國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
4. Data Descriptions
In this section, we briefly describe primary data sources of our research and how we construct our data set, e.g., how performance pricing types are defined. Then we show the explanatory variables we will use in our empirical analysis. Lastly we provide descriptive statistics regarding loan facilities, borrower characteristics for our full sample as well as subsamples of loans that include interest-decreasing and interest-increasing performance pricing provisions.
4.1 Data Sources
We obtain loan sample from the Thomson Reuters Loan Pricing Corporation Dealscan (LPC`s Dealscan) database, which contains detailed information on syndicated loan contracts since 1982. We extract our sample of loans issued by U.S non-financial borrowers from 1993 to 20103. We conduct our analysis on the facility (tranche) level because performance pricing features are determined for each facility. We obtain information on loan characteristics such as facility amount, facility maturity, loan type, as well as loan facility purpose. In addition, we record whether a loan is secured or not. We exclude loans whose size (loan amount), and maturity are missing from Dealscan.
To obtain borrower-specific information, We merge our loan data with Standard and Poor`s Compustat North American database4, such as firm size, S&P credit rating of firms, and market-to-book ratio , from the last available fiscal quarter before the
3 Prior to 1993, virtually no contracts include a performance-pricing provision according to LPC. We conclude that Dealscan’s data quality with respect to PPC in insufficient prior to 1993 as PPC was first introduced in the 1970s.
4 We use Michael Robert`s Dealscan-Compustat Linking Database to merge Dealscan with Compustat.
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
loan issue. We exclude those borrowers whose financial data (firm size, leverage, market value ... etc.) are missing in Compustat.
Using the Institutional Brokers' Estimate System (I/B/E/S) data base, we obtain the number of analysts following the firm, which is a proxy for borrower opacity. Our final dataset consists of a sample of 26,877 closed or completed loan facilities in the U.S. market. Appendix I contains the definitions of all variables in our analysis.
4.2 Performance Pricing Types
We follow Cai et al. (2012) to distinguish between different performance pricing types.
Because Dealscan database cannot indicate whether a performance pricing provision is interest-increasing or interest-decreasing directly, we use the pricing grid of Dealscan to identify the specific type. The pricing grid provides us detailed pricing information and the requirement based on either financial ratio or credit rating at each degree. Financial ratios can be divided into two groups: the first group with smaller ratios indicating better performance (e.g., leverage ratio), while the second group with higher ratios indicating better performance (e.g., EBITDA). Based on the pricing grid, we determine the highest and lowest levels of pricing. According to Dealscan, all-in-spread-drawn, which expressed as a spread over LIBOR, is a measure of the overall cost of loans at starting point of loan facilities.
A performance-pricing covenant is defined as interest-increasing if the starting all-in-spread-drawn is at the lowest level in the pricing grid and the performance measure (financial ratio or credit rating) is at the lowest or highest level depending on which group it is in. On the contrary, a performance pricing covenant is defined as interest-decreasing if the all-in-spread-drawn is the highest at starting point of the loan
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
facility and the performance measure is the highest or lowest depending on its group.
If the all-in-spread-drawn at the time of origination is neither at the highest nor lowest level of the pricing grid, we label the loan facility as one with both interest-increasing and interest-decreasing performance pricing (mixed). We further identify whether interest-increasing and interest -decreasing performance pricing covenants are based on financial ratios or credit ratings. Accounting-based PPC uses measures of financial ratios and Rating-based PPC relies on firm`s credit rating.
4.3 Measuring Relationship Strength
We follow Bharath et al. (2011) and construct three alternative proxies of relationship strength by searching the past borrowing record of borrowers. These proxies are designed to pick up the existence of prior lending by the same lender in the past. To construct these proxies, we first need to identify the lead lender(s) for each loan contract. We classify a lender as the lead lender if the variable "Lead Arranger Credit"
(provided by LPC’s Dealscan) takes on the value "Yes", or if the lender is the only lender specified in the loan contract.
Next, we search the borrowing record of borrowers over the past five years. The first lending relationship strength proxy, Rel(Dummy), is a dummy variable which equals one if the firm borrowed from the same lead lender in the previous five years and zero otherwise. If there are multiple lead lenders retained, we calculate the proxy separately for each lender and assign the highest value to the loan. The second proxy, Rel(Number), measures the relative number of loans obtained from the relationship lender. For bank m lending to borrower i, it is calculated as follows.
‧
umber of loans by borrower i l Number
Total number of loans by borrower i
Again, the highest value is assigned to a loan if there are multiple lead lenders.
The third proxy, Rel(Amount), measures the relative loan amounts obtained from the relationship lender. For bank m lending to borrower i, it is calculated as follows.
Amount bank m to in the last 5 years ($)
Re ( )
amount in the last 5 years ($)
m
of loans by borrower i l Amount
Total of loans by borrower i
Again, the highest value is assigned to a loan if there are multiple lead lenders.
4.4 Lead Arranger Ranking
To assign the lead arranger ranking score, we construct the league table for lead arrangers in USA syndicated loans market between 1993 and 2010. The lending amount for each bank is estimated from each deal’s lending amount times its bank allocation. We use bank-allocation and lending amount in LPC`s Dealscan database. If bank-allocation is missing or the sum value of each facility is less than 100%, we assume leader arrangers take responsibility for the whole lending amount. Table 4 shows five largest arrangers of USA syndicated loans from 1993 to 2010.
[Insert Table 4 here]
Because our analysis is on the facility (tranche) level, we construct a variable
“Bank Ranking” which refers to the lead arranger ranking score of each tranche according to the league table from 1993 to 2010, the lead arranger ranking score ranges from ten to zero. For example, a score of ten means the lead arranger in one
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
facility is the top lead arranger in syndicated loan market and a score of zero means it is out of the top ten lead arrangers .If there are multiple lead arrangers in one facility, we assign the highest lead arranger ranking score to the facility.
4.5 Descriptive Statistics
Table 5 presents descriptive statistics for our sample consisting of 26,877 loan tranches issued by 4,930 non-financial borrowers between 1993 and 2010. The data are winsorized at the 1% and 99% levels to remove outliers. Panel A reports loan characteristics, which are consistent with prior studies (e.g. Adam (2014)).For example, the mean/median facility amounts in our sample are $317/$106 million, which is large given the mean/median book value of assets of $4,559/$814 million and an average leverage ratio of 30%. The average maturity is 47 months. 27% of loan tranches are term loans. We can also observe that roughly 42% of loans include a performance-pricing provision. Panel B reports borrower characteristics variables. In 59% of cases, borrowers do not have a credit rating. But if a rating exists it tends to be around the investment grade threshold. Panel C shows the mean/median lead arranger ranking score we assign in our samples. Because we assign lead arranger ranking score from 10 to 0 , we observe that at least 50% facility is conducted by the top 4 lead arranger in our sample( median arranger ranking score of 7). Panel D reports descriptive statistics on the three relationship lending proxies. A lending relationship exists in 39% of all loan contracts. On average, 17% of the total capital raised over the course of 5 years was raised from the same lead lender.
Table 6 shows the various performance measures used in PPC contracts. The most common performance measure is the Debt-to-EBITDA ratio (48%), followed by
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
the senior debt rating (28%). The remaining performance measures are mostly other types of leverage ratios. In at least 4% of cases, leverage performance measures are used. We define PPC as accounting-based PPC whenever a financial ratio is used as a measure of firm performance. Rating-based PPC comprise all PPC contracts, which use borrower’s credit rating as a performance measure.
[Insert Table 5 here]
[Insert Table 6 here]