Reassessing Residential REITs: Performance and Resilience After COVID-19

Abstract

This study analyzes the monthly returns of residential real estate investment trusts (REITs) from 2010 through 2025, assessing how these investments performed both during and following the COVID-19 pandemic. Using Sharpe, Sortino, and Omega ratios and multi-factor asset pricing models, the study finds that residential REITs delivered stable risk-adjusted returns and improved downside protection relative to broad indices. Value-at-risk metrics confirmed their defensive nature, and portfolio optimization demonstrated diversification and volatility reduction when including residential REITs. Overall, residential REITs were shown to preserve capital and improve portfolio efficiency during market disruptions.

1. Introduction

Residential real estate investment trusts (REITs) serve as a vital institutional conduit connecting capital markets with the housing sector. Functioning as a pooled investment vehicle, residential REITs offer investors an indirect, professionally managed, and easily tradable method of gaining exposure to multifamily, single-family, and mixed-use rental housing. Both theoretically and practically, residential REITs have historically been perceived as defensive assets, providing diversified portfolio investors with a means to access the rental housing market in a way that complements conventional equity investments (Okoro & Ayaba, 2023; Zietz et al., 2003).

Historically, the primary appeal of REITs for both institutional and retail investors has been their low correlation with the broader stock market, their stable cash flows, and their ability to protect against inflation (Okoro & Ayaba, 2023; Zietz et al., 2003). Furthermore, their reliable income and capital, together with their responsiveness to demographic and economic trends, make them significant for long-term asset allocation decisions.

The COVID-19 pandemic significantly altered the housing landscape. Lockdowns, the rise of remote work, eviction bans, and government aid all played a role in reshaping demand for various housing options and in different areas. During the lockdowns of 2020, housing markets typically cooled off. Both tenants and workers relocated from urban centers to suburban and Sun Belt areas. By 2021, as vaccine distribution expanded and office attendance policies were reinstated, investors observed renewed growth in rental rates and a contraction in supply.

Residential REITs, encompassing both public and private entities, demonstrated superior performance compared to their office and retail counterparts, which faced challenges stemming from structural issues or region-specific disruptions. Residential REITs exhibited stable occupancy rates, consistent rent collections, and a more rapid recovery in value following the pandemic (Victor et al., 2023; Malhotra, 2023). Consequently, it is crucial to evaluate the performance of residential REITs across diverse market conditions, particularly during periods of economic strain. A critical consideration for long-term investors is the ability of these REITs to maintain their defensive characteristics and diversification advantages amidst elevated inflation, mortgage rates, and housing costs.

Literature has extensively studied REITs. Previous studies have examined REIT performance, risk, and factor exposures during various market conditions. Numerous studies utilize aggregated REIT samples that encompass residential, office, retail, industrial, and various other property categories (Chan et al. (1990); Chen and Peiser (1999); Han and Liang (1995); Ling and Naranjo (1999); Coën and Guardiola (2025); Giacomini et al. (2016); Okoro and Ayaba (2023)). Pooling of various categories of REITs is appropriate for analyzing returns at the index level but assumes that cash-flow dynamics and risk factors are homogenous across REIT sectors. Residential REITs have distinct lease terms, tenant turnover, income stability, regulation, and inflation expectations that drive return behavior and downside risk compared to other property types. By pooling residential REITs with property types that have different structural features, we lose information about their individual risk dynamics.

In addition, recent REIT studies increasingly focus on short-horizon analyses centered on the COVID-19 period. Although these windows capture immediate market reactions to an unprecedented shock, they are less suited for assessing persistent risk characteristics, resilience, or alpha behavior. Short samples risk conflating transitory crisis-driven volatility with longer-run performance dynamics, particularly in residential real estate, where adjustment mechanisms operate over extended horizons.

This paper seeks to address these gaps in the literature and provides a detailed analysis of the performance, risk characteristics, and resilience of residential REITs (equity and mortgage) from 2010 to 2025, an extended time horizon. For the study’s purposes, resilience refers to the ability to absorb macroeconomic shocks, adjust risk exposure across economic states, and maintain performance characteristics, such as controlled downside risk and stable factor-adjusted returns, over extended periods. The paper’s main goal is to determine whether residential REITs provide attractive risk-adjusted returns, alpha, and diversification benefits relative to the broad REIT and equity markets, or whether structural trends have reduced their value proposition and long-term appeal to investors. This study uses risk-adjusted performance measures (Sharpe, Sortino, and Omega ratios), multi-factor and conditional alpha models, and downside risk measures (Value-at-Risk (VaR) and Conditional Value-at-Risk (CVaR)) to decompose the performance of residential REITs across different market conditions.1

Specifically, this study analyzes residential REITs’ long-term performance before and after the pandemic, estimates multi-factor and conditional alphas with various asset pricing models, assesses downside risk and CVaR for portfolio contribution, and constructs three optimized portfolios (Global Minimum Variance (GMV), Tangency, and Risk Parity) to inform asset allocation decisions. Ali et al. (2024), for instance, use a similar approach that combines VaR, CVaR, and mean–variance optimization to analyze how diversification and risk reduction work across different asset classes. Building on this research, the current study moves step by step: first evaluating performance adjusted for risk, then analyzing tail risks, and finally optimizing portfolios, making sure statistical results lead to meaningful economic insights for portfolio management.

The remainder of the paper is organized as follows. Section 2 reviews prior literature on REIT performance, empha-sizing property-type differences and pandemic-era findings. Section 3 describes the data. Section 4 outlines the em-pirical models and performance measures. Section 5 introduces the robustness framework and conditional specifi-cations. Section 6 presents the empirical results, including correlation analysis, risk-adjusted performance, fac-tor-based alpha estimates, downside risk behavior, and portfolio optimization outcomes. Section 7 concludes with a summary of findings and directions for future research.

2. Previous Studies

In the past three decades, scholarly investigation into REITs has advanced considerably, emphasizing distinctions based on property type, risk-adjusted returns, cash flow analyses, and sector-specific developments. Early work by Gyourko and Keim (1992) argued that traditional equity indicators do not accurately reflect real estate markets, leading to the use of REIT-specific metrics like Funds from Operations (FFO), Adjusted Funds from Operations (AFFO), and Net Asset Value (NAV). Later studies confirmed these measures are more suitable, showing that REITs differ from general equities in risk and return and only demonstrate their benefits, such as competitive returns and inflation protection, when evaluated with real estate-focused indicators rather than standard accounting figures.

With the expansion of REIT markets, researchers increasingly examine property-type differences as a primary method of analysis. After the global financial crisis, and especially during COVID-19, these distinctions became increasingly important. Milcheva (2021) found that REITs moved more closely with general stock markets during the pandemic, which reduced their ability to diversify investments, but this effect varied depending on the type of property. Akinsomi (2021) showed that retail and office REITs were hit hardest during the pandemic, while residential and data center REITs remained relatively stable given their steady rent payments and better liquidity. In a study assessing how uncertainty measures—such as implied volatility—affect REIT sectors from 1990 to 2020 (which includes the global financial crisis), Demiralay and Kilincarslan (2024) found that residential REITs were the most resilient compared to other sectors. Together, these findings demonstrate that residential REITs often behave very differently from other sectors in times of economic downturn.

The ability of residential REITs to quickly adjust rents in response to inflation is another key advantage, as their short-term leases enhance inflation-hedging capabilities (Newell & Fischer, 2009; Malhotra, 2023). Cai and Xu (2022) also observed that, during the COVID-19 pandemic, residential REIT vacancy rates remained stable as populations moved from urban centers to suburbs and smaller cities due to pandemic concerns and remote work. During the third quarter of 2020, multifamily residential REITs experienced steady growth despite challenges in other sectors. They also found that the net impact of COVID-19 on residential REITs was not negative. Additionally, demographic shifts, such as migration to suburban areas and the rise of hybrid work models, supported residential REIT performance during the pandemic (Malhotra & Malhotra, 2024; Beracha et al., 2019).

Further research underscores the role of behavioral factors and capital flows in REIT performance. For instance, institutional ownership patterns influence REIT valuation and volatility (Lantushenko & Nelling, 2022), and mutual fund flows can drive sector prices independently of fundamentals (Ling & Naranjo, 2006). In emerging markets, liquidity issues, regulation, and sentiment cause significant variation in returns (Debata et al., 2018). REITs also improve portfolio diversification and reduce risk, particularly in volatile African markets (Ng et al., 2021; Okoro & Ayaba, 2023).

Internationally, residential REITs have become more prominent within global investment portfolios. Newell and Fischer (2009) initially found that residential REITs offered limited diversification before the 2008 financial crisis, but they have since become stabilizing assets. Recent studies, such as those by Newell and Marzuki (2023), suggest that institutional investors are increasingly favoring residential REITs due to their ability to provide steady income streams. Research from the Asia-Pacific region (Ooi & Liow, 2004; Newell & Marzuki, 2016) indicates that residential REITs experience less severe downturns and are less sensitive to interest rate changes compared to other property sectors.

The flexibility of residential REITs, particularly through shorter lease periods, has made their managers more adaptable in comparison to those in other sectors (Buttimer et al., 2012). However, recent studies (Malhotra et al., 2024) raise concerns about the increasing correlation between REITs and equity indices during macroeconomic uncertainty, which could reduce their effectiveness as portfolio diversifiers and hedging instruments.

To assess empirically whether residential REITs provide such defensive characteristics, this study’s approach is grounded in established asset pricing and real estate economic theories. These theories highlight the importance of sectoral differences and the dynamic nature of risk exposure. Simultaneously, sector-specific analyses reveal considerable divergence in cash-flow stability, operational effectiveness, and risk profiles among different REIT categories; for instance, residential REITs tend to demonstrate more consistent income streams compared to more cyclical sectors like hospitality.

Consequently, real estate performance is inherently multifaceted, encompassing operational underpinnings, capital arrangements, and broader economic circumstances, alongside return-focused assessments. These theoretical perspectives suggest that performance metrics devoid of conditions and analyses with short timeframes might mask economically significant differences in risk exposure, adverse performance, and portfolio effectiveness. This research uses empirical methods, including conditional factor models, downside-risk analysis, and portfolio optimization, to evaluate residential REIT performance in different market conditions.

Drawing on data from 2010 to 2025, this study employs various performance ratios, factor models, downside-risk metrics, and mean–variance optimization (including GMV, Tangency, and Risk Parity portfolios) under varying market conditions, offering a more robust analysis of their role in investment portfolios. This approach enables a robust assessment of residential REIT performance in different market conditions while addressing prior methodological gaps. The next section describes the data, variables, and empirical models used.

3. Data

The sample consists of the monthly total returns for U.S. residential REITs obtained from Morningstar Direct. Morningstar Direct is used due to its comprehensive coverage of U.S. REIT total returns and constituent-level information, and monthly frequency is appropriate for REITs due to appraisal smoothing, dividend relevance, and alignment with macroeconomic conditioning variables. The number of residential REITs in the sample is 66 at the beginning of the study period in January 2010 and increases to 69 by June 2025, reflecting new listings and corporate restructuring over time. All REITs are included contemporaneously as they enter or exit the universe, ensuring comprehensive coverage of the residential REIT sector. Returns are equally weighted throughout the analysis to avoid dominance by the largest REITs and to better capture sector-level performance and risk characteristics rather than firm-size effects.

The sample also includes benchmark data from Morningstar Direct. The Dow Jones Residential REITs USD Index serves as the sector-specific benchmark, capturing REITs primarily invested in residential properties such as multifamily and single-family housing. The Dow Jones Composite All REIT Total Return (TR) USD Index represents the broader U.S. REIT market across multiple property sectors, while the Russell 3000 Index provides a broad U.S. equity benchmark. Together, these benchmarks allow residential REIT performance to be evaluated both within the real estate sector and relative to the broader equity market.

The complete sample encompasses data from January 2010 through June 2025, with a subset analysis conducted for March 2020 through June 2025 to capture dynamics during the pandemic period. The subset analysis begins in March 2020, coinciding with the COVID-19 outbreak and lockdowns. Factor data for market, size, value, profitability, investment, and momentum factors are obtained from Kenneth French’s Data Library.2

Table 1. Summary statistics of monthly rates of returns for residential real estate investment trusts and benchmark indices for the period January 2010 to June 2025.

	Residential REITs	Dow Jones Residential REITs USD Index	Dow Jones Composite All REIT TR USD Index	Russell 3000 Index
January 2010 to June 2025
Average Monthly Return (%)	0.94	0.99	0.83	1.16
Standard Deviation (%)	3.95	5.16	4.88	4.35
Return to risk ratio (%)	0.24	0.19	0.17	0.27
March 2020 to June 2025 (post-COVID-19 period)
Average Monthly Return (%)	1.05	0.46	0.53	1.38
Standard Deviation (%)	5.19	5.94	6.09	5.28
Return to risk ratio (%)	0.20	0.08	0.09	0.26

presents descriptive statistics summarizing the monthly returns of residential REITs, the Dow Jones Residential REITs Index, the Dow Jones Composite All REIT TR Index, and the Russell 3000 Index over two distinct periods: January 2010 to June 2025 for the entire sample and March 2020 to June 2025 for the period after the COVID-19 outbreak.3

According to Table 1, residential REITs achieved an average monthly return of 0.94% for the period of January 2010 to June 2025, comparable to the Dow Jones Residential REITs Index and just above the overall REIT market. Their volatility was lower at 3.95% compared to over 5% for composite REIT benchmarks, indicating more stable returns. The return-to-risk ratio of residential REITs (0.24% per unit of risk) surpasses other REIT categories and nearly matches the efficiency shown by the Russell 3000 Index.

After the COVID-19 outbreak, residential REITs averaged a 1.05% monthly return with volatility rising to 5.19%. Despite higher market uncertainty, they showed resilience by sustaining a strong risk-adjusted performance and a 0.20 return-to-risk ratio. The other REIT indices’ returns-to-risk ratios declined by a larger percentage after 2020.

4. Model

We assess residential REITs using multi-factor asset-pricing models and key performance metrics, offering a clear view of their returns and risks. This approach highlights both specific asset behavior and general market effects, showing how residential REITs respond to systemic risks and market trends.

4.1. Risk-Adjusted Performance Metrics

This study employs three widely recognized risk-adjusted performance ratios (Sharpe, Sortino, and Omega) to evaluate how effectively residential REITs deliver returns to investors in proportion to the risks they assume.

The Sharpe ratio (Sharpe, 1964) is the most widely used risk-adjusted performance measure, which calculates the excess return an investment generates per unit of total volatility. This provides useful information on whether residential REITs offer adequate compensation for the risk they pose to investors. A high Sharpe ratio indicates that residential REITs can deliver high returns compared to their overall variation, which is a highly desirable characteristic for income-seeking investors in terms of generating steady performance over the long term.

The Sortino ratio (Sortino & van der Meer, 1991) is another risk-adjusted performance measure; however, it accounts for only downside risk. Instead of considering all volatility in performance, this measure specifically penalizes negative deviations while leaving positive ones unaccounted for. This makes the Sortino ratio an especially relevant risk-adjusted performance measure for residential REITs, as it signals whether they are effective in managing downside risk. During periods of intense market fluctuations, like the pandemic, many investors focused mainly on protecting their capital rather than pursuing growth.

Finally, the Omega ratio (Keating & Shadwick, 2002) is a performance measure that gives an overall picture of an investment’s performance by comparing the probability-weighted returns above a threshold to the losses below it, with the risk-free rate typically used as a threshold value. Since it considers both the size of returns and their likelihood, the Omega ratio can be used to evaluate the asymmetry in the return distribution of residential REITs. A high Omega ratio indicates a higher probability of positive outcomes, which is a desirable characteristic for residential REITs for income-seeking investors, as it indicates a lower likelihood of losses or downside risk.

These three measures present a nuanced picture of how residential REITs balance return generation with risk control, capturing not only their volatility profile but also their ability to protect investors from adverse market movements.

4.2. Multi-Factor Models for Alpha Estimation for Residential REITs

This study uses the Fama–French five-factor model to evaluate whether residential REITs yield excess returns that cannot be accounted for by traditional risk factors. Fama and French (2015) expanded the original three-factor Fama and French (1993) model by including profitability and investment style. These two factors are especially important for real estate enterprises, since long-term performance is driven by operational efficiency and capital allocation. For this study, factor regressions employ U.S. Fama–French factors, consistent with the domestic focus of the REIT universe and standard practice in the REIT literature.

The conventional three-factor model encapsulates the impact of market risk, business size, and value orientation as the principal determinants of anticipated returns. This concept is beneficial since residential REITs generally act like small-cap, value-oriented companies with regular income and physical assets. It can be expressed as:

$$R_{i,t}-R_{f,t}={\alpha }_i+{\beta }_1\times \left(R_{m,t}-R_{f,t}\right)+{\beta }_2\times {SMB}_t+{\beta }_3\times {HML}_t+{\epsilon }_{i,t}$$

(1)

where $R_{i,t}$ = percentage return for firm i in month t.

$R_{f,t}$ = yield on the U.S. Treasury bill in month t.

$R_{m,t}$ = return on the CRSP value-weighted index for month t.

$\left(R_{m,t}-R_{f,t}\right)=$ market risk factor represents the excess return of the overall market and accounts for the general risk associated with investing in the stock market.

${SMB}_t$ (Small Minus Big (SMB)) = realization on capitalization factors (small-cap return minus large-cap return) for month t.

${HML}_t$ (High Minus Low (HML)) = realization on value factor (value return minus growth return) for month t.

ε_i,t = an error term.

The three-factor model alone does not fully capture the complexity of residential REIT performance. Real estate companies exhibit significant variation in their approaches to leverage, reinvestment intensity, and operational efficiency—key factors influencing profitability and risk that are not addressed within the initial framework. Residential REITs reinvest large portions of earnings into property acquisition and development, and their profitability depends heavily on cost management and rental growth. These characteristics warrant a more comprehensive model that integrates firm-level behavior alongside traditional market factors.

For this reason, the study adopts the Fama–French five-factor model, which extends the earlier specification by adding profitability and investment factors. It can be expressed as:

$$R_{i,t}-R_{f,t}={\alpha }_i+{\beta }_1\times \left(R_{m,t}-R_{f,t}\right)+{\beta }_2\times {SMB}_t+{\beta }_3\times {HML}_t+{\beta }_4\times {RMW}_t+{\beta }_5\times {CMA}_t+{\epsilon }_{i,t}$$

(2)

where

${RMW}_t$ (Robust Minus Weak (RMW)) = factor that measures the return premium for firms with strong profitability relative to those with weak profits for month t.

${CMA}_t$ (Conservative Minus Aggressive (CMA)) = factor that captures differences in investment aggressiveness for month t.

Within residential REITs, a positive RMW coefficient suggests that trusts with higher profitability, reflected in superior margins or stronger rent collection, tend to outperform their less efficient counterparts. A positive CMA suggests that REITs pursuing disciplined, conservative growth strategies deliver superior returns relative to those engaged in rapid or highly leveraged expansion.

Using the five-factor model offers several advantages. It acknowledges that real estate returns depend not only on market-wide risk but also on the internal financial and operational decisions that define firm quality (Malhotra, 2024). It recognizes that profitability and investment behavior, proxied by RMW and CMA, are critical to explaining return variation across REITs. It also improves the accuracy of alpha estimation by reducing unexplained residuals that can distort performance interpretation under simpler models.

The decision to employ this model is further supported by recent literature. Studies such as Malhotra and Malhotra (2024) demonstrated that REIT returns are strongly linked to firm-level efficiency and capital management, consistent with the mechanisms captured by the RMW and CMA factors. Moreover, as the post-pandemic period has shown, the performance of residential REITs reflects not only exposure to market conditions but also differences in balance-sheet strength, geographic diversification, and reinvestment prudence.4

The Fama–French five-factor model provides a robust and theoretically grounded framework for understanding the drivers of residential REIT performance. It allows this study to separate returns arising from broad market forces from those generated through managerial skill, profitability, and disciplined capital management, offering a clearer view of whether residential REITs truly deliver persistent, risk-adjusted alpha.

5. Robustness Check

To add robustness to the empirical results, the analysis augments the traditional unconditional factor models with conditional alphas, allowing the risk exposures to vary with time or economic conditions. This approach recognizes that the operating environment for residential REITs is not static. For example, during the post-COVID recovery, an increasing interest rate cycle and shifts in housing demand likely affect how residential REITs behave relative to systematic risk factors. The inclusion of conditional specifications in conjunction with market-timing and selectivity measures can help identify whether performance is simply a function of broad risk factor exposures or whether there is evidence of managerial skill across economic regimes.

5.1. Conditional Alpha

Standard approaches for gauging REIT performance are unconditional factor models with time-invariant betas and a constant risk premium. This modeling framework, however, is particularly restrictive for real estate securities whose flows and valuations are closely tied to the macroeconomy and the state of the housing market. Ferson and Schadt (1996) demonstrated that time variation in risk exposures is generally ignored at the portfolio level to the detriment of performance measurement, because the composition of a portfolio and the exposures of its securities change endogenously with shifts in economic conditions. In the case of residential REITs, changes in interest rates, the term-structure slope, labor-market conditions, and inflation are expected to alter portfolio risk exposures, so performance evaluation is unlikely to be accurate when risk exposures are assumed to be time invariant.

Ferson and Schadt (1996) presented a conditional performance measure that includes lagged public information and macro-financial predictors of stock returns, which also allow factor loadings and expected returns to vary across time with the state of the economy. Their key insight is that the level of portfolio risk is not constant but varies with the arrival of new information. In a standard model with a time-invariant beta, changes in risk exposures are confounded with other sources of expected and unexpected returns. If a fund’s beta increases in response to new information, a traditional estimate of the alpha will capture this change in risk through a spurious increase in performance, which is a return difference that should be attributed to variations in risk exposures. To address this problem of identification, when betas are allowed to vary as functions of observable information, performance evaluation becomes more responsive to the economic forces that differentially impact real estate sectors.

Conditioning on this logic, we augment the standard multi-factor regression model for residential REITs by including lagged macro-financial instruments in interactions with the market excess return. We add interaction terms between the market risk premium (mkt − rf) and the slope of the yield curve, unemployment, Quality Spread Differential (QSD), and inflation. The slope of the yield curve is measured by the difference between the yield on 30-year and 3-month Treasury securities. Quality Spread Differential (QSD) is measured by the difference in yields on Aaa- and Baa-rated corporate bonds. The new specification allows for market betas for residential REITs to vary with key state variables while still identifying a conditional alpha over and above dynamic factor exposures. This type of conditional modeling is particularly appropriate for residential REITs because their risk characteristics adjust with monetary policy, labor-market slack, funding conditions, and inflation, which jointly determine rent demand, occupancy, and property values over the cycle.

The following equation presents the resulting conditional models, where $z_{J,t-1}$ represents the demeaned value of the unconditional elements:

$$R_{i,t}-R_{f,t}={\alpha }_i+{\beta }_i\left(R_{m,t}-R_{f,t}\right)+\mathsf{\delta }\times \left\{z_{t-1}\times \left(R_{m,t}-R_{f,t}\right)\right\}+{\beta }_2\times {SMB}_t+{\beta }_3\times {HML}_t+{\beta }_4\times {RMW}_t+{\beta }_5\times {CMA}_t+{\epsilon }_{i,t}$$

(3)

where

z_(t−1) = one or more lagged information variables (e.g., dividend yield, term spread, lagged market return).

The multiplicative interaction between z_(t−1) and the market excess return allows factor loadings to vary over time creates an interaction term (or terms) between each conditioning variable and the market excess return.

δ = a vector of coefficients for the interaction terms. This approach tests whether residential REITs’ alpha remains after accounting for changing factor exposures over time.

5.2. Downside Risk Analysis of Monthly Returns of Residential REITs

VaR and CVaR (also called Expected Shortfall (ES)) are important metrics for residential REITs to evaluate the risk they are taking in the event of a big market shock. In an investing climate characterized by post-pandemic uncertainty, increasing interest rates, and issues in housing affordability, these downside risk metrics assist investors in assessing the susceptibility of residential REITs to significant losses amid financial strain or economic recessions.

VaR identifies the most money a residential REIT portfolio could lose over a certain amount of time, assuming the market is regular. CVaR extends this analysis by providing the average of losses that exceed the VaR threshold. This gives a more complete picture of tail risk. These two measures give investors insight into the probability and magnitude of extreme losses, indicating sector stability. By measuring downside risk, VaR and CVaR help investors determine if residential REITs can be a safe part of a diversified portfolio during times of high volatility.

6. Empirical Analysis

The empirical analysis begins with an examination of the correlations among monthly returns of residential REITs, the Dow Jones Residential REITs Index, the Dow Jones Composite All REIT TR USD Index, and the Russell 3000 Index for the period January 2010 to June 2025 and for the period March 2020 to June 2025. This comparison provides insight into how closely residential REITs move with the broader REIT market and general equity indices, offering an initial indication of their diversification potential and sensitivity to macroeconomic factors.

6.1. Correlation Analysis

Table 2. Correlation among monthly returns of residential REITs, the Dow Jones Residential REITs Index, the Dow Jones Composite All REIT TR USD Index, and the Russell 3000 Index. Analysis is based on monthly returns from January 2010 to June 2025 and from March 2020 to June 2025.

	Residential REITs	Dow Jones Residential REITs Index	Dow Jones Composite All REIT TR USD Index	Russell 3000 Index
January 2010 to June 2025
Residential REITs	1
Dow Jones Residential REITs Index	0.70	1
Dow Jones Composite All REIT TR USD Index	0.74	0.89	1
Russell 3000 Index	0.69	0.60	0.76	1
March 2020 to June 2025
Residential REITs	1
Dow Jones Residential REITs Index	0.82	1
Dow Jones Composite All REIT TR USD Index	0.80	0.90	1
Russell 3000 Index	0.76	0.76	0.86	1

presents the correlation.

Across the full sample, residential REITs show a strong correlation (0.70) with the Dow Jones Residential REITs Index, confirming that their performance is largely aligned with other housing-focused REITs that share similar property and income characteristics. Residential REITs’ correlation with the broader REIT market (0.74) and with the Russell 3000 Index (0.69) indicates that while residential REITs are connected to overall market movements, they retain a degree of independence that can enhance portfolio diversification. In essence, they behave like a bridge between the stability of real estate income and the cyclical nature of equities.

The correlations become even more telling in the post-COVID-19 period. During this time, residential REITs exhibit stronger relationships with both the REIT and equity markets, with correlations that rise to 0.82 with the Dow Jones Residential REITs Index and 0.76 with the Russell 3000 Index. This pattern suggests that the pandemic and its aftermath increased the integration of residential REITs with broader financial markets. As liquidity conditions tightened and investor sentiment became more risk-sensitive, the returns of residential REITs began to move more closely with general market trends.

Even with this heightened co-movement, residential REITs continued to preserve meaningful diversification benefits. Their comovement with the overall equity market remains below 0.80, indicating that they do not move in lockstep with stocks. This moderate level of correlation enhances their value as a portfolio component, providing a cushion against equity volatility while still capturing upside potential during market recoveries.

Overall, the correlations in Table 2 illustrate how residential REITs occupy a middle ground, closely tied to real estate fundamentals but responsive to market-wide factors. Their behavior over time reflects both their defensive nature and their growing integration into capital markets, particularly during and after the COVID-19 pandemic.

6.2. Analysis of Risk-Adjusted Returns

Table 3. Risk-adjusted performance (Sharpe, Sortino, and Omega ratios) for Residential REITs, the Dow Jones Residential REITs Index, the Dow Jones Composite All REIT TR USD Index, and the RUSSELL 3000 Index during the period January 2010 to June 2025 and the sub-period March 2020 to June 2025.

	Sharpe Ratio	Sortino Ratio	Omega Ratio
Residential REITs
January 2010 to June 2025	0.21	0.35	1.80
March 2020 to June 2025	0.16	0.27	1.56
Dow Jones Residential REITs Index
January 2010 to June 2025	0.17	0.27	1.54
March 2020 to June 2025	0.04	0.06	1.11
Dow Jones Composite All REIT TR USD Index
January 2010 to June 2025	0.15	0.22	1.47
March 2020 to June 2025	0.05	0.07	1.14
Russell 3000 Index
January 2010 to June 2025	0.24	0.39	1.85
March 2020 to June 2025	0.22	0.35	1.72

assesses the risk-adjusted performance of residential REITs and benchmark indexes across two periods: the complete sample from January 2010 to June 2025 and the post-COVID-19 period from March 2020 to June 2025. The table evaluates how effectively various investment vehicles compensate investors for the risks they incur using three commonly recognized measures—Sharpe ratio, Sortino ratio, and Omega ratio.

Over the entire period, residential REITs demonstrate a Sharpe ratio of 0.21 and a Sortino ratio of 0.35, indicating that they delivered consistent excess returns relative to total and downside risk. The Omega ratio of 1.80 shows that the probability-weighted gains from residential REITs have been nearly twice as large as their potential losses. These values collectively suggest that, for more than a decade, residential REITs have had a stable, reliable performance profile to compete with broader equity benchmarks.

In the post-COVID-19 period, performance understandably moderated as economic volatility increased and interest rates rose. The Sharpe ratio declined slightly to 0.16, and the Sortino ratio eased to 0.27. However, these figures still compare favorably with broader REIT benchmarks, whose Sharpe ratios declined closer to zero. The Omega ratio of 1.56 during this period signals that, despite heightened uncertainty, residential REITs continued to deliver a favorable balance between upside potential and downside risk.

Residential REITs maintained positive risk-adjusted performance even amid turbulence, while other REIT indices experienced sharper declines. This resilience reflects several structural strengths unique to the residential sector: steady rent flows, persistent housing demand, and the relatively short lease durations that allow for quicker rent adjustments in inflationary environments.

Table 3 shows that residential REITs maintained competitive and stable performance through both expansionary and crisis periods, generating steady returns while containing downside risk.

6.3. Empirical Analysis of Net Monthly Alphas Based on Multi-Factor

Table 4. Net monthly alpha for residential REITs (January 2010 to June 2025 and March 2020 to June 2025).

	January 2010 to June 2025	March 2020 to June 2025
Adjusted R2	0.47	0.59
Alpha	0.24	−0.02
Mkt-RF	0.58 ***	0.68 ***
SMB	0.12	0.18
HML	0.17 *	0.25 *
RMW	0.09	0.13
CMA	−0.15	−0.20

*** statistically significant at 1% significance level, * statistically significant at 10% significance level.

presents alpha estimates for residential REITs from the Fama–French five-factor model. The alpha estimates for residential REITs cover monthly returns from January 2010 to June 2025 while emphasizing performance changes during the post-COVID-19 era from March 2020 to June 2025.

The alpha of the residential REITs over the full sample period is small but positive (0.24% per month) and statistically not significant. In the post-COVID-19 period, alpha becomes small and negative (−0.02%), suggesting that the returns on residential REITs are largely in line with or even slightly below the expected return given their risk exposures. This may reflect the challenges faced by the residential REIT sector in adapting to structural shifts within housing markets, increasing interest rates, and broader macroeconomic pressures following the pandemic. Nonetheless, the reported alphas do not reach statistical significance.

The factor loadings on the market factor (Mkt–RF) are highly statistically significant (0.58 before COVID and 0.68 after COVID), which is consistent with strong linkage of residential REIT returns to the general market. Exposure to the HML factor has increased in the post-pandemic sample period (0.17 before COVID to 0.25 after COVID), which is consistent with a greater value orientation, possibly in the context of a greater demand for safe and income-producing assets during periods of economic uncertainty. The factor loadings for SMB (size) and RMW (profitability) are positive, yet not statistically significant. This suggests a weak or limited relationship between residential REIT returns and the variations in average returns observed across firms based on market capitalization and profitability characteristics. The negative values for the CMA (investment) factor (−0.15 before COVID to −0.20 after) suggest that residential REITs do not act like conservative investment firms.

Overall, the model explains a substantial portion of variation in returns, with adjusted R² improving from 0.47 to 0.59 in the later period, showing that the five-factor framework captures more of the sector’s dynamics post-COVID. The findings point out that the residential REITs displayed limited ability to generate abnormal returns beyond systematic risk exposures, particularly after 2020. Their increasing alignment with value characteristics and persistent market sensitivity suggests that investors should view them as stable, market-dependent assets rather than as sources of consistent alpha generation.

6.4. Empirical Analysis of Five-Factor Models with Benchmark Indices

Table 5. Net monthly alphas for residential REITs and benchmark indexes for the Fama–French multi-factor model from January 2010 to June 2025 and the sub-period March 2020 to June 2025.

	Residential REITs	Dow Jones Residential REITs Index	Dow Jones Composite All REIT TR USD Index	Russell 3000 Index
January 2010 to June 2025	0.47	0.11	−0.21	−0.02 **
March 2020 to June 2025	−0.02	−0.89 *	−0.93 ***	−0.04 ***

*** statistically significant at 1% significance level, ** statistically significant at 5% significance level, * statistically significant at 10% significance level.

extends the Fama–French multifactor analysis by comparing the net monthly alphas of residential REITs against major benchmark indices, including the Dow Jones Residential REITs Index, the Dow Jones Composite All REIT TR USD Index, and the Russell 3000 Index, over two periods: the full sample (January 2010–June 2025) and the post–COVID-19 period (March 2020–June 2025).

During the full sample period, residential REITs generated a positive alpha of 0.47% every month, exceeding all benchmark indices. On average, residential REITs outperformed passive benchmarks and major equity indices when adjusting for common risk factors. The Dow Jones Residential REITs Index had a small positive alpha of 0.11%, while the Dow Jones Composite All REIT TR USD Index and Russell 3000 Index had slightly negative or negligible alphas of −0.21% and −0.02%, respectively. These results show that residential REITs, as a separate group, were better at generating extra returns over the long term than the larger REIT or equities markets.

Nonetheless, this advantage in performance disappeared following the pandemic. Residential REITs’ alpha went from positive to negative (−0.02%) between March 2020 and June 2025. This meant that the sector was no longer making more money than it was taking on risk. The drop was bigger for the larger REIT indices. The Dow Jones Residential REITs Index had an alpha of −0.89%, and the Dow Jones Composite All REIT TR USD Index fell even more to −0.93%. The Russell 3000 Index recorded a slight yet statistically significant negative alpha (−0.04%), indicating that equities underperformed amid the period of heightened volatility.

The post-2020 reversal is attributable to the higher sensitivity of the sector to macroeconomic/policy-related headwinds, such as higher mortgage rates, inflation, and changing housing demand post-COVID. The findings, therefore, highlight that although residential REITs may have been able to offer alpha and diversification potential in the past, their risk-adjusted returns have deteriorated in recent years, becoming more closely aligned with the overall equity market.

6.5. Empirical Analysis of Conditional Multi-Factor Model

Table 6. Net monthly alphas for residential REITs for the conditional five-factor model (January 2010 to June 2025 and the sub-period March 2020 to June 2025).

Variables	January 2010 to June 2025		March 2020 to June 2025
Conditional Fama–French 5-factor model
Adjusted R2	0.48		0.60
	Coefficient	T-statistics	Coefficient	T-statistics
Alpha	0.32	1.40	0.41	0.86
Mkt-RF	0.73	1.00	0.90	0.67
SMB	0.11	1.16	0.20	1.17
HML	0.11	1.21	0.11	0.71
RMW	0.05	0.42	0.06	0.33
CMA	−0.06	−0.46	−0.06	−0.31
(Mkt-RF) × Slope of yield curve	0.02	0.39	0.10	1.11
(Mkt-RF) × Unemployment	−0.05 *	−1.85	−0.10 **	−2.05
(Mkt-RF) × Change in Gross Domestic Product	0.00	−0.12	−0.01	−0.29
(Mkt-RF) × Quality Spread Differential	0.57 **	2.36	0.90 **	2.06
(Mkt-RF) × Inflation	−0.21	−0.77	−0.31	−0.64

** statistically significant at 5% significance level, * statistically significant at 10% significance level.

presents the conditional net monthly alphas for residential REITs, estimated using a conditional version of the Fama–French five-factor model over two periods: the full sample (January 2010–June 2025) and the post–COVID-19 period (March 2020–June 2025). Unlike the static model in Table 5, the conditional specification accounts for the changing relationships between REIT returns and risk factors over time, particularly across varying market conditions and economic cycles.

Alpha remains positive but not significant across both models. There is no evidence of abnormal performance once market factor exposure changes are considered. This suggests that unconditional alpha mainly reflects time-varying systematic risk.

The unconditional market factor (Mkt–RF) remains positive but not significant, highlighting that constant-beta models do not adequately capture the risks of residential REITs. The size, value, profitability, and investment factors also show no statistical significance, reinforcing that traditional equity factors explain little of the expected return differences when accounting for state-dependent market exposure.

The unemployment interaction term is negative and statistically significant post-COVID and marginally significant in the full sample. This indicates that residential REITs reduce market exposure when the labor market is weak, due to a decline in beta as real economic activity slows. By contrast, the interaction between Mkt–RF and QSD is positive and statistically significant across both samples and is quantitatively stronger during the post-COVID period. The findings indicate that during times of market stress and rising risk premium, market exposure for REITs increases and residential REITs become more pro-cyclical. Lastly, the lack of statistical significance associated with interactions with yield-curve slope and inflation suggests that risk premia embedded in real-economy variables matter more for pricing conditional REIT market risk than variables picking up expectations in financial markets.

The model fits the data better for the post-COVID period. This supports the view that state-dependent risk exposures are becoming more significant in determining REIT returns. The findings indicate that conditional alpha can become important even if unconditional alpha is no longer present, since abnormal performance tends to show up during certain macroeconomic situations but gets averaged out when models presume constant parameters.

6.6. Empirical Analysis of Downside Risk

Table 7. VaR (5th percentile loss) and CVaR (average loss below the 5th percentile) confidence intervals for residential REITs and stock market indices that include the Dow Jones Residential REITs Index, the Dow Jones Composite All REIT TR USD Index, and the Russell 3000 Index using data from January 2010 to June 2025.

	Residential REITs	Dow Jones Residential REITs Index	Dow Jones Composite All REIT TR USD Index	Russell 3000 Index
Value at Risk (VaR)	−4.95	−6.77	−7.02	−6.40
Conditional Value at Risk (CVaR)	−7.38	−9.69	−9.70	−8.74

evaluates the downside risk profile of residential REITs compared with major benchmarks, including the Dow Jones Residential REITs Index, the Dow Jones Composite All REIT TR USD Index, and the Russell 3000 Index, using VaR and CVaR at the 95% confidence level using data from January 2010 to June 2025. These measures provide insight into potential portfolio losses during extreme market downturns, highlighting the resilience or fragility of residential REITs under stress conditions.

Residential REITs’ VaR is −4.95% at the 5% lower tail, indicating that 95% of the time, monthly losses will be less than 4.95% in magnitude. This VaR magnitude is smaller (less negative) than that of all three benchmark indices (the Dow Jones Residential REITs Index (−6.77%), the Dow Jones Composite All REIT Index (−7.02%), and the Russell 3000 Index (−6.40%)), suggesting that residential REITs are less likely to experience extreme downside shocks and have provided more capital stability in adverse market conditions.

The findings are further supported by CVaR values, which represent the average loss that exceeds the VaR threshold. Residential REITs have a CVaR of −7.38%, indicating smaller average losses in the worst-case scenarios, compared to the other REIT and equity benchmarks with higher downside risks (the Dow Jones Residential REITs Index (−9.69%), the Dow Jones Composite All REIT TR USD Index (−9.70%), and the Russell 3000 Index (−8.74%)). Residential REITs are likely to lose less money even during the worst 5% of market months, meaning they offer better protection against losses than the overall REIT or stock markets.

These results are consistent with the literature that identifies residential REITs as a defensive REIT sector. Although residential REITs are classified as real estate assets, they often exhibit more stable cash flows and lower volatility than the other REIT sectors, especially during economic downturns. The underlying residential properties are often less sensitive to economic cycles due to the more stable and essential demand for housing compared to the more cyclical and speculative demand for commercial real estate.

Overall, Table 7 shows that residential REITs provided better downside risk control, as they experienced smaller potential losses during extreme market events compared to the three benchmarks. This makes residential REITs a viable option for portfolio managers and risk-averse investors who are looking for a defensive diversifier to reduce the probability of large drawdowns in their mixed-asset portfolios or multi-REIT portfolios.

6.7. Portfolio Optimization Analysis: GMV, Tangency, and Risk Parity Portfolios

We also build optimized portfolios to compare the optimal portfolios based on the monthly return of residential REITs and the Russell 3000 Index. The three modeling results help to demonstrate the optimal risk–return combinations that are available to an investor seeking to include residential REITs in their diversified portfolio.

Table 8. (a): Portfolio Weights: REITs and Russell 3000 Index. (b): Portfolio Performance Characteristics.

(a)
Portfolio Type	Residential REITs Weight		Russell 3000 Index Weight
Global Minimum Variance (GMV)	0.652		0.348
Tangency Portfolio	0.375		0.625
Risk Parity Portfolio	0.524		0.476
(b)
Portfolio Type	Expected Return (%)	Volatility (%)	Sharpe Ratio
Global Minimum Variance (GMV)	1.016	3.774	0.242
Tangency Portfolio	1.081	3.907	0.251
Risk Parity Portfolio	1.043	3.798	0.248

a,b reports the optimal portfolio allocations and performance characteristics for the GMV, Tangency, and Risk Parity portfolios constructed from residential REITs and the Russell 3000 Index using data from January 2010 to June 2025.

The GMV portfolio was the most heavily weighted toward residential REITs, with the portfolio weights of 65.2% residential REIT and 34.8% Russell 3000 Index. The model provided the best stability with the lowest volatility of 3.774% while delivering a monthly return of 1.016%. The Tangency portfolio, known for its optimal Sharpe ratio, primarily consists of equities: 62.5% in Russell 3000 and 37.5% in REITs. It achieves a monthly return of 1.081% with a volatility of 3.907% and holds a Sharpe ratio of 0.251. The Risk Parity portfolio split the risk contributions about evenly between REITs (52.4%) and Russell 3000 (47.6%) and generated a monthly return of 1.043% and volatility of 3.798% with a Sharpe ratio of 0.248.

The Tangency portfolio delivers the best risk-adjusted returns, supporting the earlier finding that a modest allocation to residential REITs can enhance portfolio efficiency without increasing overall risk. The GMV and Risk Parity results also illustrated the defensive power of the REIT sector to reduce the portfolio volatility without giving up the attractive return and the importance of REIT allocation in a diversified portfolio with a mix of assets for investors.

7. Summary and Conclusions

This study evaluated the performance and risk of residential REITs against benchmarks including the Dow Jones Residential REITs Index, the Dow Jones Composite All REIT TR USD Index, and the Russell 3000 Index from January 2010 to June 2025, with a subset analysis from March 2020 to June 2025 to assess performance since the start of the COVID-19 pandemic.

The study found that residential REITs provided steady long-term returns with less volatility and better downside protection than broader equity and REIT benchmarks. Although performance declined after COVID-19, residential REITs outperformed most other REIT types in risk-adjusted terms.

Multi-factor asset pricing analysis indicated that residential REITs exhibited minimal and statistically insignificant unconditional alpha across the entire sample, with their performance predominantly determined by systematic risk factors. Following the COVID-19 pandemic, alpha diminished further, mirroring heightened market integration and a greater responsiveness to macroeconomic fluctuations. Conditional factor models demonstrate that the seemingly abnormal performance observed in unconditional models is largely due to time-varying risk exposures, rather than sustained excess returns, thereby emphasizing the necessity of state-dependent modeling in the assessment of REIT performance.

VaR and CVaR analysis showed that residential REITs had smaller tail losses than general REITs and equities, highlighting their defensive nature. Portfolio optimization results further demonstrated that allocating residential REITs alongside equities improved portfolio efficiency by reducing volatility without materially sacrificing expected returns.

Residential REITs are regarded as a stabilizing asset class that provides diversification advantages and mitigates downside risk, especially during periods of market volatility. These attributes make residential REITs attractive for risk-averse and income-focused investors seeking to strengthen portfolio resilience amid uncertain economic conditions.