Distinguish fixed-ranking rationing from mixed ranked and random rationing

Determine whether the 2020 freeze in nonbank Ginnie Mae buyout exercise reflected rationing along a fixed borrower-ranking index or a mixture of ranked and random rationing.

Background

The model predicts that a liquidity-constrained intermediary preserves its ranking of opportunities while reducing the share funded. The empirical results are consistent with that prediction in several periods, especially the release period, but the freeze-period estimates alone do not statistically distinguish fixed-ranking rationing from a combination of ranking-based and random selection.

References

What the evidence does not separate is stated as well: the 2022 fall in nonbank exercise mixes a value shock with higher funding costs, and the 2020 freeze alone does not separate a fixed ranking from a mixture of ranked and random rationing.

— The Financing of Debt Relief: Intermediary Liquidity and the Value of Borrower Information  (2610.12875 - Bandyopadhyay, 9 Oct 2026) in Section 4.9, “Alternative explanations”; Section 4.10, “What the evidence establishes, and what it does not”