Banks rely upon information in their loan granting decisions. While some of the mechanisms through which lenders accumulate information are institutionalised (e.g. credit bureaus and registers), the accumulation of credit market experience often occurs as a byproduct of repeated interactions with borrowing firms, borrowers’ peers and competing lenders. Previous work by De Jonghe et al. (2020) for example shows the role of banks’ industry experience in credit reallocation in the face of an interbank funding shock while Papoutsi (2021) shows that continuous lending relationships between bank loan officers and corporate borrowers improve the outcomes of loan renegotiations.
In a new paper (Degryse et al. 2021), we study the impact on credit market outcomes of three forms of experience accumulated by banks – sectoral experience, experience on borrowing firms, and experience on co-lender banks. The way banks’ past experience interacts with their screening and monitoring of borrowers is far from obvious. Past experience provides a valuable stepping stone for monitoring activities as it can enhance the productivity of monitoring and reduce the costs of acquiring information on borrowers. Nevertheless, experience could also make lenders ‘lazy’ – counting on knowledge previously accumulated in the credit market, lenders could have a natural incentive to shirk on their costly monitoring duties.
The syndicated loan market provides a natural setting for studying the impact of credit market experience on lending outcomes, firm behaviour, and banks' (mis)behaviour vis-à-vis other syndicate members. Over the course of frequent and repeated interactions in lending consortia, syndicate members learn from the actions and decisions of other syndicate members, and garner valuable experience on firms, sectors of activity and other banks. We can then unpack lenders' experience into its multiple dimensions and study to what extent its various components improve, or worsen, credit market outcomes.
Our theoretical model underpins how past experience impacts lending decisions through its effects on monitoring costs and the outside option of the lead arranger. We first discuss the theoretical predictions for the intensive margin (share of loan retained by the lead arranger). Two forces could be at work. First, past experience makes it cheaper for the lead arranger to monitor the borrower, implying that a lower lead share is more likely to be observed (i.e., there is substitutability between lead arranger's share and experience). Second, complementarity between a lead arranger's share and past experience can arise if the past information of the lead arranger exacerbates the risk of her opportunistic behaviour. Specifically, past experience can make the lead arranger lazy, for example raising her outside option in case of borrower's default (Shleifer and Vishny 1992).
At the extensive margin, our model predicts that when a bank’s experience eases monitoring, the likelihood that the bank acts as a lead arranger increases. If, instead, experience increases the bank's outside option, its predicted impact on the probability that the bank acts as a lead arranger is ambiguous. Figure 1 is based on Ivashina (2009) and helps illustrate the setting. The downward sloping participant-demand curve represents the lead share demand of the participants, meant as the lead share that induces them to participate for a given repayment. The upward sloping lead-supply curve gives the share under which a bank is willing to act as a lead arranger. Under the substitutability assumption, in Figure 1 the participants' demand for the lead arranger's share is shifted downward. While, under the complementarity assumption it is shifted upward.
Figure 1 Supply and demand curve
To address our research question, we use syndicated loan-level data on 20,932 loans originated by 663 banks to 5,309 non-financial firms. The data span 64 industries (2 digit SIC) from 1987 until 2014. We match syndicated loans with detailed data on the characteristics of firms. Our data set allows us to construct three measures of experience accumulated by a bank.
- Firm experience. The first measure is based on the number of times that the bank has interacted with a firm in previous syndicates. A key feature of this measure is that it is constructed for all active banks and not solely for lead arrangers. This is important as a participant bank may also learn about a borrower during its interactions in a lending consortium.
- Sector experience. The second measure relies on the sectoral specialisation acquired by a bank through repeated interactions with borrowing firms operating in a specific sector.
- Bank experience. The third measure focuses on interactions among banks and consists of the number of previous interactions between the lead arranger and participants in syndicated deals. This measure captures the degree of learning from prior interactions with other banks over the course of syndicated loans.
Firm experience and sector experience are often viewed as core components of relationship lending technologies (see e.g., Boot and Thakor 2000).
We find evidence that supports the role of the three forms of banks’ experience at the intensive and extensive margin of loan syndication.
At the extensive margin, we find that a one standard deviation increase in:
- firm experience leads to a 16% higher likelihood of being lead arranger in future deals;
- sector experience leads to a 5% higher likelihood of being lead arranger in future deals;
- bank experience leads to a 4% higher likelihood of being lead arranger in future deals.
Our results suggest quite a nuanced impact of lenders' experience regarding the intensive margin, i.e., the loan share a lead arranger needs to hold to attract co-lenders. In particular, we find that a one standard deviation increase in:
- firm experience, leads to a 1% reduction in the lead share;
- sector experience, leads to a 4% increase in the lead share;
- bank experience, leads to a 2% reduction in the lead share.
Thus, the estimates suggest that moral hazard within syndicates can be more severe when the lead arranger has previously accumulated stronger sectoral experience. We surmise that this could be due to the lead arranger having a larger outside option in case of failure of the loan and that this may dilute its incentive to properly monitor the loan.
To enhance our identification strategy and dissect the scenarios in which bank experience is more likely to affect lending syndicates, we take several steps. First, we exploit cross-sectional variation in sectors and firm characteristics to study whether our estimated effects are larger in these sectors and firms where we expect our mechanism to be at work. We find that the role of sectoral experience is more pronounced in sectors with a greater degree of informational complexity of products (Rauch 1999, Caballero et al. 2018). The effects of lenders' experience are more pronounced for less profitable and more informationally opaque firms (Delis et al. 2017). Second, we exploit hand-collected information on regulatory enforcement actions enacted on banks that are active in the syndicated market. Sanctions impose a reputational stigma on punished banks (Delis et al. 2020). We find that co-lenders with prior bank experience with a punished lead arranger have a higher propensity to step up and start acting as a lead arranger. That is, experience enhances the flexibility with which banks can replace co-lenders hit by reputation shocks. Third, we mitigate any lingering concern that banks' experience may be endogenous to the propensity to grant new loans or to the structure of loan syndicates by exploiting changes in credit market experience that stem from bank mergers (Favara and Giannetti 2017). To this end, we use hand-collected information on bank mergers where both banks are active in the syndicated loan market in the year before the merger. Specifically, we instrument a bank's credit market experience with information from the acquired bank in the last quarter of the pre-merger. This instrument is likely to satisfy the exclusion and relevance restrictions because it affects only peers' activities. Our results are fully robust to this IV strategy.
Experience is traditionally viewed as a fundamental mechanism of acquisition of information and knowledge in the banking sector. Experience can reduce the costs of monitoring borrowers, thus incentivising banks' monitoring effort. It can, nonetheless, also improve banks' outside option in the event of inadequate monitoring, thereby diluting banks' monitoring incentives. Our findings suggest that both experience on borrowers and experience on co-lenders incentivise banks' screening and monitoring effort, mitigating moral hazard issues in lending syndicates. By contrast, in our data we find evidence that sectoral experience exacerbates moral hazard issues. We further document that, by affecting moral hazard issues in lending syndicates, experience also gives banks flexibility in responding to negative shocks hitting co-lenders. This dynamic view of banks' experience can yield new insights into the role of banks in the aftermath of shocks. We leave this and other relevant issues to future research.
Boot, A W, and A V Thakor (2000), “Can relationship banking survive competition?”, The Journal of Finance 55 (2), 679-713.
Caballero, J, C Candelaria, and G Hale (2018), “Bank linkages and international trade”, Journal of International Economics 115, 30-47.
Degryse, H, S Kokas, and R Minetti (2021), “Banking on Experience”, CEPR Discussion Paper 16142.
De Jonghe, O, H Dewachter, K Mulier, S Ongena, and G Schepens (2020), “Some borrowers are more equal than others: Bank funding shocks and credit reallocation”, VoxEU.org, 6 March.
Delis, M D, M Iosifidi, S Kokas, D Xefteris, and S Ongena (2020), “Enforcement actions on banks and the structure of loan syndicates”, Journal of Corporate Finance, 60, 101527.
Delis, M D, S Kokas, and S Ongena (2017), “Bank market power and firm performance”, Review of Finance, 21, 299.
Favara, G, and M Giannetti (2017), “Forced asset sales and the concentration of outstanding debt: evidence from the mortgage market”, The Journal of Finance, 72 (3), 1081–1118.
Ivashina, V (2009), “Asymmetric information effects on loan spreads”, Journal of Financial Economics, 92 (2), 300-319.
Papoutsi, M (2021) “Lending relationships in loan renegotiation: evidence from corporate loans”, VoxeU.org, 22 March.
Rauch, J E (1999), “Networks versus markets in international trade”, Journal of international Economics, 48 (1), 7-35.
Shleifer, A, and R W Vishny, (1992), “Liquidation values and debt capacity: A market equilibrium approach”, The Journal of Finance, 47 (4), 1343-1366.