMETOWN BC	1,153,654		0.004148		0.004263		0.004263
16		HOMETOWN BC	1,153,654		0.004148		0.004263		0.004263
		FIRST TX BHC	1,299,111		0.003216		0.004281		0.004281
17		FIRST TX BHC	1,299,111		0.003216		0.004281		0.004281
		Access National	847,181		0.003563		0.004324		0.004324
		Corporation
18		Corporation
		INWOOD BSHRS	1,569,067		0.002431		0.002482		0.004384
19		INWOOD BSHRS	1,569,067		0.002431		0.002482		0.004384
		State Street Corp	243,028,090		0.000062		0.000062		0.004439
	20	State Street Corp	243,028,090		0.000062		0.000062		0.004439
		POST OAK BSHRS	772,461		0.004011		0.004747		0.004747
21		POST OAK BSHRS	772,461		0.004011		0.004747		0.004747
		LEACKCO BHC	581,422		0.002289		0.002801		0.004772
22		LEACKCO BHC	581,422		0.002289		0.002801		0.004772
		Pacific Premier	1,714,187		0.002338		0.003972		0.004926
		Bancorp, Inc.	1,200,777		0.004812		0.005275		0.005275
	23	Bancorp, Inc.
		CUMMINS-AMER
		CORP
24		CORP
		MIDLAND BSHRS	1,076,128		0.004980		0.005294		0.005294
25		MIDLAND BSHRS	1,076,128		0.004980		0.005294		0.005294
		Center Bancorp,	1,673,082		0.004812		0.005154		0.005383
		Inc.	826,012		0.004845		0.005601		0.005601
	26	Inc.
		FIRST BSHRS OF
		TX
27		TX
		PLANTERS FNCL	784,115		0.003947		0.005336		0.005840
		GRP	1,393,217		0.004581		0.005768		0.005847
	28	GRP
		INDEPENDENCE
		BSHRS
29		BSHRS
		HERITAGE GROUP	624,052		0.005856		0.005863		0.005863
30		HERITAGE GROUP	624,052		0.005856		0.005863		0.005863
		GNB BC	495,827		0.005474		0.005563		0.005916


				Book Value				(Nonperforming		(Nonperforming
		Name of Bank		Book Value		(Nonperforming		(Nonperforming		Loans + Gross
		Name of Bank		of Total		(Nonperforming		Loans + Gross		Loans + Gross
		Holding		of Total		Loans) / (Total		Loans + Gross		Charge-offs + Other
		Holding		Assets		Loans) / (Total		Charge-offs) /		Charge-offs + Other
		Company		Assets		Loans)		Charge-offs) /		Real Estate Owned)
		Company		(1000s)		Loans)		(Total Loans)		Real Estate Owned)
										/ (Total Loans)


32		MBT BSHRS	562,965		0.000000		0.003311			0.005993
31		SECURITY NAT	1,218,571		0.003863		0.004625		0.006190


		CORP
33		CORP
		Independent Bk	2,163,984		0.003535		0.004369		0.006293
		Grp Inc	614,586		0.003964		0.004108		0.006437
34		Grp Inc
		MIDWEST BANC
		HC	745,840		0.004719		0.005762		0.006458
35		HC
		CITIZENS NAT
		BSHRS OF
		BOSSIER
36		BOSSIER
		MVB FC	991,730		0.003913		0.006055		0.006580
37		MVB FC	991,730		0.003913		0.006055		0.006580
		RBB BC	723,410		0.002074		0.004106		0.006673
38		RBB BC	723,410		0.002074		0.004106		0.006673
		RED RIVER BSHRS	1,301,336		0.005064		0.005679		0.006726
39		RED RIVER BSHRS	1,301,336		0.005064		0.005679		0.006726
		FARMERS &	2,076,722		0.002915		0.003557		0.006879
		MRCH BC	705,560		0.006696		0.007112		0.007112
40		MRCH BC
		DAKOTA CMNTY
		BSHRS
41		BSHRS
		LAURITZEN CORP	1,832,705		0.006729		0.007248		0.007302
42		LAURITZEN CORP	1,832,705		0.006729		0.007248		0.007302
		Bank Of New York	374,310,000		0.006732		0.007062		0.007469
		Mellon Corpor
43		Mellon Corpor
		BOU BC	834,970		0.006925		0.007587		0.007587
44		BOU BC	834,970		0.006925		0.007587		0.007587
		MANHATTAN BC	1,099,989		0.007041		0.007790		0.007808
45		MANHATTAN BC	1,099,989		0.007041		0.007790		0.007808
		Century Bancorp,	3,431,154		0.006412		0.007846		0.007846
		Inc.
46		Inc.
		NORTHERN BC	1,096,296		0.007499		0.007916		0.007916
47		NORTHERN BC	1,096,296		0.007499		0.007916		0.007916
		AMERICAN BC	1,192,869		0.006221		0.007910		0.007963
	48	AMERICAN BC	1,192,869		0.006221		0.007910		0.007963
		SECURITY NAT	744,886		0.005745		0.006530		0.007970
		CORP	1,966,948		0.006080		0.006740		0.007971
49		CORP
		Peapack-
		Gladstone
		Financial Co
50		Financial Co
		BPC CORP	662,640		0.001452		0.005943		0.008005
51		BPC CORP	662,640		0.001452		0.005943		0.008005
		Northrim	1,215,006		0.003601		0.005288		0.008300
		Bancorp, Inc.
52		Bancorp, Inc.
		STATE BSHRS	2,946,169		0.006175		0.007173		0.008397
53		STATE BSHRS	2,946,169		0.006175		0.007173		0.008397
	54	Inc.	1,896,640		0.005302		0.006206		0.008418
		Bridge Bancorp,	1,896,640		0.005302		0.006206		0.008418
		BANK OF	1,143,494		0.001941		0.004944		0.008699
		BANK OF	1,143,494		0.001941		0.004944		0.008699
		HIGHLAND PK
		FNCL CORP	9,008,870		0.006522		0.008648		0.008712
55		FNCL CORP
		Stifel Financial
		Corp.	708,060		0.007100		0.007964		0.008730
56		Corp.
		AMERICAN
		CENTRAL
		BANCORP
57		BANCORP
		AMERICAN BK	1,205,154		0.004398		0.006899		0.008782
		HOLDING CORP	1,060,887		0.003963		0.005174		0.008788
58		HOLDING CORP
		FIRST
		MANITOWOC BC	562,783		0.008540		0.008726		0.008891
59		MANITOWOC BC
		FARMERS ST
		CORP	889,577		0.001020		0.003005		0.008971
	60	CORP
		HOLD
		AVENUE FNCL


				Book Value				(Nonperforming		(Nonperforming
		Name of Bank		Book Value		(Nonperforming		(Nonperforming		Loans + Gross
		Name of Bank		of Total		(Nonperforming		Loans + Gross		Loans + Gross
		Holding		of Total		Loans) / (Total		Loans + Gross		Charge-offs + Other
		Holding		Assets		Loans) / (Total		Charge-offs) /		Charge-offs + Other
		Company		Assets		Loans)		Charge-offs) /		Real Estate Owned)
		Company		(1000s)		Loans)		(Total Loans)		Real Estate Owned)
				(1000s)				(Total Loans)		/ (Total Loans)
										/ (Total Loans)
	61	Prosperity	18,651,693		0.007357		0.008063			0.009002
		Bancshares, Inc.
62		Bancshares, Inc.
		FIRST-WEST TX	916,226		0.007154		0.008521		0.009006
		BSHRS	800,678		0.003242		0.004446		0.009185
63		BSHRS
		METROPOLITAN
		BANCGROUP
64		BANCGROUP
		CITIZENS BSHRS	665,350		0.007725		0.008268		0.009511
65		CITIZENS BSHRS	665,350		0.007725		0.008268		0.009511
		IDA GROVE	1,245,520		0.008319		0.009363		0.009531
		BSHRS	950,266		0.005302		0.005724		0.009688
66		BSHRS
		Southern Missouri
		Bancorp, Inc
67		OTTAWA BSHRS	488,233		0.006072		0.008381		0.009889
68
		CENTRAL BC	162,259		0.007789		0.007985		0.009941
69		CENTRAL BC	162,259		0.007789		0.007985		0.009941
		BANKWELL FNCL	779,618		0.008426		0.008696		0.010013
		GRP	2,117,679		0.006255		0.007095		0.010044
70		GRP
		RCB HC

Table 4 (Loans with bad credit)

The data set includesSmmary710topStatistics-tierbankofholdingTop-TiercompaniesBank HoldingattheendCompaniesof2013.NonperformingatYear-Endloans2013include past due loans less than and more than 90 days plus nonaccruing loans, lease financing receivables, placements, and other assets (BHCK525+BHCK5524 +BHCK5526); gross charge-offs (BHCK4635); and other real estate owned (BHCK2150) . Banks whose nonperforming loans exceed 15 percent of total loans as well as those whose nonperforming loans are less than 1 percent of total loans are dropped. And banks with total loans less than 15 percent of consolidated assets are dropped. These restrictions reduce the initial sample of 807 banks to 710.

Panel A: Consolidated Assets < $1 Billion (Loans with bad credit)

Assets^BookValue(1,000s)	N			Mean			Median	Std. Dev.	Minimum	Maximum

	364		662,973				654,468	167,443	92,694	998,762
Loans/Assets	364		0.6400				0.6558	0.1267	0.2838	0.9142
(Nonperforming	364	Panel B:^0.0461					0.0370	0.0310	0.0100	0.1485
Loans) / Loans	364	Panel B:^0.0461			<		0.0370	0.0310	0.0100	0.1485
Assets^BookValue(1,000s)				$1 Billion	<	Consolidated Assets < $10 Billion
				$1 Billion		Consolidated Assets < $10 Billion
	293		2,600,335				1,853,823	1,918,892	1,000,668	9,641,427
	293		2,600,335				1,853,823	1,918,892	1,000,668	9,641,427
Loans/Assets	293		0.6414				0.6537	0.1229	0.1653	0.9216
(Nonperforming	293		0.0383				0.0280	0.0277	0.0100	0.1492
Loans) / Loans		Panel C: $10 Billion			<		Consolidated	Assets < $50 Billion
					<
Assets^BookValue(1,000s)	35		21,161,256				18,473,488	9,330,543	10,989,286	47,138,960
Assets^BookValue(1,000s)	35		21,161,256				18,473,488	9,330,543	10,989,286	47,138,960
Loans/Assets	35		0.6322				0.6588	0.1366	0.3856	0.9621
(Nonperforming	35		0.0312				0.0264	0.0219	0.0102	0.1379
Loans) / Loans					<
Loans) / Loans		Panel D: $50 Billion				Consolidated Assets < $250 Billion

Assets^BookValue(1,000s)	11		104,276,621				92,991,716	43,640,904	56,031,127	183,009,992
Assets^BookValue(1,000s)	11		104,276,621				92,991,716	43,640,904	56,031,127	183,009,992
Loans/Assets	11		0.6678				0.6958	0.1388	0.2854	0.8290
(Nonperforming	11		0.0307				0.0283	0.0099	0.0146	0.0449
Loans) / Loans
Loans) / Loans
				Panel E: Consolidated Assets > $250 Billion
Assets^BookValue(1,000s)	7		1,272,854,333		1,527,015,000			923,465,518	297,282,098	2,415,689,000
Assets^BookValue(1,000s)	7		1,272,854,333		1,527,015,000			923,465,518	297,282,098	2,415,689,000
Loans/Assets	7		0.5174				0.5501	0.1409	0.3167	0.6656
_Loans)(Nonperforming_/Loans	7		0.0610				0.0569	0.0118	0.0483	0.0798

Table 5 (Loans with bad credit)

Holding companies The whose Largest consolidated Banks and assets Thexceedir Loan $50 Perfect billion perfomance are listed at below Year.–Non performance End 2013 is measured by the following data from the Y9-C report at year-end 2013 divided by Total Loans (BHCK2122). Nonperforming = past due loans less than and more than 90 days plus nonaccruing loans, lease financing receivables, placements, and other assets (BHCK525+BHCK5524 +BHCK5526); Nonperforming loans + Gross Charge-offs (BHCK4635); Nonperforming Loans + Gross Charge-offs + Other Real Estate Owned (BHCK2150).


									(Nonperforming
							(Nonperforming		(Nonperforming
	Name of Bank		Book Value of		(Nonperforming		(Nonperforming		Loans + Gross
	Name of Bank		Book Value of		(Nonperforming		Loans + Gross		Loans + Gross
	Holding		Total Assets		Loans) / (Total		Loans + Gross		Charge-offs + Other
	Holding		Total Assets		Loans) / (Total		Charge-offs) /		Charge-offs + Other
	Company		(1000s)		Loans)		Charge-offs) /		Real Estate Owned)
	Company		(1000s)		Loans)		(Total Loans)		Real Estate Owned)
							(Total Loans)		/ (Total Loans)
									/ (Total Loans)
1	JPMorgan Chase	2,415,689,000		0.042777		0.052537			0.056131
	& Co.
2	Bank Of America	2,104,995,000		0.058744		0.069468		0.071591
	Corporation
3	Citigroup Inc.	1,880,382,000		0.033841		0.056329		0.056944
4	Wells Fargo &	1,527,015,000		0.067471		0.075102		0.079751
	Company
5	U.S. Bancorp	364,021,000		0.036162		0.044326		0.048340
6	PNC Financial	320,596,232		0.037558		0.045575		0.048626
	Services Group,
7	Capital One	297,282,098		0.037057		0.064881		0.065447
	Financial Corp
8	BB&T Corp	183,009,992		0.028152		0.036616		0.038728
9	Suntrust Banks	175,380,779		0.024413		0.031124		0.033224
10	Fifth Third	129,685,180		0.017885		0.024993		0.028333
	Bancorp
11	Regions	117,661,732		0.027746		0.039622		0.041359
	Financial Corp
12	Northern Trust	102,947,333		0.016903		0.018918		0.019324
	Corporation
13	Keycorp	92,991,716		0.018109		0.024203		0.024622
14	M&T Bank Corp	85,162,391		0.036329		0.040286		0.041329
15	Discover	79,339,664		0.018484		0.044931		0.044931
	Financial
	Services
16	Comerica Inc	65,356,580		0.010943		0.014293		0.014566
17	Huntington	59,476,344		0.018639		0.025698		0.026335
	Bancshares Inc
18	Zions	56,031,127		0.020123		0.023459		0.024635
	Bancorporation

Table 6 (Loans with bad credit)

The data set Estimation includes 710 of top Stochastic-tier bank Best holding-Practice companies Loan at the Non performance end of 2013.Non performing Frontier loans include past due loans less than and more than 90 days plus nonaccruing loans, lease financing receivables, placements, and other assets (BHCK525+BHCK5524 +BHCK5526); gross charge-offs (BHCK4635); and other real estate owned (BHCK2150) . Banks whose nonperforming loans exceed 15 percent or are less than 1 percent of total loans are dropped from the initial sample. Stochastic frontier techniques are used to estimate the best-practice minimum ratio of nonperforming loans to total loans for any given amount of total loans, expressed in 100 billions, controlling for the composition of the bank’s loan portfolio, the average contractual interest rate charged on its loans, the ten-year average GDP growth rate and market concentration in the states in which the bank operates, where each state’s datum is weighted by the proportion of the bank’s total deposits located in the state. Parameters significantly different from zero at the 10 percent or stricter are given in bold.

Coefficient

Parameter

Variable

Pr > |t|

Estimate

α 0

Intercept

-0.011187

0.032787

total loansi (100 billons)

0.005420

0.000000

α ₁

Contractual lending ratei

0.434162

0.000000

Herfindahl index of market concentrationi

-0.006387

0.394509

GDP growth ratei

-0.000745

0.007770

(Small business loan volumei) / (Total loan volumei)

-0.026611

0.004986

(Total business loan volumei) / (Total loan volumei)

0.000923

0.587099

(Consumer loan volumei) / (Total loan volumei)

0.008051

0.052753

(Residential real estate volumei) / (Total loan volumei)

0.006539

0.046151

(Commercial real estate volumei) / (Total loan volumei)

0.007822

0.121623

Two-sided, normally distributed error term,

i ~ iid

0.002120

0.000836

σ _µ

(0,

μi

≥

σ_v

σν

Positive, half- normally distributed error term,

( 0 ) ~ iid

0.040618

0.000000

(0,σμ2), that gauges excess nonperformance

Table 7 (Loans with bad credit)

The data set includes 710 top-tier bank holding companies at the end of 2013. Nonperforming loans include past due loans,
	Best Practice Nonperformance and Lend						Inefficiency
gross charge -offs, and other real estate owned. Banks whose nonperforming loans exceed 15 percent or are less than 1 percent
of total loans are dropped from the initial sample. Stochastic frontier techniques are used to estimate the best-practice
minimum ratio of nonperforming loans to total loans for any given amount of total loans, controlling for the composition of the
bank’s loan portfolio, the average contractual interest rate charged on its loans, the ten-year average GDP growth rate and
market concentration in the states in which the bank operates, where each state’s datum is weighted by the proportion of the
bank’s total deposits located in the state. Lending Inefficiency is measured as the difference between the noise-adjusted actual
ratio and the best-practice minimum ratio of nonperforming loans to total loans.
		Panel A: Consolidated Assets < $1 Billion
Avg. Contractual Loan	N	Mean			Median	Std. Dev.		Minimum	Maximum

	364	0.0525			0.0511		0.0089	0.0352	0.1404
Interest Rate
(Nonperforming Loans)	364	0.0461			0.0370		0.0310	0.0100	0.1485
/ (Total Loans)
Nonperformance^{Best-Practice}	364	0.0123			0.0120		0.0045	0.0009	0.0529
Lending Inefficiency	364	0.0338	<		0.0237		0.0290	0.0007	0.1308

Avg. Contractual Loan	Panel B: $1 Billion			Consolidated Assets < $10 Billion
	Panel B: $1 Billion			Consolidated Assets < $10 Billion
	293	0.0494			0.0483		0.0086	0.0278	0.1329
	293	0.0494			0.0483		0.0086	0.0278	0.1329
Interest Rate
(Nonperforming Loans)	293	0.0383			0.0280		0.0277	0.0100	0.1492
/ (Total Loans)
Nonperformance^{Best-Practice}	293	0.0109			0.0104		0.0044	0.0017	0.0504
Lending Inefficiency	Panel	C: $10 Billion			Consolidated	Assets < $50 Billion
	293	0.0274	<		0.0190		0.0256	0.0011	0.1390

Avg. Contractual Loan	35	0.0458			0.0445		0.0067	0.0352	0.0643


Interest Rate
(Nonperforming_{/(TotalLoans)}Loans)	35	0.0312			0.0264		0.0219	0.0102	0.1379
Nonperformance^{Best-Practice}	35	0.0096			0.0096		0.0035	0.0020	0.0206
Lending Inefficiency	Panel	D: $50 Billion		Consolidated Assets < $250 Billion
	35	0.0217	<		0.0168		0.0196	0.0031	0.1174


Avg. Contractual Loan	11	0.0453			0.0401		0.0223	0.0253	0.1100
Avg. Contractual Loan	11	0.0453			0.0401		0.0223	0.0253	0.1100
Interest Rate
(Nonperforming Loans)	11	0.0307			0.0283		0.0099	0.0146	0.0449
/ (Total Loans)
Nonperformance^{Best-Practice}	11	0.0129			0.0110		0.0108	0.0017	0.0431
Lending Inefficiency		Panel E: Consolidated Assets >				$250 Billion
	11	0.0179			0.0172		0.0082	0.0019	0.0304



Avg. Contractual Loan	7	0.0538			0.0444		0.0196	0.0401	0.0921
Avg. Contractual Loan	7	0.0538			0.0444		0.0196	0.0401	0.0921
Interest Rate
(Nonperforming_{/(TotalLoans)}Loans)	7	0.0610			0.0569		0.0118	0.0483	0.0798
Nonperformance^{Best-Practice}	7	0.0424			0.0496		0.0162	0.0179	0.0589
Lending Inefficiency	7	0.0186			0.0240		0.0107	0.0040	0.0307

Table 8 (Loans with bad credit)

Best Practice Nonperformance and Lending Inefficiencyfor Banks with Consolidated Assets Greater Than $50 Billion (Loans with bad credit)

The values of lending inefficiency, the best-practice ratio of nonperforming loans measuring inherent credit risk, and the observed ratio of nonperforming loans are shown for large banks with consolidated assets greater than $50 billion. Nonperforming loans include past due loans, gross charge- offs, and other real estate owned. Stochastic frontier techniques are used to estimate the best-practice minimum ratio of nonperforming loans to total loans for any given amount of total loans, controlling for the composition of the bank’s loan portfolio, the average contractual interest rate charged on its loans, the ten-year average GDP growth rate and market concentration in the states in which the bank operates, where each state’s datum is weighted by the proportion of the bank’s total deposits located in the state.


		Name				Book Value of		Lending		Best-Practice		(Nonperforming
		of Bank Holding				Total Assets		Lending		Best-Practice		Loans) / (Total
		of Bank Holding				Total Assets		Inefficiency		Nonperformance		Loans) / (Total
		Company				(1000s)		Inefficiency		Nonperformance		Loans)
		Company				(1000s)						Loans)
		_CoJPMorgan_.		Chase &
	1	_CoJPMorgan_.		Chase &	2,415,689,000		0.0065		0.0496		0.0561
	2	_CorporationBankOfAmerica			2,104,995,000		0.0127		0.0589		0.0716
3		Citigroup Inc.			1,880,382,000		0.0040		0.0531		0.0569
	4	_CompanyWellsFargo &			1,527,015,000		0.0259		0.0538		0.0798
5		U.S. Bancorp			364,021,000		0.0262		0.0221		0.0483
	6	Services^PNCFinancialGroup,			320,596,232		0.0307		0.0178		0.0486
	7	Financial^CapitalOneCorp			297,282,098		0.0240		0.0414		0.0654
	8	Corporation^BB&T			183,009,992		0.0220		0.0166		0.0387
	9	_IncSuntrust_.	Banks,		175,380,779		0.0206		0.0126		0.0332
	10	_BancorpFifthThird			129,685,180		0.0161		0.0122		0.0283
	11	_CorporationRegionsFinancial			117,661,732		0.0304		0.0109		0.0414
	12	_CorporationNorthernTrust			102,947,333		0.0176		0.0017		0.0193
	13	Keycorp			92,991,716		0.0172		0.0073		0.0246
	14	Corporation^M&TBank			85,162,391		0.0300		0.0113		0.0413
	15	_ServicesDiscover Financial			79,339,664		0.0019		0.0444		0.0449
	16	Incorporated^Comerica			65,356,580		0.0093		0.0053		0.0146
	17	Bancshares^Huntington Inc			59,476,344		0.0169		0.0094		0.0263
	18	Bancorporation^Zions			56,031,127		0.0144		0.0102		0.0246

Table 9 (Loans with bad credit)

The data set includes 710 top-tier bank holding companies at the end of 2013. Stochastic frontier techniques

Factors Affe ting Lending Inefficiency

are used to estimate the best-practice minimum ratio of nonperforming loans to total loans specified in (1).

Lending Inefficiency, defined by (6), is measured as the difference between the observed noise-adjusted ratio

and the best-practice minimum ratio of nonperforming loans to total loans.

The regression is estimated with OLS, and standard errors are heteroscedasticity consistent. Parameter

estimates in

are significantly different from zero at stricter than 10%.

We apply the general-to-specific modeling strategy to identify the best specification for the lending inefficiency

bold

equation. The regressors of the general specification are given in the first column. The specific specification is

obtained through a full AIC-based search over all possible specifications with all combinations of regressors

derived from the general specification.

General Specification

Specific Specification

Parameter

Variable

Pr > |t|

Estimate

Intercept

-1.14127

0.2358

-0.06965

0.0003

log (Book Value Assets (1,000s))

0.13899

0.2577

[log (Book Value Assets (1,000s))]²

_––_0.023890.00442

0.2558

-0.02468

Total Loans/Assets

-0.13971

0.0024

-0.14365

0.0012

Equity Capital/Assets

41.97631

0.0036

2.73758

0.0023

Contractual Interest Rate on Loans

0.1661

<.0001

(Contractual Interest Rate on Loans)²

-330.04581

0.1610

_(Book[Contractual_{Value Assets}Interest_(1,000s))Rateon Loans] x [log

-5.07195

0.1917

-1.02265

_(Book[Contractual_{Value Assets}Interest_(1,000s))Rateon Loans] ² x [log

<.0001

0.16053

0.1942

[Contractual Interest Rate on Loans] x [log

41.64547

0.1700

(Book Value Assets (1,000s))]2

[Contractual Interest Rate on Loans]₂

²x [log

-1.35212

0.1652

(Book Value Assets (1,000s))]

0.06263

0.05884

Consumer Loans/Total Loans

0.0146

0.0163

Total Business Loans/ Total Loans

0.04302

0.0124

0.04096

0.0109

Small Business Loans/ Total Loans

-0.10344

0.0002

-0.09910

0.0001

Commercial RE Loans/ Total Loans

0.06081

<.0001

0.06109

<.0001

Residential RE Loans/ Total Loans

0.03825

0.0039

0.03818

0.0042

GDP Growth Rate

-0.00395

<.0001

-0.00405

<.0001

Number of States

0.00258

0.0203

0.00230

0.0013

[Number of States] x [Number of Branches]

-3.53221E-

0.0057

-3.31649E-7

0.0105

-0.01267

0.3103

Market Concentration

Table 10 (Loans with bad credit)

The data set Best consists Practice of 710 Non performance, top-tierbank holding Lending companies, Inefficiency, of which 244 and are publicly Financial traded, Performance at year-end 2013. The sample is partitioned into the halves with the lower and higher ratios of observed nonperforming loans to total loans. Mean values in bold print are significantly different at stricter than 10 percent.

Panel A: Half of the Sample with a Lower Ratio of Nonperforming Loans to Total Loans

_(1,000s)BookValue Assets			N	Mean	Median	Std. Dev.	Minimum	Maximum

			355	4,152,644	1,098,991	12,251,316	280,370	129,685,180
_Loans)(Nonperforming_/Loans			355	0.0208	0.0206	0.0061	0.0100	0.0325
Nonperformance^{Best-Practice}			355	0.0097	0.0095	0.0030	0.0009	0.0239
Lending Inefficiency			355	0.0112	0.0110	0.0060	0.0007	0.0262
_LoanAverage_InterestContractual_Rate			355	0.0476	0.0475	0.0059	0.0253	0.0750
Loans / Assets			355	0.6543	0.6735	0.1211	0.1892	0.9621
Tobin’s		Ratio	129	1.0564	1.0554	0.0475	0.9580	1.3129
	q

		Panel B: Half of the Sample a Higher Ratio of Nonperforming Loans to Total Loans
_(1,000s)BookValue Assets			355	29,089,290	864,288	214,309,437	92,694	2,415,689,000
_(1,000s)BookValue Assets			355	29,089,290	864,288	214,309,437	92,694	2,415,689,000
_Loans)(Nonperforming_/Loans			355	0.0634	0.0549	0.0278	0.0325	0.1492
Nonperformance^{Best-Practice}			355	0.0142	0.0131	0.0068	0.0022	0.0589
Lending Inefficiency			355	0.0491	0.0398	0.0269	0.0019	0.1390
_LoanAverage_InterestContractual_Rate			355	0.0540	0.0525	0.0110	0.0366	0.1404
Loans / Assets			355	0.6246	0.6431	0.1297	0.1653	0.9216
Tobin’s		Ratio	116	1.0360	1.0258	0.0621	0.9506	1.3058
	q

Table 11 (Loans with bad credit)

Comparison of Means between Groups Partitioned
by Nonperforming Loans Ratio and Lending Inefficiency
The data set consists of 710 top-tier bank holding companies, of which 244 are publicly traded, at year-end 2013. The
sample is partitioned into the halves with the lower and higher ratios of observed nonperforming_ploans to total loans. Each
of these halves is then partition into the more and less efficient halves by lending efficiency. The -value represents the
statistical significance of the comparison of means in the pairing. Pairs of means in		are statistically different at stricter
than = 0.10.	bold

Panel A

Panel B

Panel C

(Nonperforming Loans) /

Banks with Lower Ratios of

Banks with Higher Ratios of

(Total Loans)

Nonperforming Loans to

Nonperforming Loans to Total

Lower

Higher

Total Loans

Loans

Ratio

More

Less

More

Less

More

Efficient

Lending

n = 178

n = 177

n =178

n = 177

n = 178

n* = 62

n* = 67

n*=76

n*= 40

n*= 62

n*= 76

Assets^BookValue(1,000s)

Mean

3,153,030

5,157,905

0.12

56,597,667

1,425,498

0.02

3,153,030

56,597,667

0.02

_Loans)(Nonperforming_/Loans

0.0160

0.0256

<0.01

0.0432

0.0837

<0.01

0.0160

0.0432

<0.00

Nonperformance^{Best-Practice}

0.0101

0.0092

<0.01

0.0142

0.0141

0.89

0.0101

0.0142

<0.00

Inefficiency^Lending

0.0062

0.0163

<0.01

0.0289

0.0693

<0.01

0.0062

0.0289

<0.00

Average

0.0484

0.0467

<0.01

0.0533

0.0548

0.19

0.0484

0.0533

<0.00

Contractual Loan

Interest Rate

0.6575

0.6511

0.62

0.6367

0.6124

0.08

0.6575

0.6367

0.12

Loans / Assets

Tobin’s q Ratio*

1.0643

1.0490

0.07

1.0468

1.0155

0.01

1.0643

1.0468

0.05

Table 12 (Loans with bad credit)

Regression Analysis of Tobin’s q Ratio with the Nonperforming Loans Ratio (Loans with bad credit)

The data set includes 244 publicly traded top-tier bank holding companies at the end of 2013. The dependent

as One of the Main Fac o s for the Construction of Regressors

variable is Tobin’s ratio. Regressions are estimated with OLS, and standard errors are heteroscedasticity

consistent. Probability values are reported in parentheses under the parameter estimates. Parameter

estimates in

bold

are significantly different from zero at stricter than 10%.

We apply the general-to-specific modeling strategy to identify the best specification for the q ratio equation.

The regressors of the general specification are given in the first column. The specific specification is obtained

through a full AIC-based search over all possible specifications with all combinations of regressors derived

from the general specification.

Dependent Variable: Tobin’s q Ratio

General Specification

Specific Specification

Variable

Parameter Estimate

Pr > |t|

Parameter Estimate

Pr > |t|

Intercept

−0.4160

0.049

−0.4616

0.027

_(1,000s)ln(Book ₎Value Assets

0.1972

<.0001

0.1933

<.0001

[ln (Book Value Assets

−0.0060

−0.0059

<.0001

(1,000s))]

Loans/ Assets

−0.2794

0.196

−0.0587

0.045

(Loans/ Assets)2

0.1756

0.288

_LoansNonperforming Loans/

−3.1918

0.125

−2.8247

0.088

(Nonperforming Loans/

1.3889

0.775

Loans)

(Nonperforming Loans/

0.1705

0.180

0.1553

0.177

Loans)

[

(Book Value

Assets (1,000s))]

Adj₌₀.R_.339Sq

F=18.8

Adj₌₀.R_.354Sq

F=26.1

Table 13 (Loans with bad credit)

Regression Analysis of Tobin’s q Ratio with the Nonperforming Loans Ratio and the Mega Bank Dummy as Two of the Main Factors for the Construction of Regressors The data set includes 244 publicly traded top-tier bank holding companies at the end of 2013. The dependent variable is Tobin’s ratio. Regressions are estimated with OLS, and standard errors are heteroscedasticity consistent. Probabilityq values are reported in parentheses under the parameter estimates. Parameter estimates in bold are significantly different from zero at stricter than 10%.

The regressors of the general specification are given in the first column. The specific specification is obtained through a full AIC -based search over all possible specifications with all combinations of regressors derived from the general specification.

Mega Bank is the indicator variable of banks with assets greater than $170 billion.

Dependent Variable: Tobin’s q Ratio

General Specification

Specific Specification

Variable

Parameter Estimate

Pr > |t|

Parameter Estimate

Pr > |t|

Intercept

0.0002

<.0001

0.3425

0.3327

−1.5560

−1.5617

(Book Value Assets (1,000s))

<.0001

[ln (Book Value Assets (1,000s))]²

−0.0106

<.0001

−0.0102

<.0001

Loans / Assets

−0.2678

0.2000

−0.0556

0.0502

(Loans / Assets)²

0.1685

0.2888

Nonperforming Loans / Loans

−0.6926

0.6281

(Nonperforming Loans / Loans)²

1.3889

0.5556

(Nonperforming Loans / Loans)

−0.0391

[

(Book Value Assets

−0.0064

0.9447

<.0001

(1,000s))]

Mega Bank (=1 when book value

−0.0725

0.5551

assets > $170 billion)

(Mega Bank)

(Nonperforming

−11.1061

0.1622

−14.3103

0.0018

Loans / Loans)

(Mega Bank)

[ (Nonperforming

−33.1282

0.4794

Loans / Loans)

]

Loans^(Mega /^Bank)Loans

(Nonperforming_[(BookValue

0.8417

0.0108

0.8325

0.0004

Assets (1,000s))]

× ln

Adj₌₀.R_.369Sq

F=13.9

Adj₌₀.R_.377Sq

F=25.5

Table 14 (Loans with bad credit)

Regression Analysis of Tobin’s q Ratio with the Two Components of the Noise-Adjusted Nonperforming Loan Ratio as Two of the Main Factors for the Construction of Regressors The data set includes 244 publicly traded top-tier bank holding companies at the end of 2013. The dependent

variable is Tobin’s ratio. Regressions are estimated with OLS, and standard errors are heteroscedasticity consistent. Parameterq estimates in bold are significantly different from zero at stricter than 10%.

General Specification

Specific Specification

Variable

Parameter Estimate

Pr > |t|

Parameter Estimate

Pr > |t|

Intercept

−

1.0429

0.0021

−

1.0799

0.0004

<.0001

0.0086

0.0085

(1,000s))

(Book Value Assets

0.2777

0.2732

(1,000s))]^[(BookValue2

Assets

−

<.0001

−

0.0523

<.0001

Loans / Assets

−0.2445

0.2601

−

0.0659

(Loans / Assets)²

0.1517

0.1665

Nonperformance^{Best-Practice}

−8.3203

0.0191

−6.8334

0.0041

Nonperformance)^{(Best-Practice}

−17.8215

0.6382

Best-Practice

0.5715

0.4566

Nonperformance)

[

(Book Value Assets

0.0160

0.0012

(1,000s))]

0.6188

Lending Inefficiency

−0.6801

0.7248

−

<.0001

(Lending Inefficiency)

0.8625

0.8645

Lending Inefficiency)

(1,000s)) ]

−0.0149

0.9033

[ (Book Value Assets

_(BestLending_-PracticeInefficiency)

14.3816

0.5994

Nonperformance)

Adj₌₀.R_.352Sq

F=13.0

Adj₌₀.R_.362Sq

F=24.0

Table 15: Panel A (Loans with bad credit)

Significance Tests of Derivatives of Tobin’s q Ratio with Respect

to Each of the Two Components of the Noise-Adjusted Nonperforming Loan Ratio

In the specific specification of Table 14, the derivatives of Tobin’s q ratio with respect to the best-practice

nonperforming loan ratio and the lending inefficiency are given by

∂(Tobin’s q ratio)/∂(best-practice nonperforming loans ratio) = −6.8334 + (0.4566) × [ln (Book Value

Assets (1,000s))].

∂(Tobin’s q ratio)/∂(lending inefficiency) = −0.6188.

Derivative of Tobin’s q ratio with respect to . . .

> 0

> 0 and

< 0

< 0 and

statistically

significant

best-practice nonperforming loans ratio

111

133

lending inefficiency

244

244

Table 15: Panel B (Loans with bad credit)

Estimates of Derivatives of Tobin’s q Ratio with Respect to Inherent Credit Risk,

Individually for the Largest 50 Banks (Loans with bad credit)

The derivative of Tobin’s				ratio with respect to the best-practice nonperforming loans ratio is derived
from the specific specification of					:
			q							(Book Value
∂(Tobin’s ratio)/∂(best-practice nonperforming loans ratio) = −6.8334 + (0.4566) [
Assets (1,000s))].				bold	Table 14
									× ln
	q
Estimates of derivatives in					are significantly different from zero at stricter than 10%.

									∂(Tobin’s q
							Book-Value Assets		Ratio)/
		Name							∂(best-practice	p-value
							(1000s)
									nonperforming

									loans ratio)
1								3.032
2	Jpmorgan Chase & Co.					2,415,689,000		2.969		0.0012
3	Bank Of America Corporation					2,104,995,000		2.918		0.0012
4	Citigroup Inc.					1,880,382,000		2.823		0.0012
5	Wells Fargo & Company					1,527,015,000		2.168		0.0013
6	U.S. Bancorp					364,021,000		2.110		0.0022
7	Pnc Financial Services Group					320,596,232		2.075		0.0024
8	Capital One Financial Corp					297,282,098		1.854		0.0025
9	Bb&T Corporation					183,009,992		1.834		0.0035
10	Suntrust Banks, Inc.					175,380,779		1.697		0.0036
11	Fifth Third Bancorp					129,685,180		1.652		0.0048
12	Regions Financial Corp					117,661,732		1.591		0.0053
13	Northern Trust Corporation					102,947,333		1.545		0.0061
14	Keycorp					92,991,716		1.505		0.0068
15	M&T Bank Corporation					85,162,391		1.384		0.0076
16	Comerica Incorporated					65,356,580		1.341		0.0107
17	Huntington Bancshares Inc					59,476,344		1.313		0.0121
18	Zions Bancorporation					56,031,127		1.234		0.0132
19	Cit Group Inc.					47,138,960		1.230		0.0171
20	N Y Community Bancorp					46,688,287		1.132		0.0174
21	First Niagara Financial Group					37,643,867		1.108		0.0245
22	Popular, Inc.					35,749,000		1.024		0.0266
23	City National Corporation					29,717,951		0.980		0.0365
24	Bok Financial Corporation					27,021,529		0.970		0.0431
25	Svb Financial Group					26,417,306		0.966		0.0448
	Synovus Financial Corp.					26,201,604				0.0455

Table 15: Panel B (Continued) (Loans with bad credit)

Estimates of Derivatives of Tobin’s q Ratio with Respect to Inherent Credit Risk, Individually for the Largest 50 Banks

					∂(Tobin’s q
			Book-Value Assets		Ratio)/
	Name				∂( best-practice		p-value
			(1000s)
					nonperforming

					loans ratio)
26	East West Bancorp, Inc.	24,730,609		0.940		0.0504
27				0.934
28	Cullen/Frost Bankers, Inc.	24,388,272		0.931		0.0516
29	Associated Banc-Corp	24,226,920		0.925		0.0523
30	Firstmerit Corporation	23,912,451		0.922		0.0535
31	First Horizon National Corp	23,791,187		0.908		0.0540
32	Commerce Bancshares, Inc.	23,081,892		0.870		0.0570
33	First Citizens Bancshares, Inc	21,199,091		0.862		0.0664
34	Webster Financial Corp	20,856,659		0.820		0.0684
35	Hancock Holding Company	19,025,806		0.807		0.0807
36	Susquehanna Bancshares	18,473,488		0.805		0.0852
37	Tcf Financial Corporation	18,402,494		0.797		0.0858
38	Wintrust Financial Corp	18,097,783		0.766		0.0884
39	Umb Financial Corporation	16,911,852		. 66		0.0999
40	Fulton Financial Corporation	16,908,633				0.0999
41	Valley National Bancorp	16,156,541		0.746		0.1084
42	Bank Of Hawaii Corporation	14,127,598		0.684		0.1377
43	Privatebancorp, Inc.	14,085,746		0.683		0.1385
44	F.N.B. Corporation	13,563,405		0.666		0.1480
45	Iberiabank Corporation	13,365,550		0.659		0.1518
46	Bancorpsouth, Inc.	13,045,442		0.648		0.1584
47	International Bancshares	12,079,477		0.613		0.1809
48	Trustmark Corporation	11,790,383		0.602		0.1886
49	Texas Capital Bancshares, Inc	11,714,698		0.599		0.1907
50	Umpqua Holdings Corp	11,641,151		0.596		0.1927
	Cathay General Bancorp	10,989,286		0.570		0.2124

Loans with bad credit: How Bad Is a Bad Loan? Distinguishing Inherent Credit Risk from Inefficient Lending (Does the Capital Market Price This Difference?)

Introduction – Loans with bad credit

The Data (Loans with bad credit)

Best-Practice Loan Performance and the Efficiency of Lending

3.1 The Estimation of the Best-Practice Loan Performance Frontier – Best Practice Loans with bad credit

3.2 Credit Risk and Lending Inefficiency at Community Banks ( Loans with bad credit )

3.3 Credit Risk and Lending Efficiency at assests the Largest Banks

3.4 Factors that Influence Lending Inefficiency (Loans with bad credit)

Do Capital Markets Distinguish between Inherent Credit Risk and Lending Efficiency

4.1 Relationship of Financial Performcapitalnce to Nonppricesrforming Loans (Loans with bad credit)

4.2 Relationship of Financial Performance to Inherent Credit Risk and Lending Inefficiency

Summary and Conclusions

4.1 Relationship of Financial Perform_capitalnce to Nonp_pricesrforming Loans (Loans with bad credit)