I examine how political connections affect the private equity investments of public pension funds, using a regression discontinuity design based on close elections for state offices conferring pension board seats. I find that firms donating to winning candidates are roughly ten times more likely than firms donating to losing candidates to receive a commitment. The resulting commitments underperform other funds of donating general partners by 4.2 to 7.1 percentage points annually, charge higher fees, and tilt toward the connected politician's home state. These effects disappear following the SEC's 2011 pay-to-play rule, consistent with favoritism rather than benign explanations.

Affiliated financial intermediaries serving the same customers may share a reputation. Using investment advisers' regulatory disclosures of shared locations with affiliated banks, I find that revealed adviser misconduct reduces co-located branches' deposits by approximately four to five percent. The response concentrates in misconduct bearing on customer treatment, scales with severity, and is strongest among retail-facing advisers and common-brand affiliations, consistent with depositors updating their view of the organization. Outflows reallocate to unaffected local competitors, and exposed branches are more likely to close. Corroborated by the 2003 mutual fund scandal, the results show reputational exposure extends across separately regulated sectors.

We examine whether granular consumption behavior contains information about future default risk beyond conventional financial indicators and whether its value varies across borrowers. Using integrated credit and debit card transactions from Korea’s MyData regime, we first estimate logit models and then nonlinear XGBoost models. Seven expenditure categories are significantly associated with default risk in the logit analysis. Consumption information provides greater incremental predictive value for lower-middle-income and younger borrowers, a pattern reinforced by XGBoost, SHAP-based importance measures, and bootstrap inference. These findings highlight meaningful heterogeneity in the credit-risk information contained in consumption behavior.

This paper examines how business ties with portfolio firms affect the asset management strategies of mutual funds. By exploiting the revelation of mutual fund advisory misconduct as an exogenous shock to these business ties, I find that mutual fund management firms with collapsed trust tend to increase their portfolio weights in client stocks following the misconduct revelation. This shift towards client stocks effectively reduces the likelihood of business partnership termination. Additionally, I find that client stocks underperform compared to non-client stocks and exhibit indifference towards net-selling stocks held by the same mutual fund families. These findings raise concerns about fiduciary duty violations and underscore the need for vigilance in aligning investment decisions with shareholder interests.