Credit Risk: Contract Characteristics for Success

Transcription

1 Credit Risk: Characteristics for Success By James P. Murtagh, PhD Equipment leasing companies need reliable information to assess the default risk on lease contracts. Lenders have historically built independent systems and expertise to support their credit-granting decisions. This approach created an incentive to specialize in order to leverage the value of these systems. Through the support of the Equipment Leasing and Finance Foundation (the Foundation) and PayNet, this paper investigates specific characteristics of successful and failed equipment leases for 12 recent quarters. Lenders designated failed leases as such in the PayNet database, or reported nontrivial loss amounts for these contracts. The findings indicate that contract characteristics, industry factors, and economic factors all influence lease default rates. The results support the existence of risk premiums for particular industries and asset types. This type of analysis could also be used to create more effective risk management models as proposed under the Basel II accords. BACKGROUND In his article titled The Future of Commercial Credit Data, Thomas Ware argues that the best predictor of default risk in leasing contracts is appropriate, comparable historical data. Although this type of information is readily available in the consumer lending industry, commercial lenders have traditionally developed their own systems, based on specialization and expertise in their particular industry, equipment, or borrower niches. The PayNet database was established in 2001 to provide historical data in the equipment leasing market. The availability and successful use of industrywide historical information can provide tangible benefits for lenders. Clearly, better information should lead to better credit decisions and lower default rates. A better understanding of credit risk can also provide rational support for credit risk premiums based on company, asset, or contract characteristics. An industrywide database also provides the basis for lower costs through process automation and lower risks through diversification opportunities. These effects should improve lender cash flows and profitability. The availability of a broad industry database also provides researchers with a unique and valuable insight into the nature of successful and failed equipment lease transactions. The purpose of this research study is to investigate differences in the characteristics of successful (current) lease transactions and failed transactions based on a sample of the PayNet data. Failed contracts were defined as a contract for which the lender has reported the status as bad or reported a nontrivial loss amount. Identifying measurable and persistent differences in these characteristics expands the industry's understanding of critical factors used in making lease financing decisions and monitoring existing contracts. The study also investigates differences in characteristics between various types of lease customers and contracts to identify the most important factors in granting lease financing. This comparative information is useful to firms improving internal credit risk models, as required under the Basel II Accords. The eventual impact of the Basel II Accord (2006) on the leasing industry remains unclear. One apparent certainty is that Basel II will change the way a significant portion of the leasing industry measures risk and assesses risk exposures. In a 2003 white paper, The Basel II Accord: What Does It Mean for the North American Leasing Market? the Foundation The complete version of this study may be found in the Foundation library at

2 Understanding default risk factors can assist with risk management through diversification. To build a diversified portfolio, however, lenders must learn and understand risk factors in unrelated industries. identified direct and indirect effects of the accord. 1 The white paper is a call to action for leasing company managers. The accord will affect leasing companies in North America either directly through participation, or indirectly, as the consequences of Basel II are felt globally, the white paper states. The primary focus of the accord is the global banking industry. By extension, lease subsidiaries of banks adopting the accords will be directly affected. Also, the European Union has decided to adopt Basel II for all credit institutions operating in the EU. Clearly, lease companies operating in Europe also will be directly affected by the accords. In their study of the residential subprime lending market, Cowan and Cowan find that default correlation is important in understanding risk management in these risky portfolios. Default correlation incorporates the fact that systemic events cause defaults to cluster. Coincident defaults among borrowers may be triggered by common underlying macroeconomic or industry factors. For example, it is reasonable to expect that increases in fuel costs coupled with an economic slowdown would squeeze cash flows in transportation industries, a result seen in the following study. Understanding default risk factors can assist with risk management through diversification. To build a diversified portfolio, however, lenders must learn and understand risk factors in unrelated industries in an effort to reduce the correlations between default factors. DATA AND METHODOLOGY Through an agreement between the Foundation and PayNet, this study provides descriptive statistics on a portion of the PayNet database. 2 The study investigates the differences in the contract details and financial conditions of failed and successful (not failed) equipment leases. The contract characteristics studied include the original contract receivable amount, contract payment amount, contract payment count, loss amount, balance amount, number of times past due at various aging categories, amounts past due at these aging categories, and the average number of days past due. The analysis controls for industry classification, contract term, contract type, asset type, and type of contract guarantee. This detailed investigation provides useful information to help Table 1 GOOD AND FAILED CONTRACTS BY QUARTER January 2004 Snapshot Date Good s Failed s Total % Failed April, ,854 5,162 62, % July, ,917 6,077 68, % October, ,530 6,803 72, % January, ,674 7,713 76, % April, ,523 8,574 79, % July, ,796 9,374 83, % October, ,510 9,876 86, % January, ,266 10,859 91, % April, ,649 11,532 94, % July, ,280 12, , % October, ,185 12, , % January, ,500 13, , % Average 75,974 9,513 85, % 2 J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O. 1

3 lenders make better lending decisions and build better credit risk models. Data Description The subset of the PayNet database analyzed consists of 48 fields describing between 62,000 and 107,000 separate contracts. Data snapshots were analyzed for 12 quarters beginning in April 2001 and ending in January The numbers of successful and failed contracts in each quarterly dataset are shown in table 1 (page 2). In simplest terms, a successful or good contract is one that is not in default in the quarter studied. For the purposes of this study, a failed contract is one in which the lender has reported the status as bad or reported a loss amount over $100. s meeting these criteria were flagged with a yes indication in the contract bad status field in the quarter studied. Data fields include leasing company information and contract-level information, supporting analysis at both levels. Given the size of the each quarterly data file, the SAS System for Windows, Release 8 was used for all data analysis. The data files studied represent a random subset of the complete PayNet database for any given quarter and were provided by the company in raw form, without refinement. Descriptive Statistics (Table 1 through Table 6b) Table 1 shows the number of good and failed contracts in the quarterly data files studied. The total number of contracts included in the data files nearly doubles from 62,016 in April 2001 to 107,645 in January The number of failed contracts included in the data files nearly triples in that same time period, rising from 5,162 in April 2001 to 13,145 in January The percentage of failed contracts in a given quarter increases steadily from 8.3% in April 2001 to 12.2% in April 2003 and remains at that level for the remaining periods studied. The PayNet data categorizes each lender by a four-digit SIC (standard industrial classification) code. To more efficiently categorize the data, these Figure 1 LEASE CONTRACTS BY INDUSTRY CLASSIFICATION January 2004 Figure 2 PERCENTAGE OF CONTRACTS BY ORIGINAL CONTRACT TERM January 2004 four-digit SICs were grouped into the 10 commonly used industry classifications (fig. 1). Table 2a (next page) shows the number of good and failed contracts by industry classification and quarter. The data shows an increase in the percentage of failed leases over time in each of the SIC groups studied. The greatest proportion of failed leases for January 2004 are found in the transportation/ public utilities group (17.6%), services group Another important characteristic of the lending contract is the original contract term. Longer contract terms reduce lender liquidity and provide a longer time horizon during which the contract may fail. J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O. 1 3

4 Table 2a GOOD AND FAILED CONTRACTS By Industry Classification and Quarter Table 2a Number of good and failed contracts by industry classification and quarter SIC Status Agriculture, Forestry, Fishing Good 3,890 3,922 4,007 4,073 4,134 4,213 4,291 4,355 4,409 4,479 4,660 4,698 Failed Mining Good Failed Construction Good 6,891 8,347 8,641 9,065 9,281 9,566 9,916 10,206 10,339 10,667 10,971 11,221 Failed ,028 1,098 1,167 1,262 1,321 1,355 1,383 Manufacturing Good 9,920 11,299 11,703 12,051 12,383 12,930 13,503 13,968 14,179 14,874 15,161 15,661 Failed 990 1,208 1,297 1,440 1,577 1,691 1,887 1,972 2,053 2,207 2,279 2,328 Transportation, Public Utilities Good 8,313 8,981 9,283 9,511 9,451 10,148 10,430 10,624 10,802 11,052 11,416 11,877 Failed ,065 1,230 1,510 1,763 1,883 2,204 2,353 2,459 2,486 2,535 Wholesale Trade Good 4,943 5,248 5,391 5,500 5,589 5,736 5,816 6,108 6,292 6,685 7,038 7,305 Failed Retail Trade Good 2,413 2,615 2,676 3,816 4,135 4,185 4,250 4,472 4,519 4,687 4,799 4,863 Failed Finance, Insurance, Real Estate Good 4,458 4,799 5,066 5,276 5,438 5,822 6,243 6,602 6,832 7,360 7,775 8,103 Failed Services Good 13,929 15,361 16,209 16,631 17,182 18,123 18,886 20,460 21,568 23,389 24,162 25,076 Failed 1,424 1,682 1,957 2,325 2,533 2,696 2,776 3,086 3,309 3,594 3,706 3,895 Public Administration Good 1,587 1,745 1,921 2,092 2,244 2,350 2,401 2,656 2,870 4,226 4,337 4,812 Failed Table 2b PERCENTAGE OF FAILED CONTRACTS By Industry Classification and Quarter Agriculture, Forestry, Fishing 7.4% 7.7% 8.0% 8.2% 8.4% 8.3% 8.4% 8.6% 8.9% 8.9% 9.2% 9.2% Mining 5.2% 4.8% 5.2% 5.3% 5.5% 5.9% 7.2% 7.3% 8.2% 8.8% 8.2% 8.3% Construction 8.5% 8.2% 8.6% 9.0% 9.5% 9.7% 10.0% 10.3% 10.9% 11.0% 11.0% 11.0% Manufacturing 9.1% 9.7% 10.0% 10.7% 11.3% 11.6% 12.3% 12.4% 12.6% 12.9% 13.1% 12.9% Transportation, Public Utilities 8.2% 9.3% 10.3% 11.5% 13.8% 14.8% 15.3% 17.2% 17.9% 18.2% 17.9% 17.6% Wholesale Trade 6.0% 6.3% 7.0% 7.7% 8.2% 8.7% 8.4% 8.5% 8.6% 8.4% 8.9% 8.9% Retail Trade 8.8% 8.9% 9.7% 7.4% 7.1% 8.8% 9.1% 9.6% 9.8% 9.8% 9.9% 10.0% Finance, Insurance, Real Estate 7.4% 7.9% 8.1% 8.2% 8.5% 8.8% 7.9% 7.5% 7.7% 7.6% 7.8% 7.9% Services 9.3% 9.9% 10.8% 12.3% 12.8% 12.9% 12.8% 13.1% 13.3% 13.3% 13.3% 13.4% Public Administration 7.0% 7.3% 7.2% 8.2% 9.6% 10.0% 10.6% 12.3% 12.2% 10.2% 10.0% 9.3% Table 3a GOOD AND FAILED CONTRACTS By Original Term and Quarter Term Status months Good 1,424 1,635 1,691 1,858 1,916 1,986 2,066 2,216 2,332 2,478 2,619 2, months Failed months Good 3,322 3,798 3,995 4,394 4,539 4,787 4,917 5,182 5,317 5,526 5,703 5, months Failed months Good 17,187 19,062 19,997 20,764 21,402 22,392 23,146 24,380 25,205 26,944 27,890 29, months Failed 1,452 1,682 1,855 2,044 2,249 2,431 2,547 2,814 2,985 3,239 3,332 3, months Good 12,035 13,309 13,868 14,649 15,180 15,931 16,514 17,148 17,693 19,754 20,468 21, months Failed 1,113 1,287 1,452 1,605 1,776 1,957 2,073 2,285 2,394 2,560 2,631 2, months Good 18,275 19,943 20,647 21,353 21,794 22,819 23,747 24,913 25,588 26,629 27,424 28, months Failed 1,711 2,067 2,347 2,771 3,089 3,382 3,534 3,860 4,142 4,328 4,486 4, months Good 4,611 5,170 5,332 5,656 5,692 5,881 6,120 6,427 6,514 6,949 7,081 7, months Failed ,059 1,199 1,266 1,370 1,422 1,463 Table 3b PERCENTAGE OF FAILED CONTRACTS By Original Term and Quarter Term months 6.6% 6.5% 7.7% 8.1% 8.5% 8.7% 8.9% 8.7% 9.1% 9.0% 8.7% 8.8% months 7.7% 7.8% 8.0% 7.9% 8.3% 8.4% 8.6% 8.6% 8.8% 9.4% 9.6% 9.6% months 7.8% 8.1% 8.5% 9.0% 9.5% 9.8% 9.9% 10.3% 10.6% 10.7% 10.7% 10.6% months 8.5% 8.8% 9.5% 9.9% 10.5% 10.9% 11.2% 11.8% 11.9% 11.5% 11.4% 11.5% months 8.6% 9.4% 10.2% 11.5% 12.4% 12.9% 13.0% 13.4% 13.9% 14.0% 14.1% 14.0% 61+ months 9.9% 10.5% 11.0% 11.7% 13.3% 14.2% 14.8% 15.7% 16.3% 16.5% 16.7% 16.8% 4 J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O. 1

5 (13.4%), and manufacturing (12.9%). The failure percentage for the transportation group was more than double that in the financial or mining groups. This significantly greater failure rate in a certain industry supports the idea that industrylevel risk factors are relevant to internally developed risk-management models. Greater industry risk may lead to higher risk premiums for lenders or borrowers in these industries and place greater emphasis on effective risk models for lenders in this industry. Another important characteristic of the lending contract is the original contract term. Longer contract terms reduce lender liquidity and provide a longer time horizon during which the contract may fail. Table 3 (page 4) shows the number and percentage of good and failed contracts categorized by original contract term and quarter. The majority of the contracts studied had an original contract term between 25 and 60 months (fig. 2, page 3). The percentage of defaults increases with respect to the original contract term. Nearly twice as many 61-plus month contracts (16.8%) failed as 0- to 12-month contracts (8.8%). The PayNet database also categorizes contracts into a number of contract types. The majority of contracts are true leases (68.4%). Table 4 (next page) shows the number and percentage of good and failed contracts by contract type and quarter. Default rates increased in all categories over time, with the largest increase on a percentage basis occurring in conditional sales and loans. In the most recent period, the highest default rates are in lease purchase (14.4%) and loans (13.9%), while the lowest default rate was in the revolver accounts (6.5%). An important characteristic of the lease contract is the underlying asset of the contract. The PayNet data that was studied lists 22 separate contract asset codes ranging from agriculture to waste and refuse handling. The number and percentage of good and failed contracts by contract asset code and quarter are shown in table 5 (next page). Over 70% of the contracts are found in only four categories: copiers (36.5%), trucks (12.3%), computer equipment (11.8%), and construction/ mining (10.6%). Default rates also vary widely in the most recent period. Marine craft, medium/light duty trucks, and waste and refuse handling show defaults rates less than 7%, while other asset codes are much higher in this sample. This considerable variation in default rates supports the use of higher risk premiums for certain asset classes and emphasizes the need for asset-specific factors in risk-management models. Furthermore, an analysis of the default rates over time (table 5b) shows that default correlations are not consistent across asset types. Recognizing the variation in default correlations provides managers with an opportunity to reduce the overall portfolio risk through diversification. The PayNet data also provides an indication of the manner by which contract cash flows are guaranteed. The contract guarantor code results are shown in table 6 (page 7). Under the PayNet system, guarantor codes include corporate, personal, both, yes-undefined, no, and unknown. More than three-quarters of the contacts studied have a guarantor code of no or unknown. For these guarantor codes, contract default rates remain fairly consistent across the 12 quarters studied, providing very little additional insight into the development of effective risk management models. Differences Between the Means Analysis, January 2004 (Tables 7-12) To more clearly understand the characteristics of successful and failed leases in the sample, descriptive statistics were calculated for subsets of the January 2004 data file (tables 7-10). The factors studied include the original contract receivable amount, contract payment amount, payment count, loss amount, and balance amount. The number of times the account was reported past due and the past due balances were studied for accounts 31 to 60 days past due, 61 to 90 days past due, and 90-plus days past due. The average number of days past due was also studied. Using these statistics, differences between The considerable variation in default rates supports the use of higher risk premiums for certain asset classes and emphasizes the need for asset-specific factors in riskmanagement models. J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O. 1 5

6 Table 4a GOOD AND FAILED CONTRACTS By Type and Quarter Type Status Conditional Sale Good 11,179 11,555 11,828 12,814 13,158 13,775 14,091 14,903 15,275 15,688 16,100 16,654 Failed ,138 1,303 1,559 1,594 1,790 1,896 2,023 2,093 2,160 Leveraged Lease Good Failed Loan Good 3,016 3,403 3,667 4,094 4,195 4,920 5,308 5,416 5,444 5,596 5,995 6,328 Failed ,002 1,022 Lease Purchase Good 4,282 4,338 4,393 4,469 4,520 4,557 4,603 4,708 4,768 4,842 4,885 4,936 Failed Revolver Good Failed Rental Lease Good 1,049 1,093 1,110 1,268 1,294 1,322 1,334 1,363 1,382 1,414 1,486 1,534 Failed True Lease Good 37,303 42,484 44,487 45,980 47,305 49,169 51,118 53,789 55,637 60,511 62,418 64,685 Failed 3,393 4,121 4,587 5,244 5,849 6,269 6,699 7,351 7,775 8,358 8,619 8,936 Table 4b PERCENTAGE OF FAILED CONTRACTS By Type and Quarter Type Conditional Sale 6.3% 6.8% 7.7% 8.2% 9.0% 10.2% 10.2% 10.7% 11.0% 11.4% 11.5% 11.5% Leveraged Lease 71.4% 71.4% 71.4% 71.4% 71.4% 71.4% 42.9% 71.4% 71.4% 71.4% 71.4% 71.4% Loan 8.4% 9.0% 10.1% 10.3% 11.4% 11.8% 11.2% 12.5% 14.0% 14.4% 14.3% 13.9% Lease Purchase 13.0% 13.4% 13.7% 14.2% 14.3% 14.3% 14.5% 14.6% 14.5% 14.6% 14.5% 14.4% Revolver 4.2% 2.3% 4.4% 7.8% 7.5% 7.3% 8.8% 5.6% 9.7% 7.8% 7.2% 6.5% Rental Lease 7.9% 8.5% 9.2% 8.4% 8.5% 8.3% 8.4% 9.0% 9.3% 9.3% 9.4% 9.7% True Lease 8.3% 8.8% 9.3% 10.2% 11.0% 11.3% 11.6% 12.0% 12.3% 12.1% 12.1% 12.1% 6 Table 5a GOOD AND FAILED CONTRACTS By Asset Code and Quarter Table 5a Number of good and failed contracts by contract asset code and quarter Asset Code Status Agricultual Good 2,704 2,958 3,022 3,075 3,122 3,159 3,206 3,266 3,291 3,330 3,362 3,378 Failed Aircraft Good Failed Marine craft Good Failed Buses & Motor coaches Good Failed Construction & Mining Good 6,932 7,934 8,187 8,665 8,821 9,012 9,294 9,526 9,600 9,790 9,934 10,129 Failed ,019 1,049 1,121 1,148 1,174 1,242 Computer Good 5,463 6,078 6,314 6,512 6,770 7,250 7,657 8,167 8,609 9,945 10,582 11,213 Failed ,033 1,078 1,215 1,275 1,377 1,410 1,447 Copier & fax Good 19,900 21,459 22,815 23,691 24,579 25,802 26,785 28,542 29,817 32,065 33,035 34,419 Failed 1,772 2,114 2,431 2,849 3,148 3,352 3,586 4,019 4,279 4,529 4,690 4,866 Forklift Good ,028 1,167 1,343 1,400 1,413 1,426 1,489 Failed Logging & Forestry Good Failed Medium/Light duty truck Good 2,456 2,784 2,862 2,920 2,963 3,027 3,099 3,147 3,192 3,242 3,265 3,296 Failed Medical Good Failed Manufacturing Good 3,446 3,895 3,963 4,010 4,055 4,133 4,271 4,400 4,456 4,558 4,681 4,792 Failed Office Equipment Good 1,107 1,163 1,210 1,237 1,281 1,575 1,608 1,721 1,780 2,478 2,630 2,791 Failed Printing & Photographic Good Failed Railroad Good Failed Retail Good 1,047 1,274 1,318 2,428 2,662 2,713 2,790 2,940 3,037 3,146 3,282 3,413 Failed Telecommunications Good 1,278 1,631 1,705 1,711 1,730 1,753 1,785 1,916 1,951 1,995 2,075 2,117 Failed Truck Good 8,405 8,773 8,916 9,042 9,078 9,791 10,044 10,243 10,322 10,492 10,903 11,233 Failed ,040 1,190 1,405 1,498 1,659 1,796 1,913 1,978 2,022 Unknown Good 1,248 1,533 1,584 1,616 1,638 1,660 1,718 1,792 1,824 2,329 2,409 2,496 Failed Vending & Restaurant Good ,019 1,052 Failed Waste & Refuse Handling Good Failed J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O. 1

7 Table 5b PERCENTAGE OF FAILED CONTRACTS By Asset Code and Quarter Table 5b Percentage of failed contracts by contract asset code and quarter Asset Code Agricultural 7.4% 7.3% 7.5% 7.8% 8.1% 8.5% 8.7% 8.9% 9.0% 9.1% 9.1% 9.2% Aircraft 7.5% 6.2% 8.1% 7.6% 9.3% 9.0% 7.8% 7.7% 7.7% 7.6% 7.6% 7.5% Marine craft 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 5.3% Copier Buses & Motor fax coaches 9.3% 8.2% 8.6% 9.0% 7.8% 9.6% 10.7% 9.5% 13.7% 11.4% 14.8% 11.5% 14.9% 11.8% 24.0% 12.3% 26.0% 12.5% 25.7% 12.4% 28.2% 12.4% 28.1% 12.4% Forklift Construction & Mining 9.1% 7.4% 8.7% 6.5% 9.1% 6.6% 9.1% 7.4% 9.5% 8.2% 9.5% 8.4% 9.9% 8.5% 9.9% 7.6% 10.5% 8.1% 10.5% 11.9% 10.6% 11.8% 10.9% 11.4% Logging Computer & Forestry 8.6% 4.6% 9.0% 5.2% 9.5% 5.0% 10.8% 5.3% 12.3% 6.3% 12.5% 6.7% 12.3% 7.9% 13.0% 7.7% 12.9% 7.7% 12.2% 8.0% 11.8% 7.9% 11.4% 7.7% Medium/Light duty truck 2.8% 3.1% 3.2% 3.8% 4.3% 4.6% 4.7% 5.2% 5.5% 5.7% 5.8% 5.8% Medical 12.0% 11.5% 12.4% 12.3% 12.6% 12.2% 11.6% 11.6% 12.7% 12.3% 13.1% 13.1% Manufacturing 12.4% 12.8% 13.1% 13.9% 14.7% 15.0% 15.5% 15.8% 15.8% 15.6% 15.5% 15.2% Office Equipment 7.4% 8.1% 8.7% 9.6% 10.0% 9.0% 9.8% 10.6% 10.6% 10.6% 10.3% 10.6% Printing & Photographic 14.0% 14.3% 14.8% 16.9% 17.6% 17.5% 17.4% 16.2% 16.8% 17.0% 17.1% 16.4% Railroad 13.0% 10.0% 10.0% 9.4% 9.1% 9.1% 2.9% 8.1% 10.8% 25.6% 25.6% 25.6% Retail 9.7% 8.6% 9.0% 5.5% 5.2% 6.7% 5.8% 6.1% 6.6% 7.1% 7.1% 7.1% Telecommunications 8.9% 10.5% 11.2% 12.8% 13.4% 14.5% 14.6% 14.3% 14.5% 14.3% 14.4% 14.4% Truck 7.2% 8.2% 9.4% 10.3% 11.6% 12.5% 13.0% 13.9% 14.8% 15.4% 15.4% 15.3% Unknown 13.3% 13.0% 13.3% 13.7% 14.9% 15.6% 16.0% 16.1% 16.3% 15.0% 15.1% 15.0% Vending & Restaurant 4.8% 6.3% 7.0% 7.6% 8.7% 17.1% 10.7% 10.7% 11.0% 11.9% 12.4% 12.7% Waste & Refuse Handling 3.0% 3.3% 3.2% 3.7% 3.9% 3.8% 3.4% 3.5% 4.0% 3.9% 4.1% 3.8% Table 6a GOOD AND FAILED CONTRACTS By Guarantor Code and Quarter Guarantor Code Status Both Good Failed Corporate Good Failed None Good 43,056 47,438 49,612 52,193 53,847 56,463 58,639 62,046 64,120 68,977 71,366 74,050 Failed 3,857 4,529 5,057 5,741 6,400 6,955 7,328 8,071 8,588 9,202 9,504 9,846 Personal Good Failed Unknown Good 2,539 2,589 2,603 2,642 2,627 2,618 2,595 2,723 2,760 2,786 2,833 2,876 Failed Yes Good 11,148 12,773 13,191 13,675 13,879 14,542 15,093 15,308 15,578 16,328 16,794 17,379 Failed 1,152 1,365 1,515 1,697 1,834 2,020 2,093 2,285 2,415 2,538 2,627 2,697 Table 6b PERCENTAGE OF FAILED CONTRACTS By Asset Code and Quarter Guarantor Code Both 33.3% 25.0% 25.0% 16.7% 16.7% 14.3% 12.5% 12.5% 12.5% 12.5% 12.5% 12.5% Corporate 0.0% 0.0% 0.0% 0.0% 25.0% 25.0% 25.0% 25.0% 25.0% 25.0% 25.0% 25.0% None 8.2% 8.7% 9.3% 9.9% 10.6% 11.0% 11.1% 11.5% 11.8% 11.8% 11.8% 11.7% Personal 4.6% 6.9% 7.3% 8.5% 9.3% 11.2% 11.2% 11.7% 12.1% 14.8% 14.6% 14.4% Unknown 5.5% 6.3% 7.8% 9.0% 10.9% 12.5% 14.2% 14.9% 15.4% 16.3% 16.6% 16.5% Yes 9.4% 9.7% 10.3% 11.0% 11.7% 12.2% 12.2% 13.0% 13.4% 13.5% 13.5% 13.4% J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O. 1 7

8 Table 7 MEANS AND DIFFERENCE BETWEEN MEANS (FAILED-GOOD) All s (Jan.2004 Data) January, 2004 Data Mean Difference Variable Good Failed (Failed-Good) receivable 114, , , ** payment amount 4, , (106.83) Active scheduled payments count ** Loss amount , , ** Balance amount 21, , (7,073.38) * days past due ** days past due ** over 90 days past due ** days past due amount ** days past due amount ** 90+ days past due amount ** Average number of days past due ** ** significant at 1% level, * significant at 5% level Table 8 DIFFERENCES BETWEEN MEANS (FAILED-GOOD) By SIC Code (Jan.2004 Data) SIC n Receivable Count Loss Agriculture, Forestry, Fishing 5,173 29, ** (176.27) 5.32 ** 12, ** Construction 964 1, (399.34) 4.91 ** 5, ** Finance, Insurance, Real Estate 12,604 31, , * 4.01 ** 1, Manufacturing 17,989 (769.01) (3,624.84) ** 4.59 ** 4, ** Mining 14,412 (18,057.29) (602.03) Public Adminstration 8,019 75, ** 1, * 9.17 ** ** Retail Trade 5,403 62, * 10, ** 1, ** Services 8,802 4, ** 1, ** Transportation, Public Utilities 28,971 43, (1,021.30) 4.05 ** 2, ** Wholesale Trade 5,308 (4,514.91) (687.97) * 6.22 ** 3, ** days days 90+ days SIC n Balance Past Sue Agriculture, Forestry, Fishing 5,173 (5,204.04) * 5.86 ** 3.99 ** 5.66 ** Construction 964 (4,623.40) 7.09 ** 3.85 ** 4.19 ** Finance, Insurance, Real Estate 12,604 2, ** 2.40 ** 3.16 ** Manufacturing 17,989 (15,386.49) ** 5.98 ** 3.72 ** 5.25 ** Mining 14,412 (23,856.62) 3.75 ** 2.40 ** 4.56 ** Public Adminstration 8,019 (11,972.82) ** 6.08 ** 4.98 ** 6.35 ** Retail Trade 5,403 (14,070.52) ** 4.91 ** 2.79 ** 3.26 ** Services 8,802 (11,356.45) ** 4.87 ** 3.16 ** 3.54 ** Transportation, Public Utilities 28,971 5, ** 5.82 ** 5.33 ** Wholesale Trade 5,308 (19,924.55) ** 6.40 ** 3.70 ** 4.47 ** SIC n Day Day 90+ Day Avg Days Agriculture, Forestry, Fishing 5, * * ** Construction ** ** ** ** Finance, Insurance, Real Estate 12, * * ** Manufacturing 17, ** ** 1, ** Mining 14,412 (13.43) Public Adminstration 8, * ** * ** Retail Trade 5, ** * * ** Services 8, * ** * ** Transportation, Public Utilities 28, ** ** ** ** Wholesale Trade 5, * ** ** **Significant at 1% level *Significant at 5% level 8 J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O. 1

9 the means for successful and failed leases were calculated by subtracting the mean of the successful contracts from that of the failed contracts. A positive result (difference) indicates that the mean of the failed contracts is greater than that of the nonfailed contracts. Based on this calculation (failed versus not failed), the difference between the means for the number of times past due, amounts past due, loss amount, and average days past due should all be positive. The difference between the means for the original contract amount and payment amount are not expected to be significantly different from zero. Based on the assumption that lenders make effective initial assessments of credit risk, defaults early in the contract term are not expected. As a result, the difference in payment count is expected to be positive and the difference in balance amount is expected to be negative. The standard deviation and the z-statistic were also calculated for these differences. The differences and levels of significance for the January 2004 data are summarized in tables 7 through These results control for potentially influential factors such as SIC classification, contract term, contract type, asset type, and type of guarantee. The means and differences between means (failed-good) for the January 2004 data are presented in table 7 (page 8). For the overall sample, the only factor that was not statistically significant was the difference between the contract payment amounts. On average, failed contracts had higher original contract receivable amounts and larger reported loss amounts. Failed contracts also had lower average balance amounts, perhaps indicating an older subset of contracts. The failed contracts had a greater number of times past due in all aging categories, the average number of days past due is significantly higher, and the mean dollar amounts in all aging categories are greater. Table 8 (page 8) presents the differences between means for the same key contract characteristics. However, the results are further categorized by industry classification based on the reported SIC code for the contracts. Differences in the results across industry classification may be indicative of industry influences in contract defaults. All industry classifications show a greater number of times past due for failed contracts in all aging categories. The difference in the average number of days past due is also positive, ranging from 13.5 days (retail) to 75.3 days (construction). The failed contracts also exhibited greater past due amounts and loss amounts in nearly all industry classifications. The difference in the original contract receivable amount was significant in three SIC groups: agriculture, public administration, and retail trade. Failed contracts in these industries were characterized by larger original contract amounts than the nonfailed contracts. This difference, though generally positive in the other industry groups, was not statistically significant. The results were mixed for the original contract payment amount. The finance, insurance, and real estate industries and public administration industry show statistically significant and larger payments for failed contracts. Manufacturing and wholesale trade have smaller payments for failed contracts. Lastly, contract payment counts are greater and balance amounts are lower for failed contracts, perhaps indicating that the contracts were further along in the payment stream when they failed. The length of the contract term should be an important factor in assessing default risk. Longer term contracts offer a longer period for economic, industry, or company factors to negatively affect the firm's ability to fully repay the contract. Differences between the means for contract characteristics controlling for original contract term are shown in table 9 (next page). The difference in the original contract receivable amount is significant and positive in the 25- to 36-month term, indicating that the failed contracts in this term had larger original balances than the nonfailed contracts. The opposite is true for the contracts over 60 months. In this instance, failed contracts appear to have been smaller. This result could indicate that riskier contracts were extended for longer terms. The length of the contract term should be an important factor in assessing default risk. Longer term contracts offer a longer period for economic, industry, or company factors to negatively affect the firm's ability to fully repay the contract. J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O. 1 9

10 Table 9 DIFFERENCES BETWEEN MEANS (FAILED-GOOD) By Term (Jan Data) Term n Receivable Count Loss 0-12 months 2,989 59, (29,185.19) ** 3.18 * 12, * months 6,532 71, , ** 2.96 ** 1, * months 32,594 57, * ** 2.78 ** 1, ** months 23,891 10, ** 3, ** months 32,911 (9,814.04) (500.12) * 4.60 ** 2, ** 61+ months 8,728 (50,952.84) * (953.83) 7.76 ** 8, ** Days Days 90+ Days Term n Balance 0-12 months 2,989 14, ** 2.46 ** 5.00 ** months 6,532 12, ** 2.78 ** 4.04 ** months 32,594 8, ** 3.29 ** 4.06 ** months 23,891 (10,627.74) ** 5.72 ** 3.50 ** 4.11 ** months 32,911 (12,938.98) ** 7.80 ** 4.41 ** 4.28 ** 61+ months 8,728 (56,658.40) ** 9.07 ** 5.38 ** 6.75 ** Term n Day Day 90+ Day Avg Days 0-12 months 2, months 6,532 (1.85) * months 32, * ** ** ** months 23, ** ** 1, ** months 32, ** ** * ** 61+ months 8, ** ** 1, ** ** **Significant at 1% level *Significant at 5% level Table 10 DIFFERENCES BETWEEN MEANS (FAILED-GOOD) By Type (Jan.2004 Data) Type n Receivable Count Loss Conditional Sale 18,814 53, ** ** 4, ** Leverage Lease 7 (625,726.67) (55,540.22) N/A Loan 7,350 (122,253.35) ** (9,868.73) ** 3.82 ** 4, ** Lease Purchase 5,769 (173,356.04) ** (5,624.29) ** (65.14) ** (4,949.44) ** Revolver 382 (670,136.69) ** (118,549.57) ** (10.06) ** (2,835.56) Rental Lease 1,698 (29,112.74) (2,602.05) ** 8.24 ** * True Lease 73,621 23, ** 1, ** 5.06 ** 2, ** Days Days 90+ Days Type n Balance Conditional Sale 18,814 (12,645.76) ** 9.96 ** 6.46 ** 5.40 ** Leverage Lease 7 (5,270,487.80) 6.00 * * Loan 7,350 (27,253.21) ** 4.53 ** 2.12 ** 2.19 ** Lease Purchase 5,769 (18,652.29) ** (18.24) ** (8.79) ** (6.92) ** Revolver 382 (179,097.70) (2.83) ** (1.71) ** (6.32) ** Rental Lease 1,698 (12,657.44) ** 8.48 ** 4.96 ** 5.09 ** True Lease 73,621 (8,520.20) ** 4.94 ** 3.08 ** 4.24 ** Type n Day Day 90+ Day Avg Days Conditional Sale 18, ** ** 2, * ** Leverage Lease 7 N/A N/A N/A * Loan 7, ** ** 1, ** ** Lease Purchase 5,769 (165.99) ** (261.36) ** (1,627.09) ** (149.28) ** Revolver 382 N/A N/A N/A (0.07) Rental Lease 1, * * 1, ** True Lease 73, ** ** ** ** **Significant at 1% level *Significant at 5% level 1 0 J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O. 1

11 The difference in the contract payment amount was significant in four of the six term categories. In the short term contracts (0 to 12 month), the difference is negative, indicating that failed contracts had lower payments. This is also true for the 49- to 60-month contracts. The 13- to 24- month and the 25- to 36-month contracts show a positive and significant difference. As expected, the differences for the number of times past due, past due amounts, and loss amounts are generally significant and positive. The average number of days past due for the failed contracts exceeds that of the nonfailed contracts by 32 to 64 days. The differences in the payment counts are positive and the significant differences in contract balances are negative. This result is consistent with the results found in table 8. The PayNet data used in this study includes seven different contract types. Although the true leases account for nearly three-quarters of the sample, conditional sales, lease purchases, and loans represent a substantial portion of the sample as well. Other contract types include rental lease, revolvers, and leveraged leases. Table 10 (page 10) shows the differences between means by contract type based on the January 2004 data. Again, the differences in the number of times past due, average days past due, and amounts past due are generally significant and positive. The exceptions are the lease purchases and the revolver contracts, which exhibit negative and significant differences. These mixed results persist in the other characteristics studied. For the original contract receivable amount and the contract payment amount, the differences are negative for loans, lease purchases, and revolvers and positive for true leases. The differences in the payment counts are similarly mixed. The differences in the balance amounts are all negative, with significant results for all contract types except leveraged leases and revolvers. Table 11 (page 12) shows the differences between the means organized by the reported contract asset code. The data studied includes 21 separate asset codes. This categorization provides the finest level of detail for identifying differences in the characteristics of failed and successful lease contracts. The differences are positive for the number of times past due for all aging categories, and nearly all are significant. The difference in the average number of days past due is also positive. This difference is significant in 13 of the 21 asset codes studied. The difference in the loss amount is also positive, as expected. The difference in the payment count is significant and positive in 11 of the asset codes, ranging from 1.45 days (trucks) to days (railroad). Although fewer show statistical significance, the differences in the past due amounts are generally positive for all aging categories. This result is consistent with expectations. The difference in the original contract receivable amount is significant in only five of 21 asset types. Copiers, forklifts, and marine craft show positive differences for the original receivable amount. Manufacturing equipment and medical equipment exhibit negative differences. The differences in the contract payment amount also show mixed results with only six of 21 asset types significant. The difference is negative for computers, medium/light duty trucks, telecommunications equipment, and the unknown category. Forklifts and retail equipment exhibit a positive difference, indicating that failed contracts for these asset types had a larger average payment amount. The differences in the balance amounts are negative in 18 of the cases and significant in 12 of these. This result is consistent with previous findings, supporting the idea that the failed contracts are more mature than the nonfailed. The PayNet database captures information about the type of contract guarantee through the guarantor code. More than three-quarters of the contracts have a guarantee code of none, and 20% are either yes-undefined or unknown. Corporate, personal and both combined account for less than 1% of the sample. Although table 12 (page 13) appears consistent with previous tables, the lack of variation in guarantor codes across this sample makes it more difficult to draw conclusions from the results. As expected, the differences for the number of times past due, past due amounts, and loss amounts are generally significant and positive. J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O

12 Table 11 DIFFERENCES BETWEEN MEANS (FAILED-GOOD) By Asset Code (Jan.2004 Data) Asset Code n Receivable Count Loss Agricultual 3,720 38, (325.78) 7.29 ** 12, ** Aircraft 213 6,639, , , Buses & Motor coaches 19 2,137, , Computer 526 (50,509.07) (2,296.53) ** , Construction & Mining 11,371 (12,830.69) (294.02) 5.87 ** 6, ** Copier & fax 12,660 61, ** 1, ** 1, ** Forklift 39,285 4, ** ** 4.90 ** ** Logging & Forestry 1,680 (79,060.58) (990.26) , ** Manufacturing 207 (74,920.91) ** (6,404.06) , Marine craft 3, , ** 11, ** 4, ** Medical 1,048 (154,266.94) ** (3,984.93) (1.16) 14, ** Medium/Light duty truck 5,653 (18,024.57) (4,980.87) ** 4.91 ** 10, ** Office Equipment 3,122 47, , ** Printing & Photographic 487 (75,687.38) , Railroad , (9,375.10) ** Retail 3,672 49, , ** ** 2, Telecommunications 2,472 19, (5,305.60) ** , ** Truck 13,255 (2,603.49) (1,495.75) 1.45 ** 2, ** Unknown 2, , (18,035.16) ** 7.05 ** 4, Vending & Restaurant 1,205 1, (4,759.63) (1.52) 1, Waste & Refuse Handling , , (2.46) 47, Days Days 90+ Days Asset Code n Balance Agricultual 3,720 (11,868.06) ** 6.59 ** 4.33 ** 5.73 ** Aircraft 213 3,310, ** ** Buses & Motor coaches 19 1,374, ** 6.94 ** 2.00 Computer 526 (60,639.09) ** 7.98 ** 4.90 ** 5.91 ** Construction & Mining 11,371 (13,641.00) 6.96 ** 3.67 ** 4.59 ** Copier & fax 12,660 (5,277.70) 5.04 ** 3.88 ** 4.72 ** Forklift 39,285 (4,090.79) ** 4.72 ** 2.74 ** 3.10 ** Logging & Forestry 1,680 (49,516.68) ** 4.09 ** 2.53 ** 4.73 ** Manufacturing 207 (23,634.39) 8.35 ** 4.37 ** ** Marine craft 3,500 (10,223.92) ** 8.74 ** 4.54 ** Medical 1,048 (45,993.45) ** 5.51 ** 3.62 ** 3.22 ** Medium/Light duty truck 5,653 (32,971.28) ** 9.00 ** 5.48 ** 7.38 ** Office Equipment 3,122 (6,752.17) 6.63 ** 4.17 ** 4.09 ** Printing & Photographic 487 (52,140.50) ** 8.57 ** 5.44 ** 5.56 ** Railroad , ** 4.10 ** 3.10 ** Retail 3,672 (12,968.52) ** 8.17 ** 4.22 ** 4.34 ** Telecommunications 2,472 (1,132.99) 6.27 ** 4.18 ** 4.49 ** Truck 13,255 (17,274.55) ** 9.41 ** 5.77 ** 4.90 ** Unknown 2,938 (21,714.63) ** 7.44 ** 5.04 ** 6.67 ** Vending & Restaurant 1,205 (25,382.56) ** 4.86 ** 3.18 ** 2.44 ** Waste & Refuse Handling 573 (10,480.97) ** 6.77 ** 8.41 ** Table 11 continued on next page 1 2 J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O. 1

13 Table 11 continued DIFFERENCES BETWEEN MEANS (FAILED-GOOD) By Asset Code (Jan.2004 Data) Asset Code n Receivable Count Loss Agricultual 3,720 38, (325.78) 7.29 ** 12, ** Aircraft 213 6,639, , , Buses & Motor coaches 19 2,137, , Computer 526 (50,509.07) (2,296.53) ** , Construction & Mining 11,371 (12,830.69) (294.02) 5.87 ** 6, ** Copier & fax 12,660 61, ** 1, ** 1, ** Forklift 39,285 4, ** ** 4.90 ** ** Logging & Forestry 1,680 (79,060.58) (990.26) , ** Manufacturing 207 (74,920.91) ** (6,404.06) , Marine craft 3, , ** 11, ** 4, ** Medical 1,048 (154,266.94) ** (3,984.93) (1.16) 14, ** Medium/Light duty truck 5,653 (18,024.57) (4,980.87) ** 4.91 ** 10, ** Office Equipment 3,122 47, , ** Printing & Photographic 487 (75,687.38) , Railroad , (9,375.10) ** Retail 3,672 49, , ** ** 2, Telecommunications 2,472 19, (5,305.60) ** , ** Truck 13,255 (2,603.49) (1,495.75) 1.45 ** 2, ** Unknown 2, , ** Vending & Restaurant 1, ** Waste & Refuse Handling , **Significant at 1% level *Significant at 5% level Table 12 DIFFERENCES BETWEEN MEANS (FAILED-GOOD) By Guarantor Code (Jan.2004 Data) Guarantor Code n Receivable Count Loss Both 8 52, ** (20.43) ** N/A Corporate , , (19.89) ** 284, ** None 83,896 30, ** ** 2, ** Personal 209 (16,878.04) ** (284.32) (2.82) 6, * Unknown 3,444 (15,109.65) * (1,358.74) ** (0.13) 3, ** Yes - undefined 20,076 (3,101.22) (1,227.47) 6.09 ** 5, ** Days Days 90+ Days Guarantor Code n Balance Both 8 (17,339.14) * 0.43 N/A N/A Corporate 12 (51,207.67) 2.44 ** N/A N/A None 83,896 (3,118.86) 5.97 ** 3.66 ** 4.24 ** Personal 209 (11,853.60) ** 7.09 ** 5.93 ** 8.70 ** Unknown 3,444 (18,416.38) ** 5.71 ** 2.75 ** 2.82 ** Yes - undefined 20,076 (22,201.14) ** 8.30 ** 5.08 ** 5.58 ** Guarantor Code n Day Day 90+ Day Avg Days Both 8 N/A N/A N/A N/A Corporate 12 N/A N/A N/A N/A None 83, ** ** ** ** Personal * ** Unknown 3, ** * * ** Yes - undefined 20, ** ** 1, ** ** **Significant at 1% level *Significant at 5% level J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O

14 When comparing means between failed and successful contracts, payment counts tend to be higher and balance amounts lower for failed contracts. Some form of ongoing credit monitoring could be useful to identify changes in default risk during the life of the contract. CONCLUSIONS The PayNet database provides an extensive resource for analyzing lease contract characteristics, industry factors, and economic factors that influence lease default rates. Consistent differences in industry and asset class default rates can be seen in the sample studied. Based on industry classification, the results provide evidence of industry-level risk factors across each of the 12 quarters studied. Greater industry risk may lead to higher risk premiums in specific industries and place greater emphasis on effective risk models in these industries. Similar results are found when contract term, contract type, and asset type are studied. The results support the existence of risk premiums for particular industries and asset types. When comparing means between failed and successful contracts, payment counts tend to be higher and balance amounts lower for failed contracts. This result suggests that some form of ongoing credit monitoring could be useful to identify changes in default risk during the life of the contract. This analysis could also be used to create more effective risk management models as proposed under the Basel II accords. Future research could combine economic and industry factors such as interest rates and interest rate spreads, economic growth, and industry activity measures with PayNet data to develop more comprehensive default models. At the firm level, further research should empirically investigate the opportunities for diversification in contract characteristics and subsequent reductions in borrowing costs. Acknowledgments The author acknowledges the financial support of the Equipment Leasing and Finance Foundation and data support from PayNet. Corey Trager of PayNet was especially helpful in providing valuable insights into the available PayNet data fields and creating the quarterly data files. Foundation referees provided helpful editorial recommendations. Any remaining errors are the responsibility of the author. Endnotes 1 The Basel II Accord: What DoesIt Mean for the North American Leasing Market?, research report of the Equipment Leasing and Finance Foundation, 2003, p The mission of PayNet is to provide comparable payment history for leasing contracts. The Foundation and PayNet have established a strategic partnership to provide access to the details of millions of transactions to be used for industry research studies. The PayNet database provides information on a variety of lease contracts in a consistent, understandable format. This database contains more than 50 standardized fields including creditor business and financial background, a detailed credit check, current and historical debt levels, creditor payment habits, legal judgments, and transaction-level loss information. The data is segmented by geographical region, industry, and equipment groupings. 3 The complete descriptive statistics from tables 7-12 are available from the author upon request. References The Basel II Accord: What Does It Mean for the North American Leasing Market? research report of the Equipment Leasing and Finance Foundation, Cowan, Adrian and Charles Cowan, Default Correlation: An Empirical Investigation of a Subprime Lender, Journal of Banking & Finance, vol. 28 (2004), pp Jackson, Suzanne, Better Credit Data, Better Credit Decisions, Equipment Leasing Today, June/July 2001, pp Kleinman, Robert, The Characteristics of Venture Lease Financing, Journal of Equipment Lease Financing, Spring, vol. 19 (2001), pp Neagle, Bob, Time To Get Busy: Information-based Risk Management and Tools for Basel II, Equipment Leasing Today, June/July 2003, pp Song, Michael and Ben Sopranzetti, Leasing and the Small Firm, research report of the Equipment Leasing and Financing Foundation, Ware, Thomas, The Future of Commercial Credit Data, Equipment Leasing Today, June/July 2002, pp J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O. 1

15 James P. Murtagh, PhD James P. Murtagh is a clinical assistant professor of finance in the Lally School of Management & Technology, Rensselaer Polytechnic Institute (RPI), in Troy, N.Y. Dr. Murtagh teaches in the Lally School's undergraduate, graduate, and executive programs and provides expert financial management advice to a select client list. His expertise covers financial management, corporate finance, financial markets and institutions, and entrepreneurial finance. Dr. Murtagh is an internationally recognized scholar with experience in executive programs in Russia and the Ukraine, in addition to Korean and Chinese programs at Rensselaer Polytechnic Institute. He regularly presents research at regional, national, and international conferences and has published in the Journal of International Financial Markets, Institutions and Money, International Trade and Finance Association proceedings, and proceedings of the IEEE Engineering Management Society Management Conference. He is currently dedicated to examining the timely issues of U.S. and international banking, international capital markets, risk management in leasing contracts, and entrepreneurial finance. He received his PhD in finance from RPI in 1998, his MBA from the University of Northern Colorado, in Greeley, in 1987, and his BS in engineering from the United States Military Academy (West Point, N.Y.) in J O U R N A L O F E Q U I P M E N T L E A S E F I N A N C I N G W I N T E R V O L. 2 3 / N O