Can the Textual Tone in REIT Financial Statements Improve the Information Environment for Commercial Real Estate Investors? An Investigation

Carstens, Riëtte; Freybote, Julia

doi:10.1080/08965803.2021.1985923

Cited by 6 publications

(2 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By investigating the language in the Management Discussion and Analysis of REITs, Koelbl (2020) concluded that higher levels of pessimistic (optimistic) language predicts lower (higher) future firm performance. Likewise, Carstens and Freybote (2021) found that the tone in REIT financial statements can predict total returns in the subsequent quarter, given periods of poorly performing real estate markets. Only a few studies have investigated the dictionary-based approach concerning residential real estate markets.…”

Section: Literature Review and Hypothesis Developmentmentioning

confidence: 97%

News coverage vs sentiment: evaluating German residential real estate markets

Ploessl

Just

2022

IJHMA

View full text Add to dashboard Cite

Purpose To investigate whether additional information of the permanent news flow, especially reporting intensity, can help to increase transparency in housing markets, this study aims to examine the relationship between news coverage or news sentiment and residential real estate prices in Germany at a regional level. Design/methodology/approach Using methods in the field of natural language processing, in particular word embeddings and dictionary-based sentiment analyses, the authors derive five different sentiment measures from almost 320,000 news articles of two professional German real estate news providers. These sentiment indicators are used as covariates in a first difference fixed effects regression to investigate the relationship between news coverage or news sentiment and residential real estate prices. Findings The empirical results suggest that the ascertained news-based indicators have a significant positive relationship with residential real estate prices. It appears that the combination of news coverage and news sentiment proves to be a reliable indicator. Furthermore, the extracted sentiment measures lead residential real estate prices up to two quarters. Finally, the explanatory power increases when regressing on prices for condominiums compared with houses, implying that the indicators may rather reflect investor sentiment. Originality/value To the best of the authors’ knowledge, this is the first paper to extract both the news coverage and news sentiment from real estate-related news for regional German housing markets. The approach presented in this study to quantify additional qualitative data from texts is replicable and can be applied to many further research areas on real estate topics.

show abstract

Section: Literature Review and Hypothesis Developmentmentioning

confidence: 97%

News coverage vs sentiment: evaluating German residential real estate markets

Ploessl

Just

2022

IJHMA

View full text Add to dashboard Cite

show abstract

“…ML has been applied to predict misstatements in financial statements (Bertomeu et al, 2021), banking crises from excessive credit growth and aggregate leverage (Alessi & Detken, 2018), the percentage of assets lost if a borrower defaults or the loss given a default (Kellner et al, 2022) and the destabilising effects of systematic risk using artificial intelligence (Danielsson et al, 2022). There is scanty use of ML in empirical research on REITs; for example, Carstens and Freybote (2021) suggest that a forward‐looking tone, which refers to a tone not explained by the past performance of a REIT, predicts total returns in the next quarter of REIT financial statements when real estate markets perform poorly. Their findings indicate that tone analysis can enhance the information environment for investors in commercial real estate.…”

Section: Review Of Relevant Literaturementioning

confidence: 99%

Transitory and permanent shock transmissions between real estate investment trusts and other assets: Evidence from time‐frequency decomposition and machine learning

Ngene

Wang

2023

Accounting & Finance

View full text Add to dashboard Cite

We evaluate asset returns and volatility connectedness using the time‐frequency connectedness model and machine learning approaches. Using 48 years of monthly indices of equity real estate investment trusts (EREITs), mortgage real estate investment trusts (MREITs), stocks, commodities, and bonds, we find that shocks to EREIT, and MREIT returns have a transitory impact on other assets. However, asset volatility connectedness among assets occurs at lower frequencies as markets slowly process pricing information. Therefore, shocks to EREIT and MREIT decay gradually and spill over to the other three assets for long periods. We also find that the intensity of the time‐frequency connectedness of returns and volatility varies with business cycles and significant domestic and global non‐recession events. We attribute the dominance of real estate investment trusts (REITs) in destabilising the financial system through long‐term volatility transmission to the heavy linkage of REITs to highly illiquid underlying direct real estate markets and high REITs leverage, making them more sensitive to real estate fundamentals, monetary shocks and macroeconomic risks than stocks and bonds. The result of the algorithm‐based machine learning model broadly supports the dominance of EREITs in the transmission of returns and volatility and shows that commodities have more explanatory power in the transmission of volatility than in returns. The empirical findings have implications for strategic and tactical asset allocation and policy design for market stability.

show abstract