Research seminars

Nonbank Lending

Finance

Speaker: Sergey Chernenko
Fisher College of Business

15 March 2018 - T004 - From 2:00 pm to 3:15 pm

Download

We provide novel systematic evidence on the terms of direct lending by nonbank financial institutions. Analyzing hand-collected data for a random sample of publicly-traded middle-market firms during the 2010-2015 period, we find that lending from nonbank financial institutions is substantial, with 30% of all loans being extended by nonbanks. Firms are more likely to borrow from a nonbank lender if local banks are poorly capitalized and less concentrated. Nonbank borrowers are smaller, riskier, and significantly more likely to have negative EBITDA. Nonbank lenders are less likely to include financial covenants in their loans, but appear to engage in substantial ex-ante screening: origination of nonbank loans is associated with larger positive announcement returns while ex-post performance is not distinguishable from bank loans. We also find that nonbank borrowers pay about 200 basis points higher interest rates than bank borrowers do. Using fuzzy regression discontinuity design and matching techniques generates similar results. Overall, our results provide evidence of market segmentation in the commercial loan market, where bank and nonbank lenders utilize different lending technologies and cater to different types of borrowers.

Finance

Speaker: Matthieu Bouvard
Desautels Faculty of Management

14 June 2018 - From 2:00 pm to 3:15 pm


Finance

Speaker: Mikhail Simutin
Rotman School of Management

7 June 2018 - From 2:00 pm to 3:15 pm


Disclosure, Competition, and Learning from Asset Prices

Finance

Speaker: Liyan Yang
Rotman School of Management

31 May 2018 - From 2:00 pm to 3:15 pm

Download

This paper studies the classic information-sharing problem in a duopoly setting in which firms learn information from a financial market. By disclosing information, a firm incurs a proprietary cost of losing competitive advantage to its rival firm but benefits from learning from a more informative asset market. Firms' disclosure decisions can exhibit strategic complementarity, which is strong enough to support both a disclosure equilibrium and a nondisclosure equilibrium. Allowing minimal learning from asset prices dramatically changes firms' disclosure behaviors: without learning from prices, firms do not disclose at all; but with minimal learning from prices, firms can almost fully disclose their information. Learning from asset prices benefits firms, consumers, and liquidity traders, but harms financial speculators.

Alpha Decay*

Finance

Speaker: Anton Lines
Columbia Business School

24 May 2018 - T020 - From 2:00 pm to 3:15 pm

Download

Using a novel sample of professional asset managers, we document positive incremental alpha on newly purchased stocks that decays over twelve months. While managers are successful forecasters at these short-to-medium horizons, their average holding period is substantially longer (2.2 years). Both slow alpha decay and the horizon mismatch can be explained by strategic trading behavior. Managers accumulate positions gradually and unwind gradually once the alpha has run out; they trade more aggressively when the number of competitors and/or correlation among information signals is high, and do not increase trade size after unexpected capital flows. Alphas are lower when competition/correlation increases.

What is the Expected Return on a Stock?

Finance

Speaker: Ian Martin
LSE

17 May 2018 - From 2:00 pm to 3:15 pm


We derive a formula that expresses the expected return on a stock in terms of the risk-neutral variance of the market and the stock’s excess risk-neutral variance relative to the average stock. These quantities can be computed fromindex and stock option prices; the formula has no free parameters. We run panel regressions of realized stock returns onto risk-neutral variances, and find that the theory performs well at 6-month, 1-year, and 2-year forecasting horizons. The formula drives out beta, size, book-to-market and momentum, and outperforms a range of competitors in forecasting stock returns out of sample. Our results suggest that there is considerably more variation in expected returns, both over time and across stocks, than has previously been acknowledged.


JavaScriptSettings