Monday, November 25, 2024

New York Fed: Extend-and-Pretend in the U.S. Commercial Real Estate Market

This monster did not go away, we just don't talk about it.

From the Federal Reserve Bank of New York, October 2024:

1 Introduction
In the post-pandemic period, commercial real estate (CRE) has experienced rapidly deteriorating property values, driven by transitory forces such as the significant monetary tightening and structural ones such as the emergence of remote work (Gupta et al., 2023). Given the high leverage typically used in CRE deals, these devaluations often appear to be large enough to deplete the equity of owners and threaten losses for the lenders that hold the underlying mortgages. The discussions about the broader impact of distress in CRE for the financial sector and the economy at large have so far mostly focused on the direct effect of these losses on banks, with little attention directed to how banks manage their distressed CRE exposure.

In this paper, using detailed supervisory data, we document that banks have “extended-and-pretended” their distressed CRE mortgages in the post-pandemic period to delay the recognition of losses.1 Banks with weaker marked-to-market capital—largely due to losses in their securities portfolio since 2022:Q1—have extended the maturity of their impaired CRE mortgages coming due and pretended that such credit provision was not as distressed to avoid further depleting their capital.2 The resulting limited number of loan defaults hindered the reallocation of capital, crowding out the origination of both CRE mortgages and loans to firms. The maturity extensions granted by banks also fueled the volume of CRE mortgages set to mature in the near term—a “maturity wall” with the associated risk of large losses materializing in a short period of time.

Our conceptual framework centers on banks’ incentives to extend the maturity of their existing impaired loans to avoid writing off their capital. This incentive is particularly pronounced from 2022:Q1 onward as rapidly rising rates created large marked-to-market losses on securities held by banks, eroding their economic capital. While having a limited effect on regulatory capital, marked-to-market losses make banks more likely to be monitored by regulators and credit rating agencies and, ultimately, make them vulnerable to runs by uninsured depositors (Drechsler et al., 2024; Haddad et al., 2023).

Our empirical analysis relies on loan-level supervisory data on CRE mortgages by stress tested banks (FR Y-14Q Schedule H.2). CRE mortgages are primarily issued and held on balance sheet by banks (banks hold 50.7% of the $5.8 trillion CRE market as of 2023:Q4) and our granular data captures 26.8% of all CRE mortgages held by banks. We augment this data with supervisory loan-level C&I lending data (FR Y-14Q Schedule H.1), data for Real Estate Investment Trusts (REITs) from Capital IQ and CRSP, and bank-level information (FR Y-9C). We use the latter to measure bank marked-to-market capitalization by (i) adding unrealized gains and losses on all securities to the regulatory capital ratio and (ii) calculating the distance between this measure and the bank-specific regulatory capital threshold.

Our empirical analysis is structured in five parts. First, as motivation, we use raw bank-level data to document that credit risk in the CRE market has substantially increased in the post-pandemic period but banks—weakly capitalized ones in particular—have been sluggish in assessing the associated losses. Specifically, stock returns of REITs that invest in CRE experienced a sizable drop since January 2022, especially in the office segment of this market, likely due to the emergence of remote work. In parallel, banks suffered significant marked-to-market losses as monetary policy tightening reduced the value of their securities holdings. Nevertheless, nonperforming loans and net charge-offs have remained low by historical standards, especially for weakly capitalized banks.

Second, we use our supervisory loan-level data to provide more granular evidence of the extend-and-pretend behavior by banks in their lending to CRE owners. We label a loan as distressed if the current net operating income (NOI) of the underlying property is lower than the NOI at origination.3 We show that undercapitalized banks disproportionately extend the maturities of these distressed loans and understate their default probabilities, leading to fewer realized defaults....

....MUCH MORE (51 page PDF)

And unless the Fed can manipulate rates lower while continuing to give the appearance of fighting inflation, that maturity wall is heading straight at the balance sheets of the banks carrying the paper.

Also, it's not just banks' exposure to CRE. There are ten years of loans to private equity that will have to be rolled at higher rates than they were issued at during the happy time, 2009 - 2022.