Do Boards Shape REIT Performance? Evidence from the South African REIT Sector

1. Introduction

Board structure sits at the centre of corporate governance because it determines how effectively a firm is monitored, challenged, and guided. In listed property companies, this matters in a very practical way because boards approve strategy, capital allocation, acquisitions and disposals, debt and hedging decisions, related-party transactions, and distribution policies, all of which feed directly into cash flows and investor confidence (Jensen & Meckling, 1976; Pfeffer & Salancik, 1978). When board design is weak, agency costs rise, disclosure credibility can suffer, and market trust can unwind quickly (Ntim et al., 2019).

Within the South African REIT sector, the composition and structure of boards carry significant practical implications. Historical events have demonstrated that governance uncertainties can prompt pronounced market responses, evidenced by notable share-price declines among leading listed property companies, and broader index volatility during periods of increased scrutiny (SAREIT, 2023; JSE, 2018). These instances highlight an important industry dynamic: capital is highly mobile, funding environments are subject to rapid change, and governance signals play a critical role in shaping both company valuation and access to financing. For boards and executive management teams, the relevant issue is not the importance of governance itself, but rather which particular board attributes correlate with improved outcomes in a sector where distributions, stable cash earnings, and prudent financial management are vital considerations for investors.

International research increasingly associates board activity, board size, board independence, and board tenure with firm performance, although empirical findings remain mixed across institutional settings, sectors, and performance measures used (Petrova, 2023; Hasan et al., 2024; Noguera, 2024; Zureigat et al., 2024). Recent governance literature further suggests that the effectiveness of board structures depends on firm-specific characteristics, regulatory environments, and the strategic demands placed on boards, particularly within capital-intensive industries such as real estate investment trusts (REITs). In the South African REIT sector, the need for specific governance evidence is especially important given the post-2013 SA-REIT regulatory framework and the sector’s dependence on debt financing, property valuation accuracy, and sustainable income distributions. South African studies further highlight that governance quality, board oversight, and board composition remain important determinants of REIT stability and operating performance within the JSE-listed property sector (Mofokeng, 2024; Kitulazzi et al., 2025).

Consequently, this paper investigates the relationship between board structure and performance among REITs listed on the Johannesburg Stock Exchange (JSE). The research specifically evaluates whether board activity, size, independence, and tenure are correlated with Funds from Operations per share (FFO/P/S), dividend yield, ROA, ROE, ROIC, and EPS across 30 REITs from 2013 to 2025. To address prevalent endogeneity concerns in governance-performance studies, such as unobserved firm effects and dynamic performance persistence, the analysis employs the Difference Generalised Method of Moments (GMM-Diff), as recommended by Hill et al. (2021).

The significance of this paper in the industry is central to its contribution. Firstly, it provides unique governance insights specific to REITs in South Africa by examining board activity, size, independence, and tenure collectively, instead of just focusing on individual board attributes. Secondly, the analysis incorporates performance measures relevant to REITs, such as FFO per share and dividend yield, alongside standard profitability ratios, ensuring alignment with the metrics commonly employed by boards, funders, and investors to evaluate listed property companies (Feng et al., 2021). Thirdly, the selected timeframe from 2013 to 2024 encompasses the post-SA-REIT era and includes periods where governance signals had pronounced market implications, thereby rendering the findings directly applicable to board structuring, investor communication, and governance policy within the listed property sector (SAREIT, 2023).

The findings demonstrate critical considerations for boards and executives for the REIT sector such as determining optimal meeting frequency for effectiveness, assessing whether expanding board size contributes valuable expertise or increases coordination challenges, evaluating the balance between board independence and informed oversight, and examining how tenure affects both stability and potential complacency (Institute of Directors in South Africa (IoDSA), 2025). By correlating specific board characteristics with various performance indicators, this study offers actionable insights for REIT boards, investors, and regulators aiming to establish governance practices that promote trust and sustainable value creation within the sector (Deloitte, 2020; Ntim & Osei, 2011).

1.1. Problem Statement

Board structure is widely recognised as an important mechanism through which corporate governance influences firm outcomes. However, within the South African REIT sector, empirical evidence capable of guiding effective board design remains limited and fragmented. Recent international studies continue to report that board activity, board size, board independence, and board tenure influence organisational performance, although the direction and magnitude of these relationships vary across institutional settings, governance systems, and performance measures used (Petrova, 2023; Hasan et al., 2024; Noguera, 2024; Zureigat et al., 2024). More recent governance research further suggests that the effectiveness of board structures is highly dependent on sector-specific characteristics and the strategic demands placed on boards, particularly within capital-intensive industries such as REITs (Kitulazzi et al., 2025).

For REITs, this uncertainty carries important practical implications. REIT boards operate within an environment characterised by recurring earnings requirements, high payout obligations, debt refinancing pressures, property valuation sensitivity, and strong dependence on investor confidence. Governance failures or weak oversight may therefore have immediate consequences for financing conditions, market valuations, and distribution sustainability. Recent evidence further highlights that governance quality and board oversight remain central to operational stability and financial performance within listed property firms and emerging market REITs (Mofokeng, 2024; Noguera, 2024).

Although governance frameworks such as the Institute of Directors in South Africa (IoDSA) King V code emphasise board independence, effective oversight, and balanced board composition, policymakers, regulators, and investors still face a shortage of REIT-specific empirical evidence linking individual board characteristics to critical performance indicators such as funds from operations per share (FFO_PS), dividend yield, and return on invested capital. Existing South African studies largely focus on general listed companies or broader governance themes, with limited attention devoted specifically to the listed property sector (Viviers et al., 2023; Moloi et al., 2024; Kitulazzi et al., 2025). As a result, governance recommendations may remain overly generic and compliance-oriented rather than grounded in evidence identifying which board attributes are most strongly associated with operational efficiency, capital allocation discipline, and sustainable shareholder distributions within South African REITs.

The gap we are addressing is therefore both empirical and practical. Empirically, South African REIT governance research has not consistently tested board activity, size, independence, and tenure together using REIT-relevant performance measures. Practically, industry stakeholders need clearer evidence on which board features are associated with stronger outcomes, especially in a market where governance signals affect investor trust and access to capital (SAREIT, 2023; Feng et al., 2021). Moreover, this paper responds to that need by systematically examining the association between board structure (activity, size, independence, and tenure) and multiple REIT performance indicators for JSE-listed REITs over 2013-2025, using GMM-Diff to mitigate endogeneity concerns and produce more reliable inferences for governance practice and policy (Hill et al., 2021).

1.2. Objectives and Hypothesis

The primary research objective of this study is therefore to assess the relationship between board structure and the financial and operating performance of JSE-listed REITs. Performance is evaluated using both REIT-specific measures that are central to industry practice, such as Funds from Operations per share, and dividend yield, and conventional accounting-based measures, including ROA, ROE, ROIC, and EPS. This dual approach reflects how performance is assessed in practice by boards, investors, lenders, and analysts within the listed property market (Feng et al., 2021; Ntuli & Akinsomi, 2017).

Instead of treating board characteristics as isolated mechanisms, this paper adopts an integrated view of board structure. Agency theory suggests that more active and independent boards enhance monitoring and reduce agency costs, improving performance. Stewardship theory highlights the potential value of experience, continuity, and collective expertise, implying that board size and tenure may support performance up to a point. Resource dependence theory further emphasizes the role of boards in providing access to external resources, legitimacy, and market credibility, which are particularly important in capital-intensive sectors such as listed property (Jensen & Meckling, 1976; Donaldson & Davis, 1991; Pfeffer & Salancik, 1978). Synthesising these perspectives, the study tests the overarching proposition that board structure is systematically related to REIT performance, but that the direction and strength of this relationship differ across board dimensions and performance measures. In particular, the analysis examines whether greater board activity and independence are associated with improved performance, whether board size reflects a trade-off between resource availability and coordination costs, and whether extended board tenure enhances stability or weakens oversight through entrenchment effects (Hermalin & Weisbach, 2001; Guest, 2009).

Accordingly, the empirical analysis is guided by the following general hypothesis:

• H₀: Board structure has no significant impact on the performance of South African REITs.
• H₁: Board structure is significantly associated with the performance of South African REITs.

This hypothesis is evaluated by examining the individual and joint effects of board activity, board size, board independence, and board tenure on multiple performance measures within a dynamic panel analysis framework. This approach advances the analysis beyond mere description by providing evidence that has direct implications for board structure, investor decisions, and governance policy within the South African listed property sector.

2. Literature Review

2.1. Theoretical Framework

The analysis of the relationship between board structure and company performance requires an examination of foundational theoretical frameworks. These include agency theory, stewardship theory, and resource dependence theory, each providing unique perspectives on how corporate governance influences organizational effectiveness.

2.1.1. Agency Theory and REIT Governance

Agency theory was first conceptualized by Berle and Means (1932) and formalized by Jensen and Meckling (1976) to address the conflict between principals (shareholders) and agents (directors/managers). Agents may pursue self-serving objectives, generating costs for shareholders unless proper oversight mechanisms are established. Independent directors and active boards play a significant role in mitigating these conflicts by enhancing monitoring and reducing managerial opportunism (Tosi & Gomez-Mejia, 1989). For South African REITs, this oversight is particularly important because investors are diversified and rely on professional management for asset allocation and valuation. The 2018 Viceroy Research governance scandal, where allegations of malpractice triggered large share price declines across leading REITs, highlighted the critical need for robust governance. Although agency monitoring creates costs, such as board remuneration and slower decision-making, it is critical for protecting investor trust, ensuring dividend sustainability, and maintaining market stability (Jensen & Meckling, 1976). Therefore, agency theory suggests that independent and active boards will be positively associated with REIT performance, providing a foundation for this study’s hypotheses on board activity and independence.

2.1.2. Stewardship Theory

The stewardship theory posits a contrasting perspective, proposing that managers and directors act as responsible stewards of shareholder wealth rather than opportunistic agents. Donaldson and Davis (1991) argue that executives are intrinsically motivated to achieve organizational success, while extrinsic rewards tied to performance further align their interests with those of shareholders. From this perspective, governance mechanisms should not overemphasize monitoring but instead empower directors with autonomy and discretion. Board size and tenure are particularly relevant. Larger boards can enhance decision-making by incorporating diverse expertise and perspectives (Dalton et al., 1999), while longer tenure allows directors to develop deep institutional knowledge and promote stability in strategic direction (Jones, 1995). Within South African REITs, where real estate markets are cyclical and regulatory environments are complex, continuity and accumulated expertise are valuable for navigating uncertainty. Unlike agency theory, stewardship theory reduces monitoring costs and, instead, highlights trust, collaboration, and long-term commitment as drivers of performance (Donaldson & Davis, 1991). This provides theoretical grounding for hypothesizing positive effects of board size and tenure on REIT performance.

2.1.3. Resource Dependence Theory

The resource dependence theory (RDT), developed by Pfeffer and Salancik (1978), views boards as strategic resources that connect companies to their external environment. Boards provide access to networks, legitimacy, and external knowledge, thereby reducing dependence on uncertain markets (Haniffa & Hudaib, 2006). Larger and more diverse boards are argued to broaden these connections, while independent directors contribute credibility and bring external expertise into board deliberations (Dalton et al., 1999). In the South African REIT market, where companies rely heavily on capital raising, investor confidence, and regulatory legitimacy, these external connections are particularly important. Independent and resourceful boards not only monitor management but also attract financing, enhance disclosure credibility, and secure access to market opportunities. The RDT theory thus complements agency and stewardship viewpoints, reflecting that board structures also serve an outward-facing role, enhancing organizational resilience and adaptability.

2.2. Empirical Literature

2.2.1. Board Activity and REIT Performance

Board activity, commonly measured by the frequency of board meetings, reflects the degree of director engagement in monitoring, strategic decision-making, and financial oversight. Agency theory suggests that frequent board meetings strengthen monitoring and reduce managerial opportunism, while resource dependence theory argues that active boards improve access to expertise and external resources (Jensen & Meckling, 1976; Dalton et al., 1999). Within REITs, where boards oversee capital allocation, leverage management, property acquisitions, and dividend distribution, board activity is expected to play an important governance role.

Recent empirical evidence, however, presents inconclusive findings. Nguyen and Huynh (2023) found that meeting frequency positively influenced financial performance among Vietnamese real estate and construction firms, suggesting that active boards improve strategic responsiveness and operational oversight. Similarly, Noguera (2024) reported that stronger board and committee engagement enhanced REIT performance in US REITs, particularly through improved investment monitoring and capital allocation decisions. These findings support the argument that active boards contribute positively to operational efficiency and shareholder value in property-focused firms.

Conversely, recent literature also highlights the potential costs of excessive board activity and director busyness. Venkatesh et al. (2025) identified a nonlinear relationship between busy boards and firm performance, showing that director effectiveness declines once monitoring capacity becomes overstretched. Supporting the distraction hypothesis, Cao and Nguyen (2025) found that busy boards increased earnings management behaviour in Vietnamese real estate firms, while Alomair and Al Naim (2025) reported that governance busyness may weaken monitoring effectiveness due to competing director responsibilities. These findings suggest that greater board activity does not automatically translate into superior performance.

From a South African perspective, the Institute of Directors in South Africa (IoDSA) (2025) King V code recommends that boards meet regularly and maintain sufficient engagement to ensure effective governance oversight. Prior South African studies, including Mofokeng (2024), Moloi et al. (2024), and Zvinowanda and Odendaal (2025) report a positive association between board meeting frequency and firm performance, although evidence specific to South African REITs remains limited. Given the operational complexity and high payout requirements of REITs, the relationship between board activity and performance in the South African listed property sector therefore remains an important area for further empirical investigation.

2.2.2. Board Size and REIT Performance

Board size remains one of the most debated corporate governance mechanisms in the REIT literature. Agency theory argues that larger boards strengthen monitoring capacity and reduce agency conflicts through broader oversight and increased director expertise (Jensen, 1993; Guest, 2009). Resource dependence theory similarly suggests that larger boards improve access to external resources, industry networks, and financing opportunities, which may be particularly valuable in REITs given the sector’s capital-intensive nature. Conversely, stewardship and coordination perspectives caution that excessively large boards may create communication inefficiencies, slower decision-making, and weaker accountability (Hermalin & Weisbach, 2001). Recent literature further notes that larger boards may struggle with coordination and attendance challenges, reducing monitoring effectiveness in complex firms (Mofokeng, 2024).

However, empirical evidence remains inconclusive. Noguera (2024) showed that REIT performance depends more on the quality and composition of board committees than board size alone, particularly regarding investment oversight and financial expertise. Recent emerging markets studies generally report positive associations between board size and firm performance. Hasan et al. (2024) found that larger boards improved the performance of Bangladeshi financial institutions through stronger strategic engagement and stakeholder oversight. Similarly, Zureigat et al. (2024) and Satapathy et al. (2023) reported that board size positively influenced financial performance and market returns in Jordanian and Indian firms, respectively. However, other studies suggest that the relationship is conditional rather than universally positive. Petrova (2023) found that board structure-performance relationships vary substantially across countries, indicating that no single optimal board size exists across firms and institutional settings. Khan and Mahmood (2023) further argued that leaner boards may improve efficiency and monitoring effectiveness.

The Companies Act No. 71 of 2008 in South Africa requires a minimum of three (3) directors for public companies. Ajayi (2022) found no significant relationship between board size and REIT financial performance volatility, whereas Scholtz and Kieviet (2017) reported positive associations between board size, Tobin’s Q, and ROA. Mofokeng (2024) further highlighted the importance of effective board oversight and governance quality in enhancing the performance of JSE-listed REITs. These inconclusive findings demonstrate the complexity inherent in the board size-performance relationship and suggest that the impact of board size on REIT performance may depend on the chosen performance metrics.

2.2.3. Board Independence and REIT Performance

Board independence is widely recognised as a fundamental corporate governance mechanism that enhances accountability, monitoring, and strategic oversight within organisations. Defined as the extent to which a board operates independently from management, board independence is primarily achieved through the appointment of independent non-executive directors who contribute diverse expertise, external networks, and objective judgement (Hoitash, 2010; Fuzi et al., 2016). Theoretical perspectives provide differing explanations for its importance. Agency theory argues that independent directors reduce agency conflicts by monitoring management and safeguarding shareholder interests, particularly where managerial opportunism exists (Mohamad & Sahul Hamid, 2022). In contrast, stewardship theory views directors as strategic partners who support executives in achieving organisational goals, while resource dependency theory emphasises the value of directors’ external knowledge and access to critical resources (Donaldson & Davis, 1991; Pfeffer & Salancik, 1978).

Despite strong theoretical support, empirical findings on board independence and REIT performance remain inconclusive. Several studies report positive associations between independent directors and improved firm performance. Yeon et al. (2021) found that outside directors positively moderated the relationship between environmental management and lodging REIT performance, while Mariani et al. (2018) highlighted the importance of board independence and diversity in enhancing the financial performance of European Green REITs. Similarly, David et al. (2021) reported that independent directors positively influenced Malaysian REIT performance, although the relationship was insignificant in Hong Kong REITs, suggesting differences across markets. More recent evidence by Noguera (2024) further demonstrates that board committee composition and director expertise significantly affect U.S. REIT performance, particularly in investment committees. However, contrary evidence exists. For instance, Khan et al. (2024) reported that excessive board independence may negatively affect performance in emerging markets due to close social and financial ties between directors and dominant shareholders.

From a South African perspective, IoDSA (2025) King V code recommends that the board should have a majority of non-executive directors, most of whom should be independent. It also recommends an independent non-executive chairperson and a fully independent audit committee. These expectations are important for REITs because investors rely heavily on credible oversight of asset values, rental income quality, debt sustainability and dividend policy. South African evidence is also inconclusive. For instance, Viviers et al. (2023) study on board independence and dividend distributions shows that independence remains important in the South African governance environment, especially where shareholder protection and payout decisions are concerned. However, Kitulazzi et al. (2025) JSE-listed real estate evidence reports that the proportion of non-executive directors did not significantly affect performance metrics, while independent directors had a negative relationship with financial performance but a positive relationship with sustainability. This suggests that independent boards may place greater emphasis on long-term sustainability, risk control and compliance, which may not immediately translate into higher short-term accounting returns.

2.2.4. Board Tenure and REIT Performance

Board tenure is a contested governance mechanism because it can strengthen both board capability and board entrenchment. From a resource-dependency perspective, longer-serving directors may improve REIT performance by accumulating firm-specific knowledge, property-market expertise, regulatory understanding, and stakeholder relationships. This is especially relevant in REITs, where investment decisions require knowledge of property cycles, asset valuation, tenant risk, and capital-market conditions. This view is consistent with Noguera (2024) who found that longer-tenured directors with real estate expertise, especially on investment committees, were associated with stronger U.S. equity REIT performance. Similarly, broader real estate governance evidence shows that shareholder-oriented governance mechanisms improve performance and firm value (Morri et al., 2023)

However, agency theory raises concerns that excessive tenure may weaken independence, encourage management familiarity, and reduce critical oversight. Haider et al. (2025) argue that lower director tenure can reduce groupthink and improve investor confidence in governance, while Zeng et al. (2025) also found that longer director tenure may increase corporate misconduct risk, suggesting that tenure can create management-friendly boards. However, Cashman et al. (2018) cautions that longer board tenure sometimes link to better decision-making and development activity, but also to challenges such as reduced executive compensation monitoring.

The IoDSA (2025) King V code advises against the imposition of rigid term limits for board members. This guidance highlights the importance of retaining experienced directors, especially in the real estate sector, where expertise is essential for navigating market complexities, understanding local regulatory requirements, and making informed investment decisions. However, Shill and Strand (2021) warn of the risks associated with complacency among long-serving board members. They note that directors with lengthy tenures may be susceptible to groupthink, which can hinder the board's ability to identify emerging trends or recognize potential risks, ultimately affecting the effectiveness of governance.

3. Methodology

3.1. Sample and Data

We adopt a quantitative methodology to examine the relationship between board structure and the performance of REITs. This approach involves using numerical data and statistical techniques to derive empirical insights into how various mechanisms of board structure (such as activity, size, independence, and tenure) affect REIT performance. Our analysis focuses on South African REITs listed on the JSE between 2013 and 2025, allowing for an extensive analysis of governance trends and their impact over time. The final sample comprises 30 REIT companies. This sample size enhances the robustness of the findings and provides a comprehensive view of the sector's governance environment.

Financial and governance data were obtained from Iress® and Morningstar™, which are widely used databases in financial and investment research. The final sample was restricted to REITs with sufficient governance and financial data available across the study period.

3.2. Empirical Model

To examine the relationship between board structure and REIT performance, we employ a dynamic panel data model estimated using the Difference Generalised Method of Moments estimator developed by Arellano and Bond (1991). The estimator is appropriate for this study because it addresses endogeneity, unobserved firm-specific heterogeneity, simultaneity bias, and the dynamic persistence commonly observed in firm performance measures.

The general dynamic panel specification is expressed as:

$$Y_{it}=\alpha +{\rho Y}_{i,t-1}+{{\beta }_{1}B\_ACTIV}_{\left(it\right)}+{{\beta }_{2}B\_SIZE}_{\left(it\right)}+{{\beta }_{3}BIND}_{\left(it\right)}+{{\beta }_{4}BOARD\_TEN}_{\left(it\right)}+{\eta }_i+{\lambda }_t+{\epsilon }_{it}$$

(1)

We estimate six separate dynamic panel equations using alternative measures of REIT performance.

$$\begin{array}{c}\left({\displaystyle \genfrac{}{}{0pt}{}{\begin{array}{c}{FFO\_P/S}_{\left(it\right)}\\ {DIV\_YIELD}_{\left(it\right)}\end{array}}{\begin{array}{c}\begin{array}{c}{ROA}_{\left(it\right)}\\ {ROE}_{\left(it\right)}\\ {ROIC}_{\left(it\right)}\end{array}\\ {EPS}_{\left(it\right)}\end{array}}}\right)=\alpha +\left({\displaystyle \genfrac{}{}{0pt}{}{\begin{array}{c}{\rho FFO\_P/S}_{i,t-1}\\ {\rho DIV\_YIELD}_{i,t-1}\end{array}}{\begin{array}{c}\begin{array}{c}\rho {ROA}_{i,t-1}\\ {\rho ROE}_{i,t-1}\\ {\rho ROIC}_{i,t-1}\end{array}\\ \rho {EPS}_{i,t-1}\end{array}}}\right)+{{\beta }_{1}B\_ACTIV}_{it}+{{\beta }_{2}B\_SIZE}_{it}+{{\beta }_{3}BIND}_{it}+{{\beta }_{4}BOARD\_TEN}_{it}+{\eta }_i+{\epsilon }_{it}\end{array}$$

(2)

Where: $i=1,\dots ,30$ represents the sampled REITs; $t=2013,\dots ,2025$ represents the time dimension; $Y_t$ denotes the dependent variable measuring REIT performance; $\alpha$ is the intercept; $\rho$ captures the dynamic effect of lagged performance; ${B\_ACTIV}_{\left(it\right)}$ denotes board activity; ${B\_SIZE}_{it}$ signifies board size; ${BIND}_{it}$ represents board independence; ${BOARD\_TEN}_{it}$ denotes board tenure; ${\eta }_i$ captures unobserved firm-specific effects; ${\lambda }_t$ represents time effects; and ${\epsilon }_{it}$ is the idiosyncratic error term where ${E(\epsilon }_{it})=0.$Following the Arellano and Bond (1991), first differencing is applied to eliminate unobserved firm-specific fixed effects:

$${\mathsf{\Delta }Y}_{it}={\rho \mathsf{\Delta }Y}_{i,t-1}+{{\beta }_1\mathsf{\Delta }B\_ACTIV}_{\left(it\right)}+{{\beta }_2\mathsf{\Delta }B\_SIZE}_{\left(it\right)}+{{\beta }_3\mathsf{\Delta }BIND}_{\left(it\right)}+{{\beta }_4\mathsf{\Delta }BOARD\_TEN}_{\left(it\right)}+\mathsf{\Delta }{\epsilon }_{it}$$

(3)

The lagged dependent variable in first differences is correlated with the differenced error term. To address this endogeneity problem, deeper lags of the dependent variable in levels are used as instruments. The Arellano-Bond moment conditions are specified as:

$${\rm E}\left[Y_{i,t-s}.\mathsf{\Delta }{\epsilon }_{it}\right]=0$$

(4)

For $s\ge 2$ and $t=3$,…, $T$.

The board structure variables are treated as strictly exogenous and instrument themselves in differences:

$$E\left[X_{it}.\mathsf{\Delta }{\epsilon }_{it}\right]=0$$

(5)

Where: $X=(B\_ACTIV,B\_SIZE,BIND,BOARD\_TEN)$

The moment conditions are stacked into the matrix expression:

$$Z_i^{\text{'}}\mathsf{\Delta }{\epsilon }_i=0$$

(6)

Where: $Z_i$ represents the matrix of valid instruments for firm $i$.

The Difference GMM estimator minimises the following quadratic objective function:

$${\widehat{\beta }}_{GMM}=\arg min\left\{{(\mathsf{\Delta }Y-\mathsf{\Delta }X\beta )}^{ \prime }{ZW}_N{Z}^{\text{'}}\right(\mathsf{\Delta }Y-\mathsf{\Delta }X\beta \left)\right\}$$

(7)

The closed-form GMM estimator is expressed as:

$$\widehat{\beta }={\left[{\left(\mathsf{\Delta }X\right)}^{\text{'}}{ZW}_N{Z}^{\text{'}}\mathsf{\Delta }X\right]}^{-1}{\left(\mathsf{\Delta }X\right)}^{\text{'}}{ZW}_N{Z}^{\text{'}}\mathsf{\Delta }Y$$

(8)

The one-step weighting matrix is specified as:

$$W_1^{\left(N\right)}={\left[{\displaystyle \frac{1}{N}}\sum Z_i^{\text{'}}H{Z}_i\right]}^{-1}$$

(9)

Where: $H$ is a $\left(T-2\right)\times (T-2)$ matrix with: 2 on the main diagonal; -1 on the first off-diagonals; 0 elsewhere.

$$W_2^{\left(N\right)}={\left[{\displaystyle \frac{1}{N}}\sum Z_i^{\text{'}}\mathsf{\Delta }{\widehat{\epsilon }}_i\mathsf{\Delta }{\widehat{\epsilon }}_i^{\text{'}}{Z}_i\right]}^{-1}$$

(10)

using the one-step residuals $\mathsf{\Delta }{\widehat{\epsilon }}_i$. The two-step estimator substitutes $W_2^{\left(N\right)}$ into the GMM estimation equation.

3.3. Windmeijer Finite-Sample Correction

To address downward bias in two-step standard errors in finite samples, the finite-sample correction proposed by Windmeijer (2005) is applied.

The corrected variance estimator is given by:

$$V_{corr}\left({\widehat{\beta }}_2\right)+D.V\left({\widehat{\beta }}_1\right).D^{\text{'}}+{\left[D.V\left({\widehat{\beta }}_1\right).D^{\text{'}}\right]}^{\text{'}}$$

(11)

Where: $V\left({\widehat{\beta }}_1\right)$ is the one-step variance estimator; $V\left({\widehat{\beta }}_1\right)$ is the uncorrected two-step variance estimator; and $D$ denotes the derivative of the two-step estimator with respect to the one-step residuals.

The corrected standard errors are obtained from the square roots of the diagonal elements of $V_{corr}\left({\widehat{\beta }}_2\right)$.

3.4. Variable Definition and Measurement

Table 1 below provides a summary of variable definitions and measurements:

4. Findings

We applied the above equations to the selected dataset to analyse the descriptive statistics of the dependent (performance measures) and independent variable (activity, size, independence, and tenure), with their results shown in Table 2 to 5 below:

Table 2 presents the summary statistics for the six performance variables and the four board structure variables. FFO_PS has a mean of 1.8712 and a standard deviation of 2.9228, with values ranging from -9.90 to 15.20. DIV_YIELD averages 8.29%, while ROA, ROE and ROIC average 4.19%, 6.98% and 4.93% respectively. EPS shows a mean of 1.9844 and the widest dispersion (SD = 6.8340), reflecting the heterogeneity of REIT earnings. The skewness and kurtosis statistics indicate that most variables depart from normality. EPS, FFO_PS, DIV_YIELD and B_ACTIV are positively skewed, while ROA, ROE, ROIC, BIND and BOARD_TEN are negatively skewed. ROIC shows extreme kurtosis (297.6), driven by a small number of large negative values. Among board structure variables, B_SIZE averages 9.36 directors and BIND averages 58.26%. B_ACTIV, measured by the number of board meetings, averages 6.35 meetings per year, while average board tenure (BOARD_TEN) is 6.50 years.

Table 3 reports the Pearson and Spearman pair-wise correlation coefficients respectively. The two methods yield consistent conclusions. The accounting-based performance measures (ROA, ROE, ROIC and EPS) are positively and highly correlated with each other, with ROA and ROE displaying a correlation of 0.94 in both matrices. FFO_PS correlates positively with ROA (0.28), ROE (0.28) and EPS (0.45) and negatively with DIV_YIELD (-0.27), all significant at 1% level. Among the board structure variables, B_ACTIV is negatively related to ROA (-0.25) and ROE (-0.22), suggesting that more frequent board meetings are associated with lower accounting profitability. BIND shows a positive correlation with B_ACTIV (0.27) and BOARD_TEN (0.22). The correlation between any two independent variables remains below 0.28, well within the 0.80 threshold often cited for multicollinearity concerns. The Spearman matrix corroborates these patterns and identifies a positive ROIC-DIV_YIELD rank correlation (0.24) and negative ROIC associations with B_ACTIV (-0.15) and BIND (-0.14).

Following the analysis of the Pearson and Spearman correlation matrices highlighting relationships and potential multicollinearity among variables, it is important to assess the extent of multicollinearity using the Variance Inflation Factor (VIF). VIF measures multicollinearity in a regression model (Shrestha, 2020). Multicollinearity occurs when independent variables in the model are highly correlated, leading to unreliable estimates of regression coefficients (Shrestha, 2020). A VIF value greater than 10 indicates significant multicollinearity, while values below 10 suggest negligible collinearity (Putri et al., 2023).

Table 4 presents the variance inflation factor (VIF) and its reciprocal (1/VIF) for each board structure variable. The reported VIF values range from 1.026 (B_SIZE) to 1.134 (BIND), with a mean VIF of 1.080. All values are far below the conventional cut-off of 10 (and below the more conservative 5), and the corresponding 1/VIF (tolerance) values exceed 0.88. This confirms that multicollinearity among the four board structure variables is not a concern in the panel regression models

Table 5 reports the Arellano-Bond two-step difference GMM estimates of the relationship between board structure and REIT performance for the six dependent variables. The estimator first-differences the panel to remove firm fixed effects and uses suitably lagged values of the dependent variable (from lag 2 onward) as internal instruments, treating the four board structure variables as strictly exogenous. Standard errors are Windmeijer-corrected for finite-sample bias in the two-step estimator. T-statistics are reported in parentheses below each coefficient.

The lagged dependent variable enters significantly in the ROA ($\beta =$0.286, $t=$ 2.63), ROE ($\beta$ = 0.391, $t$ = 4.08), ROIC ($\beta$ = -0.021, $t$ = -5.34) and EPS ($\beta$ = 0.631, $t$ = 8.95) specifications, confirming that REIT performance is dynamically persistent and validating the dynamic specification. The lag is not significant for FFO_PS ($t$ = 0.06) or DIV_YIELD ($t$ = 1.34).

The results of board structure variables indicate that board BIND exerts a positive and significant influence on FFO_PS ($\beta$ =5.859, $t$ = 2.47), suggesting that REITs with a higher proportion of independent directors generate higher funds from operations per share. B_SIZE is positively associated with DIV_YIELD ($\beta$ = 0.015, $t$ = 2.15) at the 5% level. BOARD_TEN is negatively related to ROIC (-$\beta$ = 0.009, $t$ = -2.18), implying that longer-serving boards are associated with marginally lower returns on invested capital. The remaining board structure coefficients do not attain conventional significance, indicating limited direct effects of board activity and board size on accounting profitability in the JSE REIT sample after controlling for firm dynamics and unobserved heterogeneity.

The Wald chi-square tests reject the null of joint insignificance across all six models at the 5% level or better. The chi-square statistics range from 13.66 (FFO_PS, p = 0.018) to 657.81 (EPS, p < 0.001), providing evidence that the included variables jointly explain variation in REIT performance.

Table 5 also summarises the diagnostic tests for the six difference GMM models. Three tests are reported. The Arellano-Bond test for first-order autocorrelation in first-differenced residuals (AR(1)) rejects the null of no serial correlation in the ROA, ROE and EPS models at conventional levels, which is the expected outcome under a valid GMM specification. The Arellano-Bond test for second-order autocorrelation (AR(2)) returns p-values above 0.10 in all six specifications, indicating no evidence of serial correlation in the underlying levels equation and supporting the validity of the moment conditions. The Hansen J-test of over-identifying restrictions returns p-values close to one in all models, which fails to reject the null of joint instrument validity. The unusually high p-values are consistent with instrument proliferation in a panel of 30 cross-sectional units (see Roodman,2009) and should be interpreted with caution; the AR(2) test results provide the more reliable confirmation of identification.

5. Discussion

The empirical findings indicate that specific board characteristics influence different dimensions of REIT performance rather than exerting a uniform effect across all financial measures. The strongest result relates to board independence, which is positively and significantly associated with funds from operations per share (FFO_PS). The results show that an increase in the proportion of independent directors improves recurring operating cash flow, suggesting that independent boards strengthen monitoring, enhance operational efficiency, and enforce greater capital discipline. This finding is consistent with agency theory, which posits that independent directors reduce agency costs by constraining managerial opportunism and improving oversight (Jensen & Meckling, 1976). The evidence further supports international and South African governance literature that links stronger external governance structures to improved operating performance in REITs and listed firms (Yeon et al., 2021; David et al., 2021; Noguera, 2024; Viviers et al., 2023; Mofokeng, 2024).

Board size is positively related to dividend yield, with the results indicating that larger boards are associated with higher shareholder distributions. Given the REIT requirement to distribute a substantial proportion of taxable income, the effect of board size appears to operate through improved income generation, property portfolio management, and strategic decision-making rather than through discretionary payout policy. Resource dependence theory provides a useful explanation, as larger boards may contribute broader expertise, stronger networks, and improved access to capital markets (Pfeffer & Salancik, 1978). Recent South African evidence supports this argument, particularly within the REIT sector where larger and more diverse boards have been associated with stronger operational and earnings performance (Mofokeng, 2024). Similarly, South African governance research highlights that board size and composition can enhance firms’ propensity to distribute dividends and strengthen governance effectiveness in listed companies (Scholtz & Kieviet, 2017; Viviers et al., 2023). These findings suggest that in capital-intensive sectors such as REITs, the resource and advisory benefits of larger boards may outweigh the coordination and monitoring costs commonly associated with board expansion (Dalton et al., 1999; Hasan et al., 2024; Zureigat et al., 2024).

Board tenure demonstrates a negative and statistically significant relationship with return on invested capital (ROIC). The findings suggest that longer-serving directors may become entrenched and less willing to challenge existing management decisions or redeploy capital away from underperforming assets. In the REIT environment, this may result in delayed property disposals, overinvestment in mature assets, and weaker capital allocation decisions. The results support the entrenchment perspective advanced in prior governance literature, which argues that excessively long tenure may weaken board independence and reduce strategic adaptability (Huang & Hilary, 2018). More recent governance evidence similarly suggests that excessive tenure may increase management familiarity, reduce board vigilance, and increase the risk of governance failures or misconduct (Haider et al., 2025; Zeng et al., 2025). The average tenure observed in the sample closely matches the threshold identified in international studies where tenure effects begin to turn negative.

Board activity, measured by the number of annual board meetings, does not exhibit a statistically significant relationship with any of the performance indicators. This suggests that meeting frequency alone is an inadequate proxy for effective governance because it fails to capture meeting quality, strategic depth, or the effectiveness of board deliberations. The findings also support prior evidence that board meetings are often reactive responses to poor performance or crises rather than proactive drivers of firm value (Vafeas, 1999). Recent evidence further suggests that excessive board activity or director busyness may weaken monitoring effectiveness when directors become overstretched across multiple responsibilities (Venkatesh et al., 2025; Cao & Nguyen, 2025; Alomair & Al Naim, 2025). Other non-significant relationships reinforce the interpretation that board characteristics influence distinct performance channels rather than all measures simultaneously. Independence primarily affects operating cash flow, board size influences distributable income, and tenure affects capital allocation efficiency.

The dynamic panel results further indicate that REIT performance is persistent but mean-reverting across several accounting measures, particularly ROA, ROE, and EPS. EPS exhibits the strongest persistence, reflecting the stabilising effects of long-term lease structures and relatively fixed administrative costs. The negative autoregressive coefficient for ROIC suggests cyclical portfolio rebalancing, where unusually high returns in one period are followed by normalisation in subsequent periods. This pattern is consistent with evidence from international and South African REIT markets (Ling & Naranjo, 2015; Boshoff & Cloete, 2012). In contrast, FFO_PS and dividend yield do not display significant persistence, likely because these measures are closely linked to contractual lease income and mandatory distribution requirements.

6. Conclusions

We examined whether board structure is associated with the performance of South African REITs, focusing on board activity, board size, board independence, and board tenure over the period 2013 to 2025. Using a two-step difference GMM estimator that accounts for endogeneity, unobserved heterogeneity, and performance persistence, the analysis provides REIT-specific evidence on how board design relates to both cash-flow-based and accounting-based performance measures.

The findings point to specific governance practices that matter for REIT boards in different ways. Each board characteristic operates through a distinct channel, and the effect surfaces in the performance metric most sensitive to that channel.

Board independence shows the clearest performance link. The proportion of independent directors is positively and significantly associated with funds from operations per share. A 10% rise in independence lifts FFO_PS by approximately 0.59 cents per share in the JSE REIT sample, holding firm dynamics and unobserved heterogeneity constant. Boards should maintain a clear independent majority, approximately 6 independent directors on a ten-member board, with relevant experience in property markets, capital markets, and risk oversight. Independence that enables informed challenge of executive decisions translates directly into stronger recurring operating cash flow.

Board size relates positively to dividend yield. A one-director increase is associated with a 1.5% rise in dividend yield. In the sample, board size averages 9.36 directors and ranges from 5 to 16 directors. A moderately sized board of approximately 9 to 12 directors gives REITs access to the breadth of expertise required for property valuations, debt refinancing, lease structuring, and disposal decisions that drive distributable income, without crossing into the coordination problems documented in larger industrial samples. The result contrasts with the common preference for smaller boards in non-financial settings and underscores the importance of sector-specific governance evidence.

Board tenure carries a measurable cost for capital efficiency. Average tenure beyond approximately 6 years is associated with lower return on invested capital. A one-year rise in average tenure reduces ROIC by -0.9%. The sample mean tenure of 6.5 years sits at the inflection point identified in the international entrenchment literature. The evidence supports active tenure management through structured board evaluation, planned succession, and director rotation that takes effect once average service approaches 6 years. This approach preserves institutional knowledge while protecting capital allocation discipline.

Board activity, measured by meeting frequency, shows no statistically significant effect on any of the six performance measures. The mean of 6.4 meetings per year in the sample is broadly aligned with practical guidance, but the regression evidence does not support frequency as a governance lever. The result reinforces the view that meeting count is a poor proxy for board engagement. What matters is the substance of the agenda and the discipline of decision-making. Boards should concentrate on the quality of meetings dedicated to capital allocation, funding structures, portfolio transactions, and dividend policy, rather than on the number of meetings held.

The lagged dependent variable is statistically significant in four of the six specifications, with autoregressive coefficients of 0.286 for ROA, 0.391 for ROE, 0.631 for EPS, and -0.021 for ROIC. REIT performance is therefore dynamically persistent, validating the choice of a dynamic panel estimator. The high EPS persistence reflects the smoothing effect of long-dated lease contracts and sticky cost structures. The negative ROIC coefficient is consistent with portfolio rebalancing as REITs redeploy capital after periods of unusually high or low returns.

For investors and regulators, the results demonstrate concrete signals for assessing board effectiveness in the listed property sector. Genuine independence, demonstrated through informed challenge of management rather than formal compliance, is the strongest single indicator of REIT operating performance. Active tenure management, including periodic review and planned succession, protects against entrenchment and supports capital efficiency. Board size within the range observed in this sample supports distribution capacity. Meeting frequency, by contrast, should not be relied upon as a governance metric in this sector.

These findings show that board structure is not a neutral or purely formal feature of governance. It shapes financial outcomes that matter to investors, lenders, and regulators. The evidence is specific to the South African listed property market over 2013 to 2025 and moves the conversation beyond generic governance prescriptions toward sector-tailored guidance.

However, several limitations qualify the findings. The performance and governance measures used, though standard in the literature, do not capture qualitative dimensions of governance such as board culture, chair leadership, or boardroom dynamics. The Hansen test p-values approach unity in all six specifications, a pattern consistent with instrument proliferation in a 30-firm panel and one that warrants robustness checks using collapsed or restricted instrument sets. The cross-sectional dimension is limited, restricting the statistical power available for sub-sample analysis by REIT type or by listing tier. Future research could extend the analysis to other emerging market REITs, incorporate additional governance variables such as board diversity, audit committee effectiveness, and CEO duality, and adopt mixed-method designs that pair the quantitative evidence with qualitative insights drawn from director interviews and board-pack analysis.

This study affirms that board structure plays a decisive role in shaping the performance of South African REITs. Independent boards lift operating cash flow. Moderately sized boards support dividend capacity. Longer-tenured boards weaken capital efficiency. Meeting frequency does not predict performance. Strong board design is therefore central to investor confidence and to unlocking the long-term value of the listed property sector. The evidence presented here supports the view that governance is a driver of value creation in REITs rather than a compliance exercise.