Research & policy
Academic papers, official research, regulatory material, patents, and standards are grouped together with their evidence labels intact.
NBER
Sep 07, 2026
Patient limited partners can change venture-fund strategy and startup outcomes
A randomized experiment with Chinese fund managers, combined with evidence from a 2014 reform that opened renminbi venture funds to insurers, links longer-horizon limited partners to longer holding periods and more early-stage investment. The authors report better portfolio-company exits and innovation after insurer entry, while the setting and quasi-experimental design limit direct generalization to other private markets.
- Fund managers in the authors' experiment adjusted stated fund duration and project holding periods to the perceived horizon of their limited-partner base.
- After insurer entry into renminbi venture funds, treated funds shifted toward longer holding periods and earlier-stage investments and showed improved exits and innovation in the authors' estimates.
Why it mattersThe duration and composition of a fund's capital base can shape both investment selection and company-building outcomes. The result is a concrete caution for private-market retailization strategies that introduce shorter-horizon capital into venture portfolios.
NBER
Sep 07, 2026
Investor exit may concentrate risk among more optimistic market participants
Using more than two decades of investor surveys and option-implied benchmarks, the authors find that people who step out of markets are generally more pessimistic about returns and crash risk than active pricing participants. They argue that this participation-selection channel helps explain why average survey expectations can be negatively related to later realized returns.
- The authors report that non-participating investors are generally more pessimistic about returns and perceive greater crash risk than the pricing population.
- Stocks with greater estimated sensitivity to non-marginal beliefs earned lower returns in the study, especially where disagreement was high.
Why it mattersSentiment surveys may mix the beliefs of price-setting investors with those of people who are not currently taking risk. Portfolio research and market indicators should account for participation before interpreting an average belief measure as the market's marginal view.
NBER
Sep 07, 2026
Network-Q model links product-market rivalry to investment and capital allocation
A new dynamic investment model generalizes Tobin's Q to a network of product-market spillovers across heterogeneous, multi-product firms. Applied to U.S. public companies, the authors report that competition materially shapes aggregate investment and capital allocation, while mergers since 1995 are associated with only a modest aggregate investment decline but heterogeneous markup increases.
- The authors report that product-market competition is an important driver of aggregate investment and capital allocation in their U.S. public-company application.
- Their estimates associate mergers since 1995 with a modest decline in aggregate capital formation among merging firms and heterogeneous firm-level markup increases.
Why it mattersThe framework offers investors and competition analysts a way to connect firm-level cost-of-capital shocks, rival responses, concentration, and investment rather than treating each company in isolation. The empirical conclusions remain dependent on the model and identification choices described by the authors.
NBER
Sep 07, 2026
Loan pricing can double as an internal bank incentive mechanism
A theoretical banking model shows that loan rates can affect both borrower repayment incentives and how informative repayment is about hidden effort inside the bank. The author finds that restrictions on paying employees after defaults can push banks toward different loan pricing and greater risk-taking under some capital-rule designs; these are model results, not measured effects in operating banks.
- The model jointly determines bank wage contracts and loan-rate menus because repayment outcomes reveal information about unobservable effort.
- In the model, restricting failure-contingent pay can change loan pricing and, under a countercyclical capital requirement, produce alternating risk-taking regimes.
Why it mattersBank capital and compensation rules can interact with product pricing in ways that standard credit-risk models miss. The paper identifies an incentive channel worth testing before assuming a prudential rule will mechanically reduce risk.