Investing in Residential Real Estate: Understanding Homebuilder Exchange-Traded Fund Performance

Abstract

Homebuilder ETFs provide investors with a diversified portfolio of residential construction and sales companies which reduces risks associated with individual stock selection in the sector. This study examines the net monthly returns of homebuilder exchange-traded funds (ETFs) through various performance evaluation models and market situations. The results reveal that these ETFs outperformed benchmark indices in absolute returns. Despite homebuilding being part of the real estate sector, the correlation between monthly returns of homebuilder ETFs and the Dow Jones US Real Estate Index, though positive, is not very high. The performance of ETFs varied across market conditions, demonstrating both outperformance and underperformance compared to U.S. stocks. During the COVID-19 pandemic, homebuilder ETFs displayed a decline, trailing behind U.S. equities in both absolute returns and risk-adjusted performance. This result emphasizes their vulnerability during economic crises. Utilizing a modified version of the Carhart factor model, significant exposure of real estate ETFs to the stock market was observed. Moreover, an assessment of ETF portfolio managers’ skills indicated proficiency in security selection but limited capabilities in market timing. Homebuilder ETFs pose higher downside risks than other indices, evident in their elevated Value at Risk (VaR) and Conditional Value at Risk (CVaR) values.

1. Introduction

Homebuilder exchange-traded funds (ETFs) represent a strategic avenue for investors by focusing on companies involved in residential home construction and sales. These funds offer a diversified portfolio of stocks related to homebuilding, effectively mitigating the risks inherent in individual stock selection within the sector.

These ETFs encompass a wide spectrum of companies engaged in various facets of home construction. The major categories of companies contained in the top ten holdings of the ETFs include homebuilders; building material suppliers; insulation product manufacturers; roofing manufacturers; paint, stain, and coating companies; HVAC manufacturers; appliance and furnishing manufacturers; and construction materials and aggregates. Homebuilder ETFs provide investors with access to the residential real estate market without the need to cherry-pick individual stocks. This broad exposure fosters diversification within the sector and unlocks potential growth opportunities. A review of the top ten companies in the ETFs by category is shown in

Table 1. Companies included in the top ten holdings of the ETFs by major categories. There were other companies listed in the top ten holdings but were not included in the major categories.

Category	Homebuilders	Building Supplies	Insulation	Paint, Stain, and Coating
	Daiwa House Industry	Builders FirstSource	Kingspan Group	Sherwin Williams
	D.R. Horton	Home Depot	Owens Corning
	KB Homes	Lowe’s	TopBuild
	Lennar Corp.
	Meritage Homes Corp.
	Nippon Building Fund
	NVR
	Pulte Group
	Toll Brothers
	Tri Pointe Homes
Category	Roofing	HVAC Equipment	Appliance and Furnishings	Construction Materials & Aggregates
	Carlisle Companies	Carrier	Williams-Sonoma	Martin Marietta
	Owens Corning	Daikin Industries		U.S. Lime & Minerals
		Johnson Controls		Vulcan Materials
		Lennox
		Trane

.

Despite extensive research on real estate mutual funds and general real estate ETFs, the specific performance of homebuilder ETFs remains largely unexplored. Understanding the dynamics of homebuilder ETFs is vital for gaining insights into the health of the housing market and the broader economy.

Distinguishing themselves from broader real estate ETFs, homebuilder ETFs concentrate specifically on companies involved in home construction and residential real estate development. This specialized focus offers a unique perspective on performance drivers and trends within the homebuilding sector, potentially differing from broader real estate indices or real estate mutual funds.

To bridge this research gap, we conducted a thorough analysis of the risk-adjusted performance of homebuilder ETFs from February 2006 to April 2023. Spanning multiple market cycles and significant economic events like the 2008 economic crisis and the COVID-19 pandemic, this examination aims to provide a comprehensive understanding of the performance dynamics of homebuilder ETFs and their resilience to market disruptions.

Additionally, comparing the performance of homebuilder ETFs against broad stock market indices such as the Russell 3000 Index serves as a benchmark to evaluate the value proposition of these ETFs. This comparison informs investment decisions and sheds light on the effectiveness of active portfolio management strategies within the real estate sector.

Our study is significant given the importance of the homebuilding industry in the U.S. economy. Historically, the homebuilding industry has been a cornerstone of economic growth, job creation, and infrastructure development. However, recent trends suggest significant challenges, particularly in terms of profitability.

Despite near-zero interest rates during the pandemic, homebuilders have faced obstacles to sustained profitability due to rising inflation and price-based competition among regional and small-scale players. Looking ahead, a cooling housing market and rising expenses are expected to continue hindering profit growth in the homebuilding industry.

Despite the industry boasting revenue estimated at USD 123 billion in 2023 and employing over 700,000 workers, the outlook for profitability remains uncertain.1 Given the formidable challenges confronting the homebuilding sector, it becomes imperative to grasp the performance of exchange-traded funds investing in this industry. Understanding the dynamics of these funds could offer crucial insights into navigating the uncertainties and optimizing investment strategies within the homebuilding domain.

Through an examination of the performance of homebuilder ETFs and a comparison with broader market indices, our study contributes to the existing body of research on real estate investment vehicles, offering valuable insights for investors and portfolio managers navigating the real estate market.

The structure of this study is as follows: Section 2 provides a summary of previous research on ETF performance. Section 3 elucidates various models for evaluating performance. Section 4 delves into the data employed in this study. Section 5 presents a summary of empirical findings. Finally, Section 6 offers a summary and conclusion for our study.

2. Previous Studies

In our review of the literature pertaining to investment vehicles focusing on homebuilder companies, we found a lack of specific studies addressing exchange-traded funds (ETFs) in this sector. Our investigation did, however, uncover comprehensive research analyzing the performance of real estate mutual funds and ETFs.

The research by Malhotra (2024) demonstrated that real estate ETFs produced positive alphas, yet their performance varied with different market conditions. The COVID-19 pandemic revealed real estate ETFs’ crisis vulnerability through their underperformance against both U.S. and global equities. Portfolio managers displayed successful security selection abilities according to the Carhart four-factor model but demonstrated poor market timing performance in terms of market exposure. Goodwin et al. (2021) conducted a thorough assessment covering 34 real estate ETFs from May 2003 to September 2019. Employing Sharp, Sortino, and Omega ratios, their analysis revealed that while real estate ETFs demonstrated higher monthly returns compared to S&P 500 ETFs (IVV and RSP), they also exhibited slightly heightened volatility. Notably, during the 2007–2009 financial crisis, real estate ETFs experienced increased risk and substantial losses.

Conversely, recent studies on real estate mutual funds have yielded inconclusive findings regarding their performance. Malhotra (2023), utilizing three-, four-, and five-factor models, observed a significant influence of the stock market on real estate mutual fund performance, contradicting prevailing investor beliefs. Chou and Hardin (2014) noted that while real estate mutual fund returns tended to exceed benchmark pre-expenses, they struggled during the financial crisis to maintain parity or surpass benchmarks.

Bond and Mitchell (2010) reported limited evidence of real estate mutual fund managers generating excess risk-adjusted returns, without any discernible performance persistence. Kaushik and Pennathur (2012) concluded that while real estate mutual fund managers did not outperform the market from 1990 to 2008, they did exhibit outperformance up until 2006. The poor performance for the overall period was due to the 2007-2008 economic downturn.

Moreover, recent investigations by Lantushenko and Nelling (2020) observed a decline in performance among real estate fund managers, with geographically diversified real estate mutual funds failing to outperform. MacGregor et al. (2021) did not find substantial evidence supporting the outperformance of global real estate mutual funds compared to other actively managed funds, even in the long term.

Given the conflicting findings on risk-adjusted performance, market timing, and the selectivity of real estate mutual funds in the existing literature, our study aims to reassess these aspects by evaluating the monthly average returns of homebuilder ETFs over a 17-year period from February 2006 to April 2023. This reassessment extends prior research by examining the risk-adjusted performance of real estate ETFs during significant economic events such as the 2007 economic crisis and the COVID-19 pandemic, which notably affected sectors like commercial leasing, apartment rentals, and shopping malls due to widespread lockdowns and closures.

3. Data

The dataset utilized in this study was sourced from Morningstar Direct, encompassing monthly returns for homebuilder ETFs spanning from February 2006 to April 2023. It is important to note that our dataset did not include any homebuilder ETFs before November 2005. Over this period, the number of homebuilder ETFs increased from one in February 2006 to seven by April 2023, indicating a very small expansion of homebuilder ETF offerings in the market.

Table 2. Summary statistics of the average monthly rates of returns by each year for homebuilder ETFs. The S&P Homebuilders Index, DJ Real Estate Index, and Russell 3000 Index for the period of February 2006 to April 2023.

	Homebuilder ETFs	S&P Homebuilders Select Industry Index	DJ US Real Estate Index	Russell 3000 Index
February 2006 to April 2023
Mean	0.86	0.54	0.68	0.68
Standard deviation	8.98	8.13	6.20	4.62
Average return per unit of risk	0.10	0.07	0.11	0.15
July 2007 to March 2009
Mean	−4.01	−4.16	−3.99	−2.80
Standard deviation	10.50	12.01	11.28	6.23
Average return per unit of risk	−0.38	−0.35	−0.35	−0.45
February 2020 to January 2021
Mean	3.84	3.00	−0.31	1.71
Standard deviation	18.69	13.44	7.65	7.95
Average return per unit of risk	0.21	0.22	−0.04	0.22
February 2021 to April 2023
Mean	1.01	0.91	0.61	0.42
Standard deviation	9.61	7.97	4.91	5.28
Average return per unit of risk	0.10	0.11	0.12	0.08

presents summary statistics of the average monthly rates of returns for homebuilder ETFs, the S&P Homebuilders Select Industry Index, DJ US Real Estate Index, and Russell 3000 Index across various periods from February 2006 to April 2023.

The S&P Homebuilders Select Industry Index tracks the performance of U.S. homebuilding companies, encompassing entities involved in land acquisition, home construction, and property sales. Its purpose extends to aiding investors in assessing the vitality of the housing market and informing investment decisions. A rising index indicates investor confidence and the potential for elevated returns, while a decline may signal market apprehensions or obstacles.

Similarly, the Dow Jones U.S. Real Estate Index functions as a financial gauge monitoring the performance of real estate companies across the United States. It typically comprises publicly traded firms engaged in various aspects of the real estate industry, including property development, management, investment, and related services.

Conversely, Russell 3000 represents a broad index encompassing around 3000 of the largest publicly traded U.S. stocks. Encompassing large, mid, and small-cap stocks, it provides a comprehensive benchmark for U.S. equities, offering insights into the overall performance of the U.S. stock market.

During the sample period spanning from February 2006 to April 2023, homebuilder ETFs exhibited a mean monthly return of 0.86% with a standard deviation of 8.98%, indicating moderate volatility. The S&P Homebuilders Select Industry Index displayed a similar mean monthly return of 0.54%, accompanied by a slightly lower standard deviation of 8.13%. Conversely, both the DJ US Real Estate Index and the Russell 3000 Index demonstrated mean monthly returns of 0.68%, with the Russell 3000 Index presenting the lowest volatility at 4.62%.

The average return per unit of risk for the entire period was calculated based on the relationship between return and standard deviation. homebuilder ETFs yielded a return of 0.10 per unit of risk, while the S&P Homebuilders Select Industry Index provided 0.07, the DJ US Real Estate Index 0.11, and the Russell 3000 Index 0.15.

Throughout the market downturn from July 2007 to March 2009, all indices and homebuilder ETFs experienced negative mean monthly returns, indicating significant losses. homebuilder ETFs and related indices incurred more substantial losses compared to broader market indices such as the Russell 3000 Index. The average return per unit of risk during this downturn was notably negative for all assets, suggesting high risk relative to returns.

Between February 2020 and January 2021, homebuilder ETFs and the S&P Homebuilders Select Industry Index demonstrated higher mean monthly returns compared to broader market indices. However, they also displayed significantly higher volatility, particularly evident in the wide standard deviation of homebuilder ETFs. The average return per unit of risk underscores the inherent trade-off between risk and return, with homebuilder ETFs offering relatively higher returns relative to the risk incurred.

From February 2021 to April 2023, homebuilder ETFs, the S&P Homebuilders Select Industry Index, and the DJ US Real Estate Index exhibited positive mean monthly returns. Homebuilder ETFs and the S&P Homebuilders Select Industry Index maintained higher returns compared to the broader market represented by the Russell 3000 Index. The average return per unit of risk remained relatively stable across the indices and ETFs, indicating a consistent risk-return profile over this period. The analysis highlights the role of market cycles and economic indicators in influencing the performance of this sector.

4. Model

This study evaluates the risk-adjusted performance of homebuilder ETFs in comparison to the Dow Jones U.S. Homebuilder Index, the Dow Jones U.S. Real Estate Index, and Russell 3000, which serves as a proxy for U.S. stocks. Monthly returns from February 2006 to April 2023 form the basis of this evaluation.

Through evaluating the risk-adjusted performance of homebuilder ETFs in relation to these indices, the study seeks to provide insights into the dynamics of the real estate and homebuilder sectors within the broader context of the U.S. financial market landscape.

For risk-adjusted performance assessment, we used the Sharpe ratio (Sharpe, 1966), Sortino ratio (Sortino & Van Der Meer, 1991), and Omega ratio (Keating & Shadwick, 2002). To determine whether homebuilder ETFs produced positive alpha, we broadened the Carhart four-factor model by adding a fifth factor. The Carhart four-factor model proves to be an appropriate choice for homebuilder ETF evaluation because of its detailed scope and multifaceted performance analysis framework. Traditional methods, such as the Capital Asset Pricing Model (CAPM), focus only on market risk (beta), whereas the Carhart model expands its framework by including additional variables like size, value, and momentum, which allows for a more detailed ETF performance analysis.

The Carhart model captures factor-based trends in ETF returns through its inclusion of size, value, and momentum factors. The momentum factor helps determine how efficient different investment approaches are at producing returns. The extended model clarifies the risk-adjusted performance of ETFs while revealing how their investment strategies function in the real estate and homebuilder sectors.

4.1. The Sharpe Ratio, Sortino Ratio, and Omega Ratio

Portfolio returns might increase because of the higher risk taken by the investor. We used the Sharpe ratio (Sharpe, 1966) to measure how well investments compensate investors for assumed risks. The Sharpe ratio functions as an indicator of investment success, which is the result of dividing the risk premium by the portfolio return standard deviation.

The Sharpe ratio compares equity investment performance with a risk-free investment while factoring in the increased risk of holding equities. When a negative Sharpe ratio is obtained, it shows that a risk-free investment delivers better risk-adjusted returns. Risk-adjusted returns are considered good when the Sharpe ratio reaches one or more.

The Sharpe ratio indicates that portfolio diversification results in a reduced standard deviation of portfolio returns. When diversifying across multiple securities, the portfolio becomes insulated from company-specific risks, which results in lower standard deviation values. The Sharpe ratio serves as a summary indicator of portfolio performance breadth across all investments.

The Sharpe ratio is, however, affected by a benchmark problem since the standard deviation of the market portfolio is contingent upon the selected proxy. As a result, both the Sharpe ratio of the benchmark portfolio and its comparison with the Sharpe ratio of the portfolio under consideration are influenced.

The Sortino Ratio (Sortino & Van Der Meer, 1991) extends the Sharpe ratio framework by differentiating between positive and negative volatility. By distinguishing between upward and downward volatility movements, this method computes risk-adjusted returns without penalizing positive performance. When a portfolio shows a higher Sortino ratio, it demonstrates better investment performance.

The Sortino ratio exclusively zeroes in on downside volatility, which stands apart from the Sharpe ratio that evaluates total volatility. The Sortino ratio evaluates investment performance by excluding upward volatility and computing only downside standard deviation, which makes it ideal for high-volatility investment portfolios.

The Omega ratio (Keating & Shadwick, 2002) determines financial asset performance by comparing their returns against investment risk levels. The Omega ratio calculates the ratio of positive returns to negative returns and evaluates every moment of the return distribution. The expanded analysis approach supplies investors with deeper understanding which exceeds the capabilities of the Sharpe ratio.

4.2. Multi-Factor Model

Investment performance evaluation relies heavily on factor models which analyze returns according to their sensitivity to different risk factors. When excess performance cannot be accounted for by changes in established factors, it becomes the investment’s alpha. For the alpha (α) calculation of homebuilder ETFs, we implement a revised Carhart four-factor model which includes five distinct factors.

The four-factor model developed by Carhart (1997) builds on the Capital Asset Pricing Model (CAPM) as well as Fama and French’s (1993) three-factor model by including extra factors that explain investment return variations beyond mere market risk. Investment performance can be fully understood and evaluated through the Carhart model, which assesses size, value, momentum, and market risk factors.

The four-factor model established by Carhart (1997) expands on the Fama–French three-factor model through the inclusion of a momentum (MOM) factor. The extension analyzes return disparities between portfolios consisting of previous 12-month top performers and bottom performers within a market equilibrium framework that contains four separate risk factors.

Furthermore, homebuilders heavily rely on borrowing for projects and land purchases, with borrowing costs influenced by the shape of the yield curve. A steep curve may raise costs, impacting profitability, while mortgage rates, tied to long-term interest rates affected by the curve, influence homebuyers’ demand. Inverted curves signal economic slowdowns, reducing housing demand and affecting homebuilders’ decisions. Therefore, our analysis of homebuilder ETFs’ performance incorporates an additional variable—the slope of the U.S. Treasury yield curve.

Equation (1) provides the modified version of the Carhart four-factor model and is estimated using monthly returns:

$$\begin{array}{cc}\hfill R_{i,t}-R_{f,t}={\alpha }_i+& {\beta }_1\times \left(R_{m,t}-R_{f,t}\right)+{\beta }_2\times SM{B}_t+{\beta }_3\times HM{L}_t\hfill \\ \hfill & +{\beta }_4\times MO{M}_t+{\beta }_5\times SlopeofYieldCurv{e}_t+{\epsilon }_{i,t}\hfill \end{array}$$

(1)

where R_i,t = the percentage return for firm i in month t.

R_f,t = the yield on U.S. Treasury bill month t.

R_m,t = the return on the Center for Research in Security Prices (CRSP) value-weighted index for month t.

$\left(R_{m,t}-R_{f,t}\right)=$ the market risk factor variable representing both the additional returns of the whole market and the total risk associated with stock market investments.

SMB_t (small minus Big) = the capitalization factor realization, which is measured by subtracting large-cap return from small-cap return during month t. The SMB factor evaluates past performance discrepancies between stocks of small-cap companies and those of large-cap companies. The SMB factor calculation takes the return from a small-cap portfolio and subtracts the return from a large-cap portfolio. When SMB is positive, it shows that small-cap stocks have performed better than large-cap stocks.

HML_t (high minus low) = the HML (high minus low) factor identifies how value stocks perform compared to growth stocks during month ‘t’, and this measurement is known as value factor realization which looks at the return difference between both types of stocks. The calculation of this difference requires subtracting the return from a portfolio containing value stocks with low price-to-book ratios from the return of a portfolio made up of growth stocks with high price-to-book ratios. When HML values are positive, it shows that value stocks have performed better than growth stocks.

MOM_t = the momentum factor at time t shows how well stocks with previous strong performances have performed. This factor measures the tendency of stocks with recent positive returns to sustain their strong performance and the tendency of stocks with recent negative returns to maintain poor performance.

Slope of the yield curve = the slope of the yield curve is measured as the yield on 30-year Treasury bonds minus the yield on 30-day Treasury bills at time t.

ε_i,t = an error term.

The size factor, SMB, which measures the difference between small-cap returns and large-cap returns, displays a positive slope βs for small-company stocks while showing a negative slope for large-company stocks. The value component shows sensitivity when βv estimates are positive while showing negative estimates for the βv point to growth factor sensitivity. When the intercept (α) is positive, it demonstrates better results than the three-factor model, but when the intercept (α) is negative, it shows worse performance.

The momentum factor identifies the pattern which shows that assets with strong recent performance maintain good performance, while those with poor recent performance continue to show poor results. This factor uses historical returns from the past 6 to 12 months to evaluate momentum.

5. Robustness Check

To ensure the robustness of our findings, we also computed the Generalized Sharpe Ratio (GSR) and Calmar Ratio to evaluate the risk-adjusted performance of homebuilder exchange-traded funds. Furthermore, to fortify the reliability of our findings, we extended our evaluation to include rolling alphas, conditional alphas, and metrics for market timing and selectivity specifically tailored for homebuilder exchange-traded funds. These supplementary analyses are intended to solidify the robustness of our conclusions, ensuring their stability across a range of methodological approaches and scenarios.

5.1. Dynamic Alpha Evaluation

Within the realm of financial performance assessment, a common pitfall arises from relying solely on a single alpha value computed across the entire evaluation period. This method overlooks the dynamic nature of performance, potentially masking periods of both strength and weakness. Thus, we opt for computing rolling monthly alphas—a series of alpha values calculated over consecutive, abbreviated intervals—to capture the evolving essence of investment performance.

Rolling alpha computation involves calculating the alpha over successive, typically overlapping time frames. This technique provides a more nuanced understanding of performance progression, revealing how alpha fluctuates over time. A typical approach involves computing rolling alphas over 12-month intervals, with the window shifting forward by one month each time. This systematic method enables a detailed analysis of performance trends, aiding in the identification of periods characterized by strong or weak performance relative to the benchmark.

The examination of rolling alphas is crucial for detecting shifts in market conditions and changes in investment strategies. Notably, persistent negative rolling alphas can serve as an early indication, signaling potential issues with the chosen investment strategy or managerial competency. Embracing this proactive approach allows for timely adjustments to mitigate risks.

In this study, we utilize the extended Carhart’s factor model to compute rolling monthly alphas. This ensures a comprehensive evaluation of performance stability over time and facilitates meaningful comparisons against benchmark indices. By adopting such a multifaceted methodology, we enhance our capacity to evaluate the effectiveness of our investment strategies and promptly respond to evolving market dynamics.

5.2. Conditional Factor Models

Financial studies evaluating managed funds often grapple with the challenge of metrics sensitive to temporal fluctuations in risks and risk premia. Scholars acknowledge variations in ETFs’ risks and risk premia, advocating for approaches that account for common time variations alongside average performance (Chen et al., 2023; Hurlin et al., 2019). By integrating a vector of lagged public information variables, this method adapts the traditional Jensen alpha model (Jensen, 1968), substituting unconditional betas with time-varying conditional expected returns and betas, a strategy historically successful in predicting stock returns (Christopherson et al., 1998; Elton et al., 2010; Ferson & Schadt, 1996).

To address potential bias, we adopt the model proposed by Ferson and Schadt (1996) and employ the unconditional approach. We integrate one-month lagged public information variables, encompassing the three-month Treasury bill rate (TR3M), the term structure slope (SLOPE), the corporate bond market quality spread (QS), and the dividend yield on the S&P 500. By incorporating these variables, our model captures market dynamics influencing managed funds’ risk-adjusted performance, enhancing the robustness of our analysis, and furnishing more reliable insights for investors. The resulting conditional model is shown in Equation (2), where Z_j,t−1 is the demeaned value of the unconditional elements.

$$\begin{array}{cc}\hfill R_{i,t}-R_{f,t}={\alpha }_i+& {\beta }_i\left(R_{m,t}-R_{f,t}\right)+\delta {\left\{z_{t-1}\times \left(R_{m,t}-R_{f,t}\right)\right\}}_2+{\beta }_2\times SM{B}_t\hfill \\ & +{\beta }_3\times HM{L}_t\hfill \\ & +{\beta }_4\times MO{M}_t+{\beta }_5\times Slope of Yield Curv{e}_t+{\epsilon }_{i,t}\hfill \end{array}$$

(2)

5.3. Market Timing and Selectivity

Selectivity refers to an investment manager’s ability to pick stocks that will provide the projected returns in the future. Market timing, on the other hand, refers to an investment manager’s capacity to adjust their portfolio holdings to anticipate changes in the asset portfolio or market price movement in general. Mutual fund market timing and selectivity have already been investigated (Treynor & Mazuy, 1966; Kon & Jen, 1979; Henriksson & Merton, 1981; Lee & Rahman, 1990). Previous research found that mutual fund managers had only minimal performance in market timing and selectivity. To account for market timing and selectivity, Treynor and Mazuy (1966) included a quadratic term to the Capital Asset Pricing Model (CAPM). Treynor and Mazuy (1966) introduced a quadratic term in another CAPM-based model that has become a standard for gauging timing skill to address managers’ abilities to foresee market swings. Treynor and Mazuy (1966), Kon and Jen (1979), Henriksson and Merton (1981), and Lee and Rahman (1990) found that mutual fund managers are only moderately successful in terms of market timing and selectivity. We used two models to look at market timing and selectivity. The basic model was designed by Treynor and Mazuy (1966). This model adds a quadratic component to CAPM or the market model to represent market timing and selectivity. The formula is as follows:

$$R_{i,t}-R_{f,t}={\alpha }_s+{\beta }_1\left(R_{m,t}-R_{f,t}\right)+{\beta }_2\times {\left(R_{m,t}-R_{f,t}\right)}^2+{\epsilon }_{i,t}$$

(3)

The coefficient β₂ reveals if the manager can properly anticipate market performance by assessing whether the relation between the portfolio return and the market return is non-linear. A β₂ that is both positive and significant implies superior market timing abilities. A negative and significant β₂ suggests poor market timing. If β₂ is not more than 0, the manager lacks market timing abilities. Similarly, αs denotes selectivity.

5.4. Conditional Market Timing and Selectivity

To further evaluate the security selection and marketing timing skills of homebuilder ETF portfolio managers, we developed conditional marketing and selectivity models to discern market timing and selectivity based on publicly available information. Following Ferson and Schadt (1996), Equation (4) illustrates the conditional market timing and selectivity of these funds.

$$R_{i,t}-R_{f,t}={\alpha }_i+{\beta }_i\left(R_{m,t}-R_{f,t}\right)+\delta \left\{z_{t-1}\times \left(R_{m,t}-R_{f,t}\right)\right\}+{\beta }_2\times {\left(R_{m,t}-R_{f,t}\right)}^2+{\epsilon }_{i,t}$$

(4)

5.5. Value at Risk (VaR) and Conditional Value at Risk (VaR)

Value at Risk (VaR) and Conditional Value at Risk (CVaR), alternatively referred to as Expected Shortfall (ES), are two commonly employed risk metrics in finance for evaluating potential losses within investment portfolios amid adverse market conditions. These measures facilitate the assessment of the resilience of exchange-traded fund performance by providing insights into the downside risks associated with the fund’s investments. VaR serves as a statistical measure estimating the maximum potential loss that an investment portfolio may incur over a specified time period at a given confidence level. In the context of mutual funds, VaR elucidates the potential downside risk associated with the fund’s investments. A lower VaR indicates reduced exposure to extreme losses, whereas a higher VaR suggests heightened potential losses during adverse market conditions. A comparative analysis of VaR across different mutual funds enables investors to evaluate which funds demonstrate stronger performance in terms of risk management.

Conditional Value at Risk (CVaR) extends the concept of VaR by offering insights into the expected magnitude of losses exceeding the VaR threshold. CVaR quantifies the average loss anticipated when losses surpass the VaR threshold. Similar to VaR, CVaR assists investors in understanding the potential downside risk of a fund. A fund with a lower CVaR implies that losses exceeding the VaR threshold are, on average, less severe, indicating superior risk management and more resilient performance.

6. Empirical Analysis

To commence our empirical examination of the performance of homebuilder ETFs compared to benchmark indices specific to the homebuilder industry, real estate industry, and U.S. equities, we conducted a series of analytical procedures, encompassing correlation analysis, risk-adjusted performance evaluation, and alpha computations.

6.1. Correlation

Table 3. Correlation using correlation test for homebuilder ETFs, S&P Homebuilders Index, DJ Real Estate Index, and Russell 3000 Index from February 2006 to April 2023.

	Homebuilder ETFs	S&P Homebuilders Select Industry Index	DJ US Real Estate Index	Russell 3000 Index
February 2006 to April 2023
Homebuilder Exchange-Traded Funds	1.00
S&P Homebuilders Select Industry Index	0.96	1.00
DJ US Real Estate Index	0.69	0.70	1.00
Russell 3000 Index	0.80	0.77	0.74	1.00
July 2007 to March 2009
Homebuilder Exchange-Traded Funds	1.00
S&P Homebuilders Select Industry Index	0.98	1.00
DJ US Real Estate Index	0.59	0.45	1.00
Russell 3000 Index	0.26	0.10	0.80	1.00
February 2020 to January 2021
Homebuilder Exchange-Traded Funds	1.00
S&P Homebuilders Select Industry Index	1.00	1.00
DJ US Real Estate Index	0.92	0.94	1.00
Russell 3000 Index	0.88	0.89	0.94	1.00
February 2021 to August 2023
Homebuilder Exchange-Traded Funds	1.00
S&P Homebuilders Select Industry Index	0.99	1.00
DJ US Real Estate Index	0.67	0.64	1.00
Russell 3000 Index	0.82	0.80	0.71	1.00

presents correlation coefficients between various indices related to homebuilders, the S&P Homebuilders Index, DJ Real Estate Index, and Russell 3000 Index for different time periods ranging from February 2006 to April 2023.

Homebuilder ETFs have consistently mirrored the performance of the Dow Jones U.S. Homebuilder Index, showcasing a strong correlation coefficient of 0.96 in their monthly returns. Moreover, our analysis in Table 2 reveals that homebuilder ETFs and the DJ US Real Estate Index exhibit a relatively lower correlation between their monthly returns at 0.69. The correlation between monthly returns of homebuilder ETFs and U.S. equities as proxied by the Russell 3000 Index was high at 0.80 during the sample period of February 2006 to April 2023.

Notably, during the economic crisis that unfolded from August 2007 to August 2009, the monthly returns of homebuilder ETFs exhibited a significant uptick in correlation with the Dow Jones U.S. Homebuilder Index, reaching a substantial 0.98. On the other hand, the correlation between the monthly returns of homebuilder ETFs and the DJ US Real Estate Index declined to 0.59. The correlation between the monthly returns of homebuilder ETFs and the Russell 3000 Index was very low at 0.26.

Amid the COVID-19-induced lockdowns, there was a striking surge in the correlation between monthly returns of homebuilder ETFs, the S&P Homebuilders Select Industry Index, DJ US Real Estate Index, and Russell 3000 Index with correlation coefficients of 1.00, 0.92, and 0.88, respectively. This period of extreme market volatility and uncertainty led to a convergence in the performance of these financial instruments.

However, in the post-COVID-19 vaccination era, we witnessed a substantial shift in these correlations. Specifically, the correlation between the monthly returns of homebuilder ETFs and the DJ Real Estate Index dwindled to a more modest 0.67, reflecting a departure from the tight coupling observed during the pandemic. However, the correlation between monthly returns of homebuilder ETFs and the S&P Homebuilders Select Industry Index as well as the Russell 3000 Index continued to be high at 0.99 and 0.82, respectively.

The Cumulative Wealth Index (CWI) serves as a measure to evaluate the growth of an initial investment over time, considering dividend reinvestment. We established CWI values for each investment category, calculating the return on a hypothetical USD 1000 investment made at the beginning of February 2006. Furthermore, CWI values were computed for benchmark indices mentioned in Figure 1, enabling an assessment of cumulative returns based on an initial investment.

An analysis starting in February of 2006 and ending in April of 2023, with the ending wealth being the highest for the Russell 3000 Index, followed by the DJ Real Estate Index, and the Homebulider ETFs, and the lowest wealth increase being found for the S&P Homebuilders Index.

According to the monthly returns data, an initial investment of USD 1000 in the Dow Jones U.S. Real Estate Index in February 2006 would have resulted in a total wealth of USD 2712 by April 2023. Comparing these figures with alternative investment choices highlights substantial differences in growth. For instance, directing USD 1000 towards the Russell 3000 Index in February 2006 would have led to a more modest accumulation of USD 3235 by April 2023. Conversely, investing USD 1000 in homebuilder ETFs in February 2006 would have generated an accumulated wealth of USD 2502 by April 2023. Similarly, a USD 1000 investment in the S&P Homebuilders Index during February 2006 would have yielded an accumulated wealth of USD 1517 by April 2023.

In absolute terms, homebuilder ETFs exhibited comparatively lower performance than U.S. equities in generating cumulative wealth from February 2006 to April 2023. However, homebuilder ETFs marginally underperformed the DJ Real Estate Index while significantly outperforming the S&P Homebuilders Index during the same period.

6.2. Empirical Analysis of Sharpe Ratio, Sortino Ratio, and Omega Ratio

Table 4. Summary of risk-adjusted performance of homebuilder ETFs using Sharpe ratio, Sortino ratio, and Omega ratio. Risk-adjusted performance is based on monthly returns from February 2006 to April 2023.

	Sharpe Ratio	Sortino Ratio	Omega Ratio
February 2006 to April 2023
Homebuilder Exchange-Traded Funds	0.09	0.12	1.24
S&P Homebuilders Select Industry Index	0.05	0.08	1.16
DJ US Real Estate Index	0.10	0.13	1.32
Russell 3000 Index	0.13	0.18	1.40
July 2007 to March 2009
Homebuilder Exchange-Traded Funds	−0.40	−0.42	0.38
S&P Homebuilders Select Industry Index	−0.36	−0.39	0.42
DJ US Real Estate Index	−0.37	−0.38	0.33
Russell 3000 Index	−0.48	−0.47	0.27
COVID-19-Induced Lockdowns to First Vaccination (February 2020 to January 2021)
	Sharpe Ratio	Sortino Ratio	Omega Ratio
Homebuilder Exchange-Traded Funds	0.20	0.30	1.74
S&P Homebuilders Select Industry Index	0.22	0.32	1.82
DJ US Real Estate Index	−0.04	−0.05	0.88
Russell 3000 Index	0.21	0.34	1.69
February 2021 to April 2023 (Post-COVID-19 Vaccination roll out period)
Homebuilder Exchange-Traded Funds	0.09	0.15	1.24
S&P Homebuilders Select Industry Index	0.10	0.15	1.27
DJ US Real Estate Index	0.10	0.17	1.31
Russell 3000 Index	0.06	0.09	1.15

summarizes the results for the Sharpe ratio, Sortino ratio, and Omega ratio based on monthly returns from February 2006 to April 2023.

The table provides a comparative analysis of the risk-adjusted performance of homebuilder ETFs in conjunction with other major indices utilizing three distinct metrics: the Sharpe ratio, the Sortino ratio, and the Omega ratio. It encompasses multiple discrete time periods, including the global financial crisis, the COVID-19-induced lockdowns, and the post-COVID-19 vaccination rollout period.

In the period spanning from February 2006 to April 2023, homebuilder ETFs demonstrated moderate performance in terms of risk-adjusted metrics, manifesting a Sharpe ratio of 0.09, a Sortino ratio of 0.12, and an Omega ratio of 1.24. A comparative analysis with benchmark indices revealed that these ETFs surpassed the S&P Homebuilder Index but fell short of the DJ US Real Estate Index and the Russell 3000 Index. Specifically, the S&P Homebuilder Index yielded a Sharpe ratio of 0.05, a Sortino ratio of 0.08, and an Omega ratio of 1.16 during the same period, while the DJ U.S. Real Estate Index exhibited analogous ratios of 0.10, 0.13, and 1.32, respectively. U.S. equities, as represented by the Russell 3000 Index, displayed a Sharpe ratio of 0.13, a Sortino ratio of 0.18, and an Omega ratio of 1.40, thereby indicating the superior risk-adjusted performance of U.S. equities.

Amidst the global financial crisis spanning from July 2007 to March 2009, homebuilder ETFs experienced notable performance downturns, as evidenced by negative scores across all three ratios (Sharpe ratio = −0.40, Sortino ratio = −0.42, and Omega ratio = 0.38). This suggests that during economic downturns, homebuilder ETFs may represent more volatile and riskier investments. U.S. equities demonstrated the poorest performance, with the Russell 3000 Index exhibiting a Sharpe ratio of −0.48, a Sortino ratio of -0.47, and an Omega ratio of 0.27. The S&P Homebuilders Select Industry Index generated a Sharpe ratio of −0.36, a Sortino ratio of −0.39, and an Omega ratio of 0.42 based on monthly returns from July 2007 to March 2009. During the same period, the DJ US Real Estate Index displayed similar ratios of −0.37, −0.38, and 0.33.

The period of COVID-19-induced lockdowns (February 2020 to January 2021) witnessed an enhancement in the risk-adjusted performance of homebuilder ETFs, as indicated by elevated scores across all three ratios. The ETFs generated a Sharpe ratio of 0.20, a Sortino ratio of 0.30, and an Omega ratio of 1.74. Comparable ratios for the S&P Homebuilders Select Industry Index were 0.22, 0.32, and 1.82; for the DJ US Real Estate Index, they were −0.04, −0.05, and 0.88; and for the Russell 3000 Index, they were 0.21, 0.34, and 1.69.

Following the COVID-19 vaccination rollout (February 2021 to April 2023), homebuilder ETFs maintained relatively stable risk-adjusted performance, with scores akin to those of the pre-COVID-19 era. This suggests that the sector successfully navigated the challenges posed by the pandemic and sustained its performance during the ensuing recovery period. Indeed, homebuilder ETFs, the S&P Homebuilder Index, and the DJ US Real Estate Index all outperformed U.S. equities in terms of Sharpe ratio, Sortino ratio, and Omega ratio. The Sharpe ratio, Sortino ratio, and Omega ratio for homebuilder ETFs and the S&P Homebuilder Index were 0.09, 0.15, and 1.24, while the same ratios for the DJ US Real Estate Index were 0.10, 0.15, and 1.27. Comparable ratios for the DJ US Real Estate Index were 0.10, 0.17, and 1.31, while U.S. equities generated a Sharpe ratio of 0.06, a Sortino ratio of 0.09, and an Omega ratio of 1.15. Consequently, U.S. equities delivered the lowest performance, with both the homebuilder and real estate sectors outperforming U.S. equities in terms of risk-adjusted performance.

6.3. Generalized Sharpe Ratio and Calmar Ratio2

To evaluate risk-adjusted performance for homebuilder exchange-traded funds, we also computed the Generalized Sharpe Ratio (GSR) and Calmar Ratio as core tools to validate our research findings. These additional performance evaluation metrics account for both higher moments of return distributions and the relationship between returns and drawdowns.

The Generalized Sharpe Ratio (GSR) extends the traditional Sharpe ratio and includes moments beyond mean and standard deviation like skewness and kurtosis in its calculation. The extension offers extensive performance analysis for portfolios or assets whose return distributions show deviations from normal distribution. While the S&P Homebuilder Index comes in 0.05, the Russell 3000 returns lead with 0.15, the GSR values for the ETF monthly returns reach 0.08. The DJ Real Estate Index results in 0.08. By including skewness and kurtosis, these metrics provide a more detailed performance analysis than the traditional Sharpe ratio.

The Calmar Ratio assesses the compound annual growth rate (CAGR) relative to the maximum drawdown (MDD). The Russell 3000 Index showed the best performance during this time with a Calmar Ratio of 0.13, while the high drawdown risk of the ETF and S&P Homebuilder Index makes them unappealing for safety-focused investors. The Dow Jones Real Estate Index delivered a Calmar Ratio of 0.0952. The homebuilder ETFs achieved a 5.46% CAGR yet its drawdown of −74.12% resulted in an unsatisfactory Calmar Ratio of 0.07.

6.4. Empirical Analysis of Modified Carhart Multi-Factor Model

Table 5. A summary of the results for net monthly alphas of homebuilder ETFs based on the multi-factor model (modified and extended Carhart four-factor model) used in this study. Risk-adjusted performance is measured based on monthly returns from February 2006 to April. 2023.

	February 2006 to April 2023	July 2007 to March 2009	February 2020 to January 2021	February 2021 to April 2023
Adjusted R2	0.66	0.72	0.85	0.62
Alpha	0.11	−1.27	19.79 *	0.82
Mkt-RF	1.40 ***	0.77 ***	2.47 ***	1.35 ***
SMB	0.52 ***	0.76	1.98	−0.15
HML	0.19	0.66	0.05	0.21
MOM	−0.06	−0.52 *	1.94 **	−0.44
Slope of Yield Curve	−0.17	0.36	−18.18 **	−0.61

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

summarizes the findings regarding the net monthly alphas of homebuilder ETFs based on the multi-factor model, specifically the modified and extended Carhart four-factor model, employed in this study. The analysis spans four distinct time periods: February 2006 to April 2023, July 2007 to March 2009, February 2020 to January 2021, and February 2021 to April 2023. Risk-adjusted performance is assessed using monthly returns data from these time frames.

During the period from February 2006 to April 2023, the adjusted R-squared value—a measure of the model’s explanatory power—stood at 0.66, suggesting a moderate level of explanatory efficacy. The alpha, representing the excess return of homebuilder ETFs beyond the expected level based on the model’s factors, was 0.11, indicating a slight positive alpha during this period.

Throughout the global financial crisis from July 2007 to March 2009, the adjusted R-squared value increased to 0.72, indicating a better alignment between the model and the observed data. However, the alpha decreased significantly to −1.27, signifying a considerable underperformance of homebuilder ETFs relative to the model’s factors.

The period from February 2020 to January 2021, characterized by the emergence of COVID-19-induced lockdowns, exhibited a high adjusted R-squared value of 0.85, suggesting substantial explanatory power of the model amid heightened volatility. Notably, the alpha surged to 19.79 during this period, indicating a significant outperformance of homebuilder ETFs compared to the model’s factors.

In the subsequent post-COVID-19 vaccination rollout period from February 2021 to April 2023, the adjusted R-squared value decreased to 0.62, suggesting a slightly weaker fit of the model with the observed data compared to the preceding period. Nevertheless, the alpha remained positive at 0.82, indicating a sustained outperformance of homebuilder ETFs relative to the model’s factors.

When we analyze the variables within the model, it is apparent that the market risk premium (Mkt-RF) consistently affects the returns of homebuilder ETFs across all investigated periods. Homebuilder ETFs exhibit both statistically and economically noteworthy exposure to the overall stock market, as indicated by positive coefficients on the excess market returns. Despite efforts to mitigate systematic variation with the stock market, these funds commonly possess stock market betas ranging from 0.7 to 2.47, illustrating a substantial correlation with the stock market.

Additionally, factors such as size (SMB), value (HML), momentum (MOM), and the slope of the yield curve demonstrate varying impacts on the performance of homebuilder ETFs across different time spans, as evidenced by their coefficients. SMB is a factor that captures the difference in returns between small-cap stocks and large-cap stocks. A positive coefficient on SMB indicates that small-cap stocks tend to outperform large-cap stocks, on average, over the sample period of February 2006 to April 2023.

Furthermore, the coefficient on momentum during the period of July 2007 to March 2009 is negatively and weakly statistically significant. A negative coefficient on momentum indicates that during the economic crisis period, stocks characterized by strong past performance (referred to as momentum stocks) witnessed diminished returns or even losses. This stands in contrast to the usual behavior observed in more stable market conditions, where momentum tends to exhibit a positive correlation with future returns.

The negative coefficient on momentum during the economic crisis may indicate heightened risk aversion among investors. During times of economic uncertainty and market stress, investors often favor safer assets or defensive strategies, potentially leading to decreased performance of riskier assets such as momentum stocks. Investors may have shifted focus away from high-risk, high-return strategies like momentum investing toward safer or value-oriented investments during the crisis, contributing to negative momentum in stock performance.

Conversely, the positive and statistically significant coefficient on momentum during the COVID-19-induced lockdowns, preceding vaccinations, suggests a reversal in momentum stock behavior. This contrasts with the negative coefficient observed during the economic crisis, implying increased returns for stocks with robust past performance. Government interventions and monetary policies aimed at alleviating the pandemic’s economic impact likely influenced market sentiment and momentum, bolstering investor confidence. Moreover, the transition to remote work heightened housing demand, further contributing to positive market momentum.

Furthermore, the negative and highly statistically significant coefficient on the slope of the yield curve in explaining the excess return of homebuilder ETFs during the COVID-19-induced lockdowns and before vaccinations underscores the significance of monetary policies aimed at mitigating pandemic effects. This finding highlights the substantial role of such policies in influencing ETF excess returns from July 2007 to March 2009.

Table 6. Net monthly alpha of homebuilder exchange-traded funds, S&P Homebuilders Index, DJ Real Estate Index, and Russell 3000 Index based on monthly returns from February 2006 to April 2023.

	Net Alpha
	Homebuilder ETFs	S&P Homebuilders Select Industry Index	DJ US Real Estate Index	Russell 3000 Index
February 2006 to April 2023	0.11	−0.63	−0.20	−0.14 ***
July 2007 to March 2009 (economic crisis period)	−1.27	7.20	−0.68	−0.10
COVID-19-Induced Lockdowns to First Vaccination (February 2020 to January 2021)	19.79 *	3.90	3.09	−0.15
February 2021 to April 2023 (post-COVID-19 vaccination roll-out period)	0.82	0.98	−0.58	−0.03

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

presents the net monthly alpha of homebuilder exchange-traded funds (ETFs) alongside the S&P Homebuilders Select Industry Index, DJ US Real Estate Index, and Russell 3000 Index based on monthly returns from February 2006 to April 2023.

Over the period from February 2006 to April 2023, homebuilder ETFs exhibited a net monthly alpha of 0.11, indicating a slight positive excess return compared to the benchmark indices. Conversely, the S&P Homebuilders Select Industry Index, DJ US Real Estate Index, and Russell 3000 Index displayed negative net alphas of −0.63, −0.20, and −0.14, respectively. Notably, the negative net alpha for the Russell 3000 Index is statistically significant at the 1% level, suggesting considerable underperformance compared to the other indices.

During the economic crisis period from July 2007 to March 2009, homebuilder ETFs experienced a substantial negative net alpha of −1.27, indicating significant underperformance relative to the benchmark indices. In contrast, the S&P Homebuilders Select Industry Index displayed a notable positive net alpha of 7.20, while the DJ US Real Estate Index and Russell 3000 Index also exhibited negative net alphas.

Throughout the period from the COVID-19-induced lockdowns to the first vaccination (February 2020 to January 2021), homebuilder ETFs demonstrated a remarkably high net alpha of 19.79, which is statistically significant at the 1% level, suggesting significant outperformance compared to the other indices. The S&P Homebuilders Select Industry Index and DJ US Real Estate Index also exhibited positive net alphas, albeit lower than that of homebuilder ETFs, while the Russell 3000 Index displayed a negative net alpha.

In the post-COVID-19 vaccination rollout period from February 2021 to April 2023, homebuilder ETFs maintained a positive net alpha of 0.82, indicating continued outperformance relative to the benchmark indices. The S&P Homebuilders Select Industry Index also exhibited a positive net alpha, while the DJ US Real Estate Index and Russell 3000 Index displayed negative net alphas.

The net monthly alpha analysis provides insights into the relative performance of homebuilder ETFs compared to benchmark indices across different market conditions. Homebuilder ETFs demonstrate periods of both underperformance and significant outperformance, indicating dynamic shifts in their performance relative to the broader market indices over time.

Table 7. Summary of average rolling monthly alphas of homebuilder exchange-traded funds (ETFs), S&P Homebuilders Select Industry Index, DJ US Real Estate Index, and Russell 3000 Index based on monthly returns from February 2006 to April 2023. Analysis is based on extended Carhart’s factor model.

	Homebuilder Exchange-Traded Funds (ETFs)	S&P Homebuilders Select Industry Index	DJ US Real Estate Index	Russell 3000 Index
Average rolling monthly alpha	−2.98	−2.89	0.23	−0.05
t-statistics
Average rolling alpha of homebuilder exchange-traded funds (ETFs) versus S&P Homebuilders Select Industry Index	−0.37
Average rolling alpha of homebuilder exchange-traded funds (ETFs) versus DJ US Real Estate Index	−3.27 ***
Average rolling alpha of homebuilder exchange-traded funds (ETFs) versus Russell 3000 Index	−4.56 ***

*** statistically significant at 1% significance level.

presents a summary of the average rolling monthly alphas of homebuilder ETFs in comparison to the S&P Homebuilders Select Industry Index, DJ US Real Estate Index, and Russell 3000 Index.

As indicated in Table 7 above, the average rolling net monthly alpha for homebuilder ETFs stands at −2.98, contrasting with the average net monthly alphas of −2.89, 0.23, and -0.05 for the S&P Homebuilders Select Industry Index, DJ US Real Estate Index, and Russell 3000 Index, respectively. These findings suggest that homebuilder exchange-traded funds (ETFs) have exhibited underperformance when compared to both the DJ US Real Estate Index and the Russell 3000 Index over the specified time period according to the Carhart’s factor model analysis.

Table 8. Summary of results for net monthly alphas based on conditional Carhart multi-factor model used in this study. Risk-adjusted performance is measured based on monthly returns from February 2006 to April 2023.

	Multi-Factor Model	Adjusted R-Square
February 2006 to April 2023	−0.05	0.70
July 2007 to March 2009 (economic crisis period)	−0.27	0.73
COVID-19-Induced Lockdowns to First Vaccination (January 2020 to 2002021)	28.44 ***	0.99
February 2021 to April 2023 (post-COVID-19 vaccination roll-out period)	1.61	0.64

*** Statistically significant at 1% level.

provides valuable insights into the performance of homebuilder ETFs based on conditional multi-factor asset pricing model when we incorporate publicly available information.

Table 8 summarizes the results for net monthly alphas based on the conditional Carhart multi-factor model utilized in this study, evaluating risk-adjusted performance using monthly returns data from February 2006 to April 2023.

From February 2006 to April 2023, the multi-factor model indicates a net monthly alpha of −0.05. The adjusted R-squared value, representing the proportion of return variation explained by the model, stands at 0.70 during this period.

During the economic crisis from July 2007 to March 2009, the net monthly alpha decreases further to −0.27, while the adjusted R-squared value slightly increases to 0.73, indicating a better model fit with the data amidst this tumultuous period.

Throughout the period from COVID-19-induced lockdowns to the first vaccination from January 2020 to January 2021, the net monthly alpha notably rises to 28.44, which is statistically significant at the 1% level, indicating substantial model outperformance. The adjusted R-Squared also increases to 0.99, indicating a highly explanatory model during this volatile period.

In the post-COVID-19 vaccination rollout period from February 2021 to April 2023, the net monthly alpha increases to 1.61, indicating positive performance. However, the adjusted R-squared value decreases to 0.64, suggesting a slightly weaker model fit compared to the previous period.

Table 9. A summary of results from Treynor and Mazuy (1966) model. For the Treynor and Mazuy (1966) models, α_s measures selectivity, whereas β₂ measures market timing. T-stats are in parentheses. Results are based on the monthly returns February 2006 to April 2023. We also evaluated market timing and selectivity by dividing the sample into pre-economic crisis and post-economic crisis periods.

Treynor and Mazuy (1966) Model
	αs	β2
February 2006 to April 2023	−0.48 (−1.06)	0.004 (0.36)
July 2007 to March 2009	−0.40 (−0.20)	−0.02 (−0.50)
February 2020 to January 2021	3.49 (1.01)	−0.07 (−1.72)
February 2021 to April 2023	−0.51 (−0.33)	0.04 (0.93)

summarizes the results of Treynor and Mazuy’s market timing and selectivity model. The analysis also divides the sample into pre-economic crisis and post-economic crisis periods, shedding light on the dynamics of selectivity and market timing over different economic conditions.

Table 9 presents result from the Treynor and Mazuy (1966) model, evaluating selectivity and market timing based on α_s and β₂ coefficients, respectively. T-stats are provided in parentheses, and the analysis spans monthly returns from February 2006 to April 2023, covering periods before and after an economic crisis.

From February 2006 to April 2023, the αs coefficient is −0.48 with a T-stat of −1.06, indicating negative selectivity. The β₂ coefficient is 0.004 with a T-stat of 0.36, suggesting negligible market timing ability.

During the period from July 2007 to March 2009, the α_s coefficient decreases to −0.40 with a T-stat of −0.20, implying slightly less negative selectivity. However, the β₂ coefficient drops to −0.02 with a T-stat of −0.50, indicating minimal but negative market timing ability.

From February 2020 to January 2021, there is a notable shift with an αs coefficient of 3.49 and a T-stat of 1.01, suggesting positive selectivity. The β₂ coefficient is −0.07 with a T-stat of −1.72, indicating negative market timing.

In the subsequent period from February 2021 to April 2023, the αs coefficient returns to negative territory at −0.51 with a T-stat of −0.33, signaling negative selectivity. The β₂ coefficient increases to 0.04 with a T-stat of 0.93, suggesting a slight positive market timing ability.

Table 10. A summary of results from the conditional Treynor and Mazuy (1966) model. For the Treynor and Mazuy (1966) models, α_s measures selectivity, whereas β₂ measures market timing. The results are based on the monthly returns from February 2006 to April 2023. The T-stats are in parentheses.

Treynor and Mazuy (1966) Model
	αs	β2
February 2006 to April 2023	−0.43 (−1.03)	−0.004 (−0.32)
July 2007 to March 2009	−0.98 (−0.45)	−0.04 (−0.77)
February 2020 to January 2021	2.04 (0.54)	−0.04 (−0.31)
February 2021 to April 2023	−1.26 (−0.73)	0.04 (0.96)

summarizes the result of conditional market timing and security selection skills of portfolio managers.

Table 10 presents findings from the conditional Treynor and Mazuy (1966) model, analyzing αs and β₂ coefficients to evaluate selectivity and market timing, respectively. T-stats in parentheses indicate significance levels. The analysis spans from February 2006 to April 2023, covering various economic periods.

Between February 2006 and April 2023, the αs coefficient is −0.43 with a T-stat of −1.03, suggesting slightly negative selectivity. The β₂ coefficient stands at −0.004 with a T-statistic of −0.32, indicating minimal market timing ability.

During the economic downturn from July 2007 to March 2009, the αs coefficient decreases to −0.98 with a T-statistic of −0.45, signaling more negative selectivity. The β₂ coefficient declines to −0.04 with a T-statistic of −0.77, implying slight negative market timing ability during this turbulent period.

In the period from February 2020 to January 2021, the αs coefficient increases to 2.04 with a T-stat of 0.54, suggesting positive selectivity. However, the β₂ coefficient remains at −0.04 with a T-statistic of −0.31, indicating negligible market timing ability.

From February 2021 to April 2023, the αs coefficient decreases substantially to −1.26 with a T-statistic of −0.73, indicating more negative selectivity during this latter period. The β₂ coefficient rises to 0.04 with a T-stat of 0.96, suggesting slightly positive market timing ability.

The examination of selectivity and market timing across different economic periods provides insights into the dynamics of investment strategies and their performance under varying market conditions.

Table 11. Value-at-Risk at 95% confidence interval and Expected Shortfall for homebuilder ETFs and stock market indices, namely S&P Select Homebuilding Index, DJ US Real Estate Index, and Russell 3000 Index. Data include monthly returns from February 2006 to April 2023.

	Homebuilder Exchange-Traded Funds	The S&P Select Homebuilding Index	DJ U.S. Real Estate Index	Russell 3000 Index
Value at Risk at 95% confidence interval	−13.97	−12.95	−8.01	−8.30
Expected Shortfall	−18.61	−17.68	−15.21	−10.34

Table 4 displays the 95% Value at Risk and Expected Shortfall for homebuilder ETFs, the S&P Select Homebuilder Index, DJ US Real Estate Index, and Russell 3000 index. The data cover monthly returns from February 2006 to April 2023.

summarizes the Value at Risk and Conditional Value at Risk for homebuilder ETFs, the S&P Select Homebuilder Index, DJ US Real Estate Index, and Russell 3000 Index. The data cover monthly returns from February 2006 to April 2023.

Beginning with VaR, it represents the maximum potential loss at a 95% confidence level. homebuilder ETFs exhibit a VaR of −13.97, indicating a 5% probability of experiencing losses exceeding 13.97%. This signifies a considerable downside risk associated with these ETFs. The S&P Select Homebuilding Index closely follows with a VaR of −12.95, still reflecting significant potential losses. On the other hand, the DJ U.S. Real Estate Index and Russell 3000 Index display lower VaR values of −8.01 and −8.30, respectively, indicating comparatively lower downside risks in comparison to the homebuilding assets.

The Expected Shortfall (ES), also known as Conditional VaR, provides insights into the average severity of potential losses beyond the VaR level. Homebuilder ETFs demonstrate an ES of −18.61, implying substantial average losses beyond the VaR threshold. Similarly, the S&P Select Homebuilding Index exhibits an ES of −17.68, indicating significant average losses beyond the VaR level. Conversely, the DJ U.S. Real Estate Index and Russell 3000 Index showcase lower ES values of −15.21 and −10.34, respectively, suggesting relatively smaller average losses beyond the VaR threshold.

Homebuilder ETFs exhibit a higher downside risk compared to other investment options analyzed in the data. The Value at Risk (VaR) and Expected Shortfall (ES) values for homebuilder ETFs are notably higher than those for other indices, such as the S&P Select Homebuilding Index, DJ U.S. Real Estate Index, and Russell 3000 Index. This indicates that investors in homebuilder ETFs face a greater likelihood of experiencing significant losses, both in terms of the maximum potential loss (VaR) and the average severity of losses beyond that threshold (ES).

7. Summary and Conclusions

This research extends previous investigations of real estate ETFs and mutual funds with detailed scrutiny of homebuilder ETFs performance behavior through various market scenarios. The results confirm that homebuilder ETFs experience increased volatility and risk during economic downturns, which matches Goodwin et al.’s (2021) findings through negative alpha data from both the 2008 financial crisis and the COVID-19 pandemic. The findings of this research agree with those of Malhotra (2024) by demonstrating significant beta values which clarify the effect of general stock market trends on homebuilder ETFs.

The main contribution of this study stems from using an extended Carhart multi-factor model to uncover detailed factor exposures, which show how momentum effects vary between crisis periods and recovery times. Previous studies by Kaushik and Pennathur showed uncertain results for real estate mutual funds, whereas this research demonstrates how homebuilder ETFs achieved dynamic outperformance during specific periods like the post-COVID-19 recovery phase. Contrary to Bond and Mitchell (2010), who observed minimal excess returns in real estate funds, this study discovers a small positive alpha for homebuilder ETFs when examined over an extended timeframe. Rolling alphas and conditional factor models increase the strength and reliability of these findings.

The study findings highlight the distinct risk–return characteristics of homebuilder ETFs, which provide practical guidance for managing investment portfolios. Market fluctuations require increased attention because homebuilder ETFs heavily rely on stock market trends. Risk mitigation for portfolio managers and investors can be achieved through hedging and the addition of non-correlated asset classes during market declines while taking advantage of superior returns during recovery periods. The application of the Value at Risk (VaR) and Conditional Value at Risk (CVaR) reveal the sector’s elevated risk of loss, which underscores the importance of strong risk management strategies.

This study makes an important addition to sector-specific ETF research by combining advanced modeling methods with a comparison with the literature. This work delivers actionable advice for investors and financial analysts who need to understand homebuilder ETFs dynamics concerning risk assessment and portfolio diversification along with market timing strategies.