We aim to build on this work by using the \citet{NBERw24029} measure of firm-level political risk to examine comprehensively how political risk affects private and public credit markets. We not only examine how borrowers respond to the political risk of their lenders, but we also study the political risk of the financial institution itself when it prices loans and makes lending decisions. What’s more, the granularity of our political risk measure, which is available at the quarterly level, allows us to examine how the political risk of lenders transmits through their connections with other financial institutions, showing potential network effects of these exposures. Relatedly, we show that banks are able to transmit the costs of their political risk to relationship borrowers. We combine this data with detailed information about the political activities of both borrowers and lenders, to see if managing political risk actively through lobbying and providing campaign donations to politicians can mitigate the consequences of political risk. In addition to managing political risk actively, we examine whether financial institutions respond passively to political risk, for example by cutting back on their lending share in syndicate loans. 
Previous attempts to measure firm-level political exposure have either mobilized specific shocks to political uncertainty (e.g., by using closely-contested election campaigns) or have adapted the \citet*{bakerbloomdavis2016} (BBD) measure of economy-wide policy uncertainty (EPU) to the firm-level \citep{RN2651,RN2662,RN2649,RN2661,Francis_2014}. While important first steps, such approaches suffer from a number of significant drawbacks. For example, while elections are a salient potential source of political risk, they are just one reason why firms might be exposed to more or less uncertainty. As a consequence, these findings might not generalize to other (non-election) political events. The BBD measure, on the other hand, was designed to reflect the over-time variation in the policy uncertainty of the US economy and, as documented in \citet{NBERw24029}, attempts to bring this measure to the firm-level are likely to simply pick up heterogeneous exposure to aggregate political risk rather than capturing firm-level political risk. As such, BBD-based exposure measures cannot exploit within firm variation in political risk or variation stemming from different firms within the same industry being differentially exposed to political risk over time.  Election-based (aggregate) shocks likewise do not reflect the fact that most of a given firm’s exposure appears to stem from political events that are specific to its industry or even vary within the sector over time. And indeed, \citet*{akey2017policy} show there is little persistence in a firm’s “EPU sensitivity” across election cycles.
To investigate the effects of political risk on credit markets, we use a recently developed and comprehensively validated measure of firm-level political risk, PRiskit, which measures the share of conversation about politics in a given firm’s quarterly earnings conference call with financial analysts \citep{NBERw24029}.  Intuitively, when analysts ask more questions about political topics or management volunteers more discussion of politics in their opening statement, it seems reasonable to conclude that the firm is exposed to more political risk. We exploit PRiskit to first establish the existence of firm level political risk in credit markets by showing the political risk affects both bond markets outcomes, including within a firm, such as bid-ask spreads, yields, and volume, and loan market outcomes.
Turning attention to the supply side of the market, we document that lender political risk matters inasmuch as the PRiskit of financial institutions affects the supply of credit as measured by loan and deposit growth. More importantly, taking advantage of the granularity of our data, we show that lender political risk is transmitted to borrowers, but only, as expected, when the borrower is in a long term relationship with the lender.
Having established that political risk affects credit market outcomes, we examine more closely the network effects and potential externalities as sources of political risk. Conceptually, there are three ways in which political risk can be transmitted in credit markets. Political risk can emanate directly from the borrower or from the lender, and we show that lender-level political risk affects the credit spreads of borrowers. Moreover, political risk can emanate also indirectly as lenders hold portfolios of borrowers, which are differentially exposed to political shocks. In addition, lenders tend to seek out the same set of financial institutions (business partners) to form syndicates, which exposes them to any political events experienced by those related banks.
We then ask whether financial institutions or borrowers can mitigate the impact of political risk on market outcomes by engaging actively in the political process by lobbying politicians or giving money to their election campaigns. Alternatively, lenders might passively manage the role of political risk by reducing their share retained in the syndicate.