A first look at bank loans to real estate funds

Prepared by Marcus Bierich, Pierce Daly, Aoife Horan, Ellen Ryan and Manuela Storz

Published as part of the Macroprudential Bulletin 25, November 2024

This special focus provides an initial granular analysis of bank lending to real estate investment funds (REIFs) in the euro area, while also considering the resulting potential financial stability implications. Financial linkages between banks and non-bank financial intermediaries are becoming increasingly important for policymakers, as they can result in the spillover of financial stability risks to and across these two sectors (see Franceschi et al., 2024). Estimates of bank exposures to commercial real estate (CRE) markets often also exclude bank loans to REIFs. This increases the likelihood of structurally underestimating banks’ exposure to the CRE market, particularly in the light of the rapid growth rate of the REIF sector (see Daly et al., 2023)). This underscores the need to examine the size and characteristics of bank loans to REIFs in the euro area. Our granular analysis applies a novel identification method to loan-level AnaCredit data through which we can identify approximately 80% of the outstanding loans granted to REIFs by euro area banks, while examining their characteristics in granular detail. We complement this analysis with aggregate data which enable us to capture the total size of bank exposures to the euro area REIF sector.^[1]

Lending to REIFs exposes banks to losses in the event of REIF market stress and can create feedback loops between banks and funds, with risks exacerbated by risky lending practices. The current higher interest rate environment raises concerns about debt servicing and rollover risks for banks loans originated during the period of very low interest rates, particularly where these funds relied on short-term or variable rate borrowing. Increasing financing costs could be particularly problematic for those REIFs which face declining profitability levels resulting from the falling demand for CRE assets and from dwindling CRE prices. The funding structure of open-ended REIFs could increase this vulnerability further, raising the potential for outflows from open-end funds and fire sales of real estate. Among the more highly leveraged funds, concerns regarding the seniority of bank debt over fund shares could exacerbate this dynamic further. For banks, rising credit risks may be amplified by risky lending practices, such as unsecured lending, which means a greater potential for losses. As discussed in the main article of this Bulletin, financial stress across parts of the REIF market could also have reputational implications for the wider euro area REIF sector, CRE markets and banks associated with REIFs or broader CRE lending. For example, this stress could occur if fire sales were to aggravate a market downturn or if disorderly fund defaults were to create market uncertainty and risk aversion. Where banks have lent to funds which belong to their own asset management company, fund stress may also trigger reputational spillovers to the wider banking group (Fecht and Wedow, (2014)).

Overall bank lending to REIFs appears to be contained, however, risks could arise from pockets of high financial leverage within this sector. Aggregate data show that loans from euro area banks to REIFs amounted to €121 billion as at end 2023, with most of these loans being granted to German and Italian REIFs. With the exception of REIFs in Luxembourg, debt financing for REIFs is largely sourced from banks and, apart from REIFs in Ireland, it stems largely from within the euro area (Chart 1, panel a). Granular AnaCredit data take us one step further and reveal that 95% of lending from euro area banks to REIFs occurs at the domestic level (Chart 1, panel b). From the bank perspective, total loans to REIFs are equivalent to only 0.4% of total assets, or 9% of total CRE lending. Similarly, aggregate REIF leverage by banks is generally not very high, with an aggregate debt-to-net asset value (NAV) ratio of 16%. This varies substantially across jurisdictions and, in some cases, the total leverage is higher, including where non-banks provide financing for these funds. Moreover, existing analysis examining leverage at fund levels typically identifies concentration in this risk among a tail of highly leveraged REIFs.^[2] So, while this segment of bank lending is unlikely to pose a threat to the overall banking system, risky lending practices could still cause stress to some banks.

Chart 1

The majority of loans from euro area banks to REIFs are granted to funds domiciled in Germany and Italy and they are typically provided by domestic banks

a) Loan liabilities and leverage of euro area REIFs, by country and lender type

(left-hand scale: EUR billions; right-hand scale: percentage of loans/net asset value)

b) Loans from euro area banks typically granted to REIFs in the same jurisdiction

(percentage of total loans to REIFs)

In some countries, funds are exposed to higher financing costs, although funds in these countries are also typically characterised by a close-end structure. Granular AnaCredit data enable us to examine the features of bank loans to funds. First, we find that approximately 43% of bank loans to REIFs are exposed to higher financing costs, either as a result of variable rates or short-term maturity borrowing. 12% of bank loans to REIFs have a remaining maturity of under two years, exposing pockets of these funds to refinancing risk in an environment of higher financing costs and falling asset prices. There is a significant degree of cross-country heterogeneity as a result of differences in the use of variable rate lending across these various countries, which largely mirrors geographic patterns in the wider bank lending sector. As a result, actual increases in financing costs faced by REIFs since the start of monetary tightening vary considerably, from 0.8 percentage points on average in Germany to 4 percentage points in Finland (Chart 2, panel a). Higher financing costs do impact the debt-servicing capacity and investor returns of REIFs, creating the potential for investor outflows if these funds are open-ended. Reassuringly, at the country level, the liquidity risk and the interest rate risk appear to offset each other – with German funds being largely open-ended but typically reliant on fixed rate loans and Italian funds being based on variable rate loans but with closed-end structures (Chart 2, panel b). However, this interplay of risks may still be important for other, smaller countries where REIFs are predominately open-ended.

Chart 2

REIFs in certain jurisdictions are particularly exposed to higher interest rates but these funds also tend to have closed-end structures

a) Share of REIF loans exposed to higher interest rates due to short maturities or variable rates

(left-hand scale: percentage of total loans to REIFs; right-hand scale: percentage points of average change)

b) Share of bank loans to open-ended REIFs vulnerable to higher rates

(x-axis: share of REIF NAV in open-end funds; y-axis: percentage of loans vulnerable to higher rates; bubble size: total NAV of national REIFs)

While aggregate exposures are contained, this loan portfolio appears to be risky in terms of both (a) loss given default and (b) probability of default. AnaCredit data also enable us to consider the type of collateral held by banks against these loans. Almost 18% of bank lending to euro area REIFs is unsecured, in spite of the risky nature of CRE loans (Chart 3, panel a). Such risky lending practices are also reflected in the credit risk metrics, with AnaCredit data revealing that 5.7% of the loans to REIFs are non-performing loans (NPL), compared with 1.4% of banks’ CRE-related loans to firms (Chart 3, panel b). It is also possible to gain insight into the type of assets that REIFs hold through the type of collateral that they post. Some 21% of all real estate collateral posted is backed by residential real estate, with the remainder stemming from the riskier CRE market.

Chart 3

A substantial share of REIF loans is unsecured and the REIF loan book has a significantly higher NPL ratio than the loan book for CRE loans to firms

a) REIF loans by type of collateral

(percentage of total loans to REIFs)

b) NPL ratio for euro area REIF loans versus euro area NFC CRE loans

(percentages of relevant stock of loans which are non-performing)

Overall, financial stability risks stemming from bank loans to REIFs appear to be contained, albeit further analysis is required to assess potential risks from highly leveraged individual REIFs. While bank loans to REIFs account for a small share of the assets of banks and REIFs and REIFs appear to have low leverage levels overall, leverage risks could be concentrated within a few individual REIFs. As shown in the case of Ireland, small cohorts of highly leveraged REIFs can be present and these funds could be at risk of forced deleveraging and sales in the event of market stress. Consequently, future analyses should examine the risks that may stem from pockets of highly leveraged REIFs across various euro area countries, while assessing the appropriate policy response to address these risks in a timely manner.

References

Daly, P., Dekker, L., O’Sullivan, S., Ryan, E. and Wedow, M. (2023), “The growing role of investment funds in euro area real estate markets: risks and policy considerations”, Macroprudential Bulletin, ECB, Frankfurt am Main, April.

Daly, P., Moloney, K. and Myers, S. (2021), “Property funds and the Irish commercial real estate market”, Financial Stability Notes, Vol. 2021, No 1, Central Bank of Ireland, Dublin, February.

Fecht, F. and Wedow, M. (2014), “The dark and the bright side of liquidity risks: Evidence from open-end real estate funds in Germany”, Journal of Financial Intermediation, Vol. 23, Issue 3, July, pp. 376-399.

Franceschi, E., Grodzicki, M., Kagerer, B., Kaufmann, C., Lenoci, F., Mingarelli, L., Pancaro, C. and Senner, R. (2023), “Key linkages between banks and the non-bank financial sector”, Financial Stability Review, Special Feature B, ECB, Frankfurt am Main, May.