Financial constraints, investment opportunities and the market value of REIT liquidity.

(1)

Financial Constraints, Investment Opportunities and

the Market Value of REIT Liquidity

University of Amsterdam

Faculty Business and Economics

MSc Business Economics

Program: Finance & Real Estate Finance

April 2014

Sander Klemann 5732689

(2)

- 2 - Abstract

In this thesis the impact of financial constraints and investment opportunities on the market value of REIT cash and credit lines is tested. A Fama and French (1998) valuation model is used to find that on average the market values REITs cash holdings at a premium. The average market value of one dollar held in cash by REITs is ranging from $1.108 to $1.244 depending on the estimation method used. Unused Credit lines are less valuable but still increase firm value. REITs with large growth opportunities have their liquidity valued higher compared to REIT with no such growth options. Furthermore liquidity is substantially more valuable in financially constraint REITs whether due to firm characteristics or external market circumstances. Overall the results provide evidence that support the financial slack arguments of Myers and Majluf (1984) and Jensen (1986) showing that shareholders weigh the benefits of liquidity against the agency costs associated with it.

(3)

- 3 -

Section 1: INTRODUCTION

This thesis focuses on the market valuation of Real Estate Investment Trusts (REITs) liquidity and the relation to financial constraints and investment opportunities. In a perfect market cash would be valued at face value meaning that cash is a zero net present value investment. Putting an extra dollar in a firm would increase the firm’s value with exactly one dollar. It does not matter whether the dollar is held by the firm or its shareholders because everyone know exactly what it is worth, how it can be used and what it opportunity costs are if held in cash. However the real world market is far from perfect and due to information asymmetries, taxes and agency costs the value of liquid assets varies over time and between firms. Past research into liquidity has focused on the firm perspective to find the determinants of cash and credit lines. It brought forward several factors such as capital market access and cash flow volatility that affect the firm’s decisions in setting the levels of available liquidity. This thesis covers the market perspective and tests how the market values REIT liquidity. I test if the determinants of REITs liquidity are aligned with the drivers of its market value. The aim is to give REIT managers a shareholder point of view, to give insight into how the market is evaluating their liquidity choices and under which circumstances it is profitable to raise or lower levels of cash or credit lines. I test if the market valuation varies with different levels of financial constraints, available investment opportunities and other firm and market characteristics. I use a model created by Fama and French (1998) to measure the market value of a REIT’s cash holdings and unused credit lines. The dataset is taken from the SNL database on US REITs and covers 147 unique equity REITS. Although the market value of cash and other forms of liquidity has been established before only one paper by Hill, Kelly, and Hardin III (2012) did so focusing on the REIT sector. Their results are verified and further explored with a focus on how valuations have changed based on firm and market characteristics. The results show that on average the market values one dollar held in cash at a premium ranging from $1.108 to $1.244 depending on the estimation method used. Unused Credit lines are less valuable but still increase firm value. REITs with many investment opportunities have their liquidity valued higher compared to REIT with no such options. Furthermore liquidity is substantially more valuable in financially constraint REITs whether due to firm characteristics or external market effects.

The following section contains a review of the relevant literature on cash and credit line valuation. Section 3 covers the hypotheses followed by the model and methodology in Section 4. Section 5 outlines the data and the descriptive statistics. Section 6 contains the results, followed by

(5)

- 5 -

robustness checks in Section 7. In the final section I conclude and give recommendations for future research.

Section 2: LITERATURE REVIEW

The primary focus of this study is on how the market values cash and credit lines in REITs with varying financial constraints and investment opportunities. To this end I first cover the basic characteristics of the REITs industry and explain why it is a unique market to test financial theories. Second the determinants of cash and credit lines need to be covered. Next the drivers of market value from a theoretical and empirical perspective are explained plus the differences between cash and credit lines.

2.1 Background on the REITs Industry

In 1960 the United States congress created REITs to make real estate investment available to the general public. Up until then real estate as an investment vehicle was only available to large corporations or wealthy individuals with enough capital to make the large investments needed. By introducing REITs small investors were able to gain exposure to the real estate market in the same way they could buy stocks or bonds. There are two types of REITs that differ in their type of investments: mortgage REITs, which focus solely on providing mortgages or invest in existing mortgages and mortgage-backed securities and equity REITs which invest and own properties and earn most of their revenues from rents. Less than 10% of all REITs are mortgage REITs.

REITs pay no taxes over the generated income however to be classified as a REIT certain conditions have to be met: 75% of their assets need to be in real estate and 75% of their income need to be from those real estate assets. Most importantly, they are required to distribute at least 90% of their taxable income to their shareholders (NAREIT, 2014). This effectively reduces REITs to a vehicle that collects capital from many different investors, invests the capital into large real estate projects and passes the returns back to the investors. Due to this obligatory distribution requirement, REITs cannot (fully) finance investments with internally generated funds. They are all exogenously capital constrained which makes the REIT market a perfect laboratory to test varying effects of financial constraints, agency problems and liquidity management.

Furthermore REITs operate under strict corporate governance and reporting rules which increase transparency and lower information asymmetries. This is further amplified by REITs’

(6)

- 6 -

frequent interactions with the capital market which subjects them to increased monitoring. Moreover, due to its strict regulations the REIT industry is largely homogeneous which leads to more consistent outcomes. Lastly REITs are obliged to publicly provide information on their investments, liquidity positions, credit lines and debt structures. All these factors make the REIT industry a valuable environment for testing the financial theories behind liquidity valuation.

2.2 Cash

In a perfect capital market the value of a dollar held by a firm would be valued at face value. So whether the firms hold this dollar or pays it to the investor as dividend the value stays the same. In case financing is needed, it can always be obtained from the market at the right price (Modigliani & Miller, 1958). However in the real world the market is imperfect: there are taxes, agency problems and information asymmetries between managers and stockholders, plus there is a conflict of interest between stocks and bondholders. Hence a dollar held might be valued above or below face value. There are two motives why firms may hold cash: the Transaction motive and the precautionary motive. The transaction motive is based on Miller and Orr (1966) argument stating that cash is needed for daily transactions and without it the firm cannot function. This motive is of little influence because in the REIT sector the need for working capital is very low since long term leases have predictable costs and cash flows. The precautionary motive means that firms hold cash not to forgo on positive NPV projects when external financing is not available or is too costly. The pecking order theory (Myers & Majluf, 1984) states that internal financing is always preferred to external financing due to information asymmetries between the better informed managers and the shareholders. Thus in markets with either high external financing costs or high information asymmetries maintaining internal liquidity can be valuable. It enables firms to undertake profitable investments without external support and to do so in a timely fashion. Without these internal funds value adding growth options can remain on the table and underinvestment occurs. So whether firms should maintain liquidity and if this is valued by the market depends (in part) on the availability of these growth options.

Jensen (1986) argues against holding cash and states that having excessive (free) cash leads managers to waste these funds. Having enough internal financing options managers do not need shareholder approval and invest in negative NPV projects and/or use it for personal gain. By paying out cash and financing investments with debt the managers are controlled by the capital market

(7)

- 7 -

thereby increasing transparency and reducing managerial opportunism. These different arguments suggest that firms should hold cash, so investment opportunities can be profited from, but not too much so that managers will waste it. Hence the cash amount held should reflect the company specific growth options.

In their 1999 paper Opler, Pinkowitz, Stulz, and Williamson examine all publicly traded US firms from 1971 to 1994 to determine the firm and market characteristics that determine firm cash holdings. They find evidence of a static tradeoff theory where managers maximize shareholder wealth. Firms with high growth opportunities or volatile future cash flows hold more cash than other firms. Furthermore those with good access to external capital, such as large firms or firms with a credit rating, tend to hold less cash. This is because the likelihood of underinvesting decreases as the access to capital markets improves. Their results confirm earlier findings into the determinants of cash holdings by Kim, Mauer, and Sherman (1998) who create and then test a model of optimal corporate investment in liquid assets. In addition they find that cash holdings are positively related to the level of asymmetric information. These results are consistent with the view that firms hold cash for precautionary purposes and show an empirical relation between cash holdings and both firm’s and market’s characteristics. The important question to ask is whether shareholders benefit from an extra dollar in the firm’s cash balance or whether it worth more to them paid out as dividend.

In a paper by Pinkowitz and Williamson (2002) the value shareholders place on cash holdings of regular firms is discussed. They find that a dollar held in cash is on average valued at face value by the market. Their research covers four hypotheses; the first focuses on available growth opportunities which makes cash more valuable; the second one on the predictability of investment opportunities, which makes cash less valuable, the third one on probability of financial distress which increases cash value and the fourth on capital market access which decreases cash value. By using an econometric model developed by Fama and French (1998) they are able to confirm the first three hypotheses on investment opportunities, their predictability, and financial distress but find no conclusive evidence that capital market access effects cash valuations. Hence, in regular firms the market’s valuation of cash and the firm’s choices are largely in line when investment opportunities are concerned.

When it comes to REITs, cash is just a small proportion of total liquidity at only two percent of assets (Ott, Riddiough, & Yi, 2005). However Hardin III, Highfield, Hill, and Kelly (2009) argue that this is a deliberate choice by the REIT managers because most REITs choose to pay out more than the

(8)

- 8 -

required 90% to its investors whilst they could just as well hold it in cash. In their paper the determinants of REIT cash holdings are studied to find a reason why REITs hold so little cash. They find similar results as the study focused on non-REITs by Opler et al. (1999). The level of cash holdings are inversely related with funds from operations (cash flow) and long and short term debt levels (leverage) while they are positively related to the cost of external financing and growth options. Also an explanation for the low levels of cash is given: managers deliberately hold little cash to reduce free cash flow agency problems which increases transparency and decreases the future cost of external capital. They also find that cash holdings are negatively related to available credit lines and their use. This result would suggest that there is substitutive relation between cash and credit lines which will be covered more extensively in the next section. In their concluding remarks Hardin III et al. (2009) suggest that further research into the market value of REIT liquidity is necessary.

Gamba and Triantis (2008) use a theoretical model to measure the value of financial flexibility which is a firm’s ability to ‘mobilize its financial resources in order to take preventive and exploitive actions in response to uncertain future contingencies in a timely manner to maximize the firm value’

(Byoun, 2011, p. 4). Financial flexibility is measured in their model by the marginal value of cash. Their results show that whether an added dollar will be valued at a premium depends on internal firm characteristics and external market circumstances. The value of cash increases when firms have positive future growth opportunities, volatile cash flows or if external financing is expensive. The value decreases if the cost of holding cash increases or if the reversibility of capital increases. In their model the costs of external financing proxies for the level of financial constraints which has an inverse relationship with the marginal value of cash. This is later empirically confirmed by Denis and Sibilkov (2010) who show that the market value of cash in non-REIT firms is significantly higher for financially constrained firms. Gamba and Triantis (2008) note that in their model financial constraints have only a small effect on the value of cash for mature firms compared to a larger effect in startups. This is mainly due to the fact that mature firms are able to finance their investments with internally generated funds. On top of that, the results are highly dependent on the tax disadvantage associated with cash holdings. Denis and Sibilkov (2010) have shown that the market value of cash in non-REIT firms is significantly higher for financially constrained firms. Both taxes and the ability to finance investment internally are of little influence in the REIT sector. It is therefore interesting to see if the market’s valuation of REIT liquidity is in line with the determents of cash holdings like in non-REITS or if the different market structure leads to different valuations.

(9)

- 9 - 2.3 Credit Lines

Together with cash holdings credit lines are the essential part of REITs liquidity. In 2009 credit lines made up almost 75% of the total liquidity available to the REIT market (compared to roughly 45% in regular firms). Almost two third of the lines of credit was unutilized representing 8,9% of the total assets in the sector (Sufi, 2009). Credit lines are pre-arranged lines of credit that allow companies to borrow up to a certain amount under pre-determined conditions. A basic upfront fee is paid to the bank each year for providing the credit line option and either a floating or fixed rate on the withdrawn amount. REITs make especially heavy use of these credit lines due to the lack of alternative ways of funding large investments in a short time frame normally associated with real estate. Investment opportunities arise and disappear very quickly and require substantial amounts of money. Thus having to wait for days or even months on agreements with debt1 issuers or setting up equity issues is usually not a viable option. The credit lines are used as a way to temporarily finance investments until more permanents modes of financing, either equity or debt, are found. Brown and Riddiough (2003) call this ‘bridge financing’. The funding works in cycles, first REITs arrange credits lines with a partner bank thereby creating sufficient liquidity to finance upcoming opportunities. When an opportunity arises the acquisition is made by drawing down on the credit lines. The firm then continues to find a permanent source of financing by either issuing seasoned equity offerings or long-term debt. With the newly found funds the credit lines are paid off and the cycle begins again. Cash and credit lines differ from each other in that credit lines are off-balance sheet and have a fixed and variable cost component. Furthermore credit lines decrease agency costs because they imply increased bank monitoring (Yun, 2009). However Hill et al. (2012) find that despite these differences credit lines and cash are largely substitutes. In their paper the market value of REITs cash holdings in relation to available credit lines is researched. The value of an additional dollar of cash diminishes when more credit lines are available. Their main conclusion on the value of cash is that the market value of an additional dollar of REITs cash holdings is worth on average $1.34. Furthermore they find that liquidity increased in value during the recent global financial crisis. This

1

Whether REIT should use debt at all is an interesting topic, attracting capital structure researchers to the REIT industry. Due to the none-firm specific and tangible nature of the REIT’s assets, debt is easily acquired. But since there are no taxes there are no tax shields benefits. Thus following the Trade of Theory (Kraus & Litzenberger, 1973) REITs should not use debt at all. However most of the REITs have been doing so extensively until an industry wide deleveraging after the global financial crisis of 2008. For more info on this topic see Feng, Ghosh, and Sirmans (2007) or Giambona, Harding, and Sirmans (2008)

(10)

- 10 -

thesis is partly based on their paper and makes use of the same model. I expand and test their results and further explore the relation between the market value of liquidity and financial constraints. Also the impact of growth opportunities is examined and improvements on the used estimation method are made to yield more accurate estimations.

Section 3: HYPOTHESES DEVELOPMENT

As mentioned above the market value of liquidity is expected to be different cross-sectional. The value of cash and unused credit lines will dependent on external factors such as the liquidity of the capital market but more importantly on individual firm characteristics like investment opportunities, capital market access and the level of financial constraints. The remainder of this section explains the three hypotheses and their empirical predictions.

3.1 Market Value of Liquidity in REITs

In the REIT sector there is a low need for operating working capital because the cash flows and the costs are quite stable. For regular firms who have predictable costs and cash flows, it is easy to finance investments with retained earnings and hence there no need to maintain financial flexibility. However due to the mandatory payout ratio for REITs the possibility to fund new investment from retained earnings is limited. Therefore REIT liquidity is valued for its ability to fund new investments and maintain dividends. On top of that, the transparent nature of the REIT industry and the frequent interactions with the capital market, which increases bank monitoring, causes a decrease in agency costs associated with free cash flow. These factors should increase the value of liquidity therefore I predict that the market value of a dollar of cash will be valued at a premium.

Hypothesis 1

REIT’s cash holdings will be valued at a premium by the market. 3.2 Investment Opportunities

As mentioned above holding cash and having credit lines available is costly to a REIT. Myers and Majluf (1984) argue that investors would value financial slack more if the firms have good investment opportunities. In other words the cost of keeping cash balances or paying for credit lines can only be justified if the company also got opportunities to use them profitably. Jensen (1986) free cash flow

(11)

- 11 -

argument states that if firms have few good investment opportunities managers will waste available funds. Thus an important determinant of the market value of liquidity should be the investment opportunities available to the REIT. Therefore it is expected that the value of REIT liquidity will increase if there are more investment opportunities available.

Hypothesis 2

The market value of cash and credit lines is higher for REITs with more investment opportunities.

Adam and Goyal (2008) state that the best proxy which can be used for investment opportunities is Tobin’s Q which is defined as the market value to book assets ratio. However since our dependent variable is essentially Tobin’s Q, directly measuring the interaction between Tobin’s Q and cash or credit lines is not possible. To overcome this problem two different proxies are used. Adam and Goyal (2008) determine that the price earnings ratio is the second best option for determining investment opportunities. I will define the price earnings ratio as the price at the end of the quarter divided by funds from operations (FFO) which is defined asnet income plus depreciation and amortization.2 As an alternative proxy for investment opportunities past year FFO growth is used as suggested by Pinkowitz and Williamson (2002).

3.3 Financial Constraints

When a firm is limited in acquiring external financing it is financially constrained. For firms with good access to capital markets (low financial constraints) and hence a lower cost of external capital, liquidity should be valued lower because financing of new investments do not depend on them. Financial constraints are thought to decrease the level of investment and create strong incentives to hoard cash (Fazzari, Hubbard, & Petersen, 1988; Kaplan & Zingales, 1995). Therefore a negative relationship between financial constraints and the market value of liquidity is expected. The hypothesis is as follows:

Hypothesis 3

The market value of cash and credit lines is higher in REITs with higher (external) financial constraints.

2_{If net income would be used the earnings would be understated. The large depreciation expenses}

have a different economic meaning in the REIT sector since real estate is less prone to obsolescence. FFO is a frequently used measurement of operating cash flow in the REIT sector.

(12)

- 12 -

To determine the level of constraints in REITs I use a method derived by Kaplan and Zingales (1997) who did extensive research into financially constraint firms. They created the KZ-index to measure a firm’s degree of financial constraint. The KZ index is derived from a combination of firm’s characteristics such as leverage and cash holdings and will be fully defined in appendix. A higher KZ value indicates higher levels of financial constraints. Thus the KZ-index should be positively related to the value of cash holdings and credit lines.

An alternative measure of financial constraints is Credit Lines available scaled by Total Assets. It has been shown by Sufi (2009) to be a good estimate of the level of constraint. Since there is a substitute relationship between cash and credit lines (Hardin III et al., 2009) I expect that the value of cash will decrease even more if there is a higher proportion of the credit lines available. To measure the effect of external constraints I use the 2008 crisis as natural experiment. In times of tight credit periods, such as in the recent global financial crisis, external financial constraints are high thus I predict that internal liquidity will be valued higher during this period. Another proxy for financial constraints is whether a REIT has a credit rating. Opler et al. (1999) find that having a credit rating, or even better an investment grade rating, makes firms hold less cash due to lower financial constraints. This would suggest that the market would value cash holdings of firms with a credit rating at a lower rate.

Section 4: MODEL AND METHODOLOGY

4.1 Model

To estimate the market value of dollar of cash hold by a REIT I use a valuation model of Fama & French (FF) (1998). Their original goal was to research the influence of debt and dividends on firm value. Their model was later adapted and used by several different studies to estimate the contribution of other factors, such as cash, on firm value (Dittmar & Mahrt-Smith, 2007; Drobetz, Grüninger, & Hirschvogl, 2010; Hill et al., 2012; Pinkowitz, Stulz, & Williamson, 2006). The model I will use is an adaptation of the FF model of Pinkowitz et al. (2006) which focuses on estimating the value of cash and cash equivalents. The main difference with the original FF valuation model is that total assets are split into an cash and a non-cash component. To make the model suitable for the REITs industry the R&D variable is dropped from the original model. Furthermore earnings are replaced by funds from operations (FFO). The base model is then defined as follows:

(13)

- 13 - (Equation 1) Where:

 Xt is the level of variable X in year t divided by the level of assets in year t

 dXt is the change in the level of X from quarter t − 2 to quarter t

 dXt+2 is the change in the level of X from quarter t to quarter t+2

 Mrktval is the market value of the firm calculated at the end of quarter as the sum of the Market value of equity, Total Preferred Equity and Total Liabilities

 Cash is Cash and cash equivalents as defined by the appropriate accounting standard

 CLun is Unused Credit Lines defined as Credit lines Available minus Credit lines Drawn

 Div is dividends defined as common dividends paid

 FFO is Funds from Operations defined as net income excluding gains or losses from sales of properties or debt restructuring and adding back real estate depreciation.

 NA is Net Assets defined as Total Assets minus Cash

 IntExp is interest expense defined as interest on debt and other borrowings

To control for heteroskedasticity all variable are scaled by Net Assets at time t. The original model uses a two period future change in the control variables to account for investors’ expectations. The past and future changes in Net Assets, dNAt and dNAt+2, proxy for the investment component of

expected cash flows, whereas the Div and IntExp variables account for current and future changes in dividend and leverage policy. FFO proxies for expected growth of profits. Following Kothari and Shanken (1992) a two period change in Market Value, dMrktVali,t+2, is added to account for

unexpected components in the other future variables (Fama & French, 1998, pp. 823-824).

The most important estimate is β1 which is the coefficient on Cash Holdings. It measures the

market to book value of an additional dollar held by the REIT. So if a REIT’s cash balance is increased with one dollar the market value rises with β1. In such β2 has the same interpretation but for Unused

Credit Lines.The set of control variables includes dummy variables for the property sector the REIT operates in. The number of property categories from the SNL database has been reduced to seven by combining a few similar SNL investment sectors into one dummy. REITs within a sector type Regional

(14)

- 14 -

Mall, Retail:other, or Shopping center get the dummy ‘Retai’ and REITs with property focus types of Diversified, Specialty and Other are put together into one category named ‘Other’. The other property

sectors stay the same. The seven property sector dummies left are: Hotel, Industrial, Residential,

Office, Retail, Storage and Other. Since Other is the largest category this is used as the base category. 4.2 Methodology

In their article about the FF valuation model Fama and French (1998, p. 822) explain in detail the econometric issues associated with their valuation method. They indicate that these issues could be mitigated by estimating the model with the Fama and MacBeth (1973) method. Therefore, just as Pinkowitz et al. (2006), this estimation technique will be used for all regressions. The procedure is as follows: In the first step a cross-sectional regression is performed for each year. Then, in the second step, the final coefficient estimates are obtained as the average of the first step coefficient estimates. Thus the reported coefficients are the mean of the cross-sectional regression coefficients. To account for autocorrelation Newey-West (1987) standard errors are used which are derived in the same manner. Using the Fama Macbeth’s method any survivorship bias and/or serial correlation should be reduced. For robustness I also estimate the regressions with a fixed effect estimator following Drobetz et al. (2010).

I examine my hypotheses by dividing the sample in subgroups based on the firm characteristic in question. To do so dummy variables are created that segment all observations into the different subgroups. This means the cross-sectional regression takes the following form:

∑

(Equation 2)

Xk,i represent all variables from equation 1 excluding Cash and Unused Credit Lines. The Cash and

Unused Credit Lines variables (Ci) are interacted with three (J=3) dummy variables represented by δj.

They are determined by ranking the observations per characteristics, per year and segment the REITs into three groups: high (j=3), average (j=2) and low (j=1). High (low) REITs are those in the top (bottom) 20% of all REITs in a given year, while average observations are those in the middle 60%. For example, when the effect of growth options on the value of cash is examined, the price/earnings ratio is one of the characteristics used. So, δ1 equals one for the REITs with the lowest 20%

(15)

- 15 -

price/earnings ratio in a given year and δ3 equals one for REITs for the highest 20%. Hence, the

estimated coefficients γ1 and γ3 represent the market value of cash for firms who have the

respectively lowest and highest 20% of the price/earnings ratios in a given year. To capture the market value of Cash or Unused Credit Lines and not the value of the characteristics, separate intercepts (αj) are being used for each group. An advantage of this specification is that firms are

allowed to switch groups from year to year and therefore the short term nature of liquidity management is captured. This method is adapted from Pinkowitz et al. (2006) who use it in the same manner to value cash holdings.

Section 5: DATA AND DESCRIPTIVE STATISTICS

5.1 Data

The baseline data starts from 1999Q3 with the introduction of the REIT Modernization Act and ends in 2013Q3. The initial sample data is retrieved from the SNL REIT database where only equity REIT’s from the United States are selected. The sample is based on quarterly data to account for the short term nature of liquidity management. Furthermore five sequential observations are needed to account for the two period lagged and future variables. Hence the retrieved data starts from 1999Q1. Observations that did not fit the following bounds were eliminated: 0.3 < Tobin’s Q ≤ 4.0; 0.0 ≤ Cash/Net Assets ≤ 0.5; 0.0 ≤ Leverage ≤ 1.0; interest expense <0; Credit Lines Unused<0. The first bound, based on Tobin’s Q, is used to eliminate outliners that would have distorted the regression results. The cash stock is limited to 50% because REITs have to comply with regulations limiting the amount of assets which can be held in non–real estate. REIT with cash holding higher than 50% are presumably in liquidation mode and will lose their REIT status shortly. Hence they will not reflect normal circumstances and are therefore dropped. The other bounds are used to eliminate outliers that are impossible and are probably a result of faulty data input. The same rules for elimination are used in Riddiough and Wu (2009). A total of 87 observations (1.5%) are dropped from the dataset due to these rules. Also observations with missing data are dropped which substantially restrict the sample size. The final sample consists of 5667 quarterly observations divided over 147 unique US based equity REIT’s.

(16)

- 16 - 5.2 Descriptive Statistics

The descriptive statistics can be found in Table 2. The size of the REITs within the sample varies greatly with Net Assets ranging from $9 million to $31 billion with a standard deviation of $4.3 billion. The market value is on average 133% of Net Assets. The average REIT holds 2.45% of its Net Assets in cash while Unused Credit Lines are equal to 9.56% of Net Assets. Furthermore it earns 1.4 cent for every dollar in Net Assets. The market value has decreased over the sample period with on average 10.5% shown by the mean dMrktval t+2. The other control variables have been stable over the sample

period since the means of the future difference variables are close to zero. The Pearson correlations of the important variables can be found in Table 3. The control variables: Div and FFO are positively correlated with market value, which supports prior theory. Cash and Clun are significantly related with market value which supports the used model and the hypotheses. The significant but small negative relation between Cash and Clun suggests cash and credit lines are substitutes till a certain degree.

Table 4 contains the descriptive statistics of the high and low financially constrained subgroups as determined by the KZ index. The difference in REIT size between the subgroups is noticeable. Highly constrained firms have on average one billion more in Net Assets than the low constrained firms. It also shows that low constrained REITs hold more cash than high constrained REITs (4.0% compared to 2.7% of Net Assets) and have more Unused Credit Lines (12.7% compared to 6.8% of Net Assets). Of all the low constrained firms 49% have a credit rating comparable to 43% in the highly constrained group. If having a credit rating is an indicator of the level of constraints as suggested by Faulkender, Milbourn, and Thakor (2006), Almeida, Campello, and Weisbach (2011) and Faulkender and Petersen (2006), the KZ index does not differentiate that well. However when looked at whether the rating is of an investment grade level there is a large difference. From the high constrained REITs with a credit rating only 39% has an investment grade rating compared to 73% of the low constrained REITs. Overall it seems the KZ index does differentiate well between constrained and unconstrained REITs.

(17)

- 17 -

Section 6: Results

6.1 Market Value of Cash and Credit Lines

Table 5 shows the regression results for the basic model estimating the market value of REIT liquidity estimated with Fama Macbeth (1973) and Fixed Effects estimators. The average market value of one dollar held in cash by a REIT is $1.244 with a 0.319 standard error for the Fama Macbeth estimation. When a Fixed Effect method is used this value is slightly lower at $1.108 with a standard error of 0.227. The average market value of a dollar in Unused Credit Lines is $0.353 with a standard error of $0.123 and $0.781 with a standard error of 0.148 for the Fama Macbeth and Fixed Effect estimator respectively. The results show that there is a direct and significant relation between liquidity and market value. The market value of an additional dollar of cash held is slightly above face value. A dollar in Credit Lines is valued at a discount which is to be expected since there are costs associated with it. The results on the cash variables are comparable to the results of Hill et al. (2012) although they do find a higher premium for Cash holdings. This difference could be caused by the use of a different estimation technique and the different sample period.3 The control variables are largely consistent over the different models and similar to previous studies in non-REIT studies valuating cash using the Fama French model (Dittmar & Mahrt-Smith, 2007; Hill et al., 2012; Pinkowitz et al., 2006). Although I do not make a direct comparison, cash seems to be more valuable for REIT compared to non-REITs. Previous research in non-REITs shows that the market values cash holdings at face value or slightly below (Pinkowitz & Williamson, 2002). This higher valuation of cash is likely due to the unique nature of the REIT industry. The high transparency and strict corporate governance regulations do not allow manager to misuse excess liquidity. These factors apparently outweigh the cash value decreasing characteristics of stable cash flows and low working capital needs. When it comes to credit lines even though credit lines are costly, the market on average assigns positive value to them. The positive and significant Cluncoefficient shows the market values the increased flexibility gained by Unused Credit Lines. This positive valuation is in line with Campbell, Devos, and Spieler (2008) who find positive shareholder wealth effects after the announcement of new credit line arrangements.

3

Their sample period is from 1999 till beginning 2009 and includes the height of the global financial crisis, a period in which Hill et al. (2012) show cash is more valuable. Also the use of OLS estimators instead of Fama Macbeth could explain this difference. Table 5 contains my base model estimated with OLS which indeed shows a higher cash valuation than the other estimators.

(18)

- 18 -

6.2 Investment Opportunities and Market Value of REITs liquidity

Table 6 contains the results of the second hypothesis. For ease of presentation only the Cash and Unused Credit Lineinteraction coefficients of the high and low groups, represented by γ1 and γ3 from

equation 2, are shown. The coefficients are compared and I test whether they are significantly different from each other. The REITs are divided into the high and low growth opportunities groups based on three different proxies. The first proxy used for growth opportunities is FFO growth which is defined as the percentual growth of FFO in one year. The coefficients for both Cash and CLun are not significant. In retrospect the use of sales growth as indicator might not applicable in the REIT sector because REIT earnings are relatively stable over time and only rises suddenly when acquisitions are made. The second proxy is yearly average net real estate investment growth. It also shows no significant results. However when the price earnings ratio is used there is a significant difference between the cash coefficients of the high and low subgroups. REITs with a higher price/earnings ratio have significantly higher coefficients on cash holdings. On average REITs that fall in the group with the highest 20% of price earnings ratios in a given year have their cash holdings valued at $1.35 on the dollar. The middle 60% group has their cash holdings valued at roughly face value while for the low subgroup its cash is valued at a significant discount. This increased value of cash from REITs with growth opportunities reflects Pinkowitz et al. (2006) results in non-REITS. Furthermore it is in line with the theoretical predictions of Jensen (1986) and Myers and Majluf (1984) who argue that the market valuation of a firm’s liquidity is mostly effected by its growth opportunities. The fact that cash is valued at a discount for firms with low investment opportunities accounts for Jensen’s prediction that managers will waste excess cash if no growth opportunities are available. Looking at the Unused Credit Line coefficients, only the subgroup with the lowest price/earnings ratios shows a small negative significant result. This was to be expected since credit lines are costly. Hence paying for credit lines will decrease the firm’s value if they cannot be used for positive NPV projects. In other words (costly) financial flexibility is only valuable if that flexibility can be put to use. In conclusion the second hypothesis stating that liquidity is more valuable for REITs with more investment opportunities can only be confirmed for cash holdings. The results show holding cash can be valuable if REITs have enough growth opportunities and although only sparsely supported credit lines seem to decrease firm value if there are no growth opportunities for which they can be used.

(19)

- 19 -

6.3 Financial Constraints and Market Value of REITs Liquidity

The results from the first two hypotheses have shown that the value shareholders place on cash is in general just above face value but if there are no growth opportunities leaving cash in the firms is destroying value for the shareholder; the dollar is worth more paid out to the investor. The next factor tested is the ease at which a firm can obtain new liquidity. If external capital is cheap, internal liquidity should be valued less. If however there are market circumstances where the REIT is restricted from raising capital, the value of liquidity should increase. To test this third hypothesis several variables are used to proxy for financial constraints or external capital costs. The results are shown in the second part of Table 6.

The first proxy used is the KZ-index. It shows that REITs who are highly financially constrained (high KZ index) have a substantially and significantly higher coefficient of cash holdings than firms who are less financially constrained. A dollar in cash is valued at three times its face value if held by highly constrained REITs, while for a low constrained REIT a dollar is only valued at 5 cents. A dollar in available credit lines for high constrained REITs is valued at 91 cents on the dollar compared to only 9 cent for the low constrained REITs. This results corresponds to the predictions from the financial theory on the value of liquidity discussed earlier (Denis & Sibilkov, 2010; Faulkender & Wang, 2006). It is also in line with the cash level determinants as examined by Hardin III et al. (2009). Hence the market’s valuation of liquidity is in line with the drivers of the mangers’ decisions. Riddiough and Wu (2009) found that financially constrained REITs reduce dividend payouts when investments are made. This would explain why the value of liquidity of constrained REITs is so much higher because it allows REITs to maintain dividend payouts while simultaneously investing in new projects. Furthermore this result shows that even though the REIT industry is highly dependent on credit lines, holding cash can be value increasing if the REIT is sufficiently financially constrained or is expected to be so in the future. REITs that are financially constrained and have difficulty in acquiring new credit lines will also have higher valuations placed on the credit lines still available to them. The next result, which uses Credit Lines Available to differentiate the subgroups, shows that cash will decrease in market value when the REIT has more credit lines available. This confirms the substitutive relationship between cash and credit lines found earlier by Hill et al. (2012). Whether cash is valuable thus also depend on the amount of credit lines available to the REITs. This substitutive relationship also explains the low level of cash holdings in REITs compared to non-REITs. Thus even though REITs can reduce dividend payouts (Hardin III et al., 2009) to hoard cash because REITs can rely on credit lines the need to do so

(20)

- 20 -

is reduced. The results show that REITs falling in the highest 20% subgroup of available credit lines have their cash holdings valued negatively. This confirms Jensen (1986) predication that excess cash holdings will destroy firm value if it is not needed although it is hard to differentiate this effect from the substitution effect.

The credit rating proxy shows mixed results. Although the Fama Macbeth regression shows no significant difference between the coefficients, the Fixed Effect estimator does shows a higher valuation for both cash and credit lines for REITs with a credit rating. This goes against the prediction that having a credit rating would make external financing easier to obtain and thereby decrease the value of internal liquidity. However the opposite is found, the market values cash higher if REITs have a credit rating. A possible explanation is that having a credit rating indicates increased transparency or it is an indicator of improved corporate governance leading to less managerial opportunism and/or agency costs thereby increasing the value of cash. To further test this explanation I made an additional regression using a subsample of only REITs with a credit rating. I test whether the market values cash differently if the REIT’s credit rating is of an investment grade. The results indeed show that cash valuations of REITs with an investment grade rating are substantially lower than those with just a credit rating. Hence having a credit rating increases the market value of cash. However if the rating is of investment grade the cash value decreases again. A possible explanation is that having a credit rating indicates less misuse of the cash thereby increases the value. However once the REIT achieves an investment grade rating cheap outside liquidity make cash holdings obsolete thereby decreasing it value again. Yet an alternative explanation could be that credit ratings are an indicator of REIT investment quality. Meaning that the market sees REITs with a credit rating as more capable of profiting from its investment opportunities effectively and hence the cash holding needed for those investments are more valuable. In conclusion more research is needed to determine the effect of the different factors at play when credit ratings are concerned.

The last regression tests whether REIT liquidity is valued differently during the recent global financial crisis of 2008 when there was a substantial decrease in outside liquidity. Hence all REITs were externally financially constrained. The subgroups in this regression are not based on firm characteristics but on the date of the observation. Thedummy variable equals one if the observation falls between the quarters 2007 Q3 and 2009 Q2. Only the credit lines coefficients show significant results. Just as in Hill et al. (2012) I find an increased value of Unused Credit Lines during the crisis in which external liquidity was more expensive for the REIT sector as a whole. This shows that the

(21)

- 21 -

market value of credit lines is also dependent on industry wide external financial constraints. The value of Unused Credit Lines rises with 45 cents during the crisis. A dollar in available credit lines is even valued at a premium at $1.31 per dollar during the crisis. Thus in periods of tight credit markets the perceived value of internal liquidity rises.

Section 7: Robustness Checks

To check the third hypothesis for robustness the regressions on financial constraints are run again on a subsample of the data. I have taken the observations with highest and lowest 20% financial constraints (based on the KZ–index) and ran a separate regression on just these observations. The same is done with the dataset split in half. The results are shown in Table 7. The coefficients for the 50/50 split data show similar results as in the full data set regression and therefore do not change the conclusions. The regressions done on the high and low quintiles show no significant results. This is most likely due to the severe data size reduction.

For robustness and to compare my result with Hill et al. (2012) I estimated all regressions with OLS. Over all regressions the absolute coefficients are found to be marginally higher compared to the other estimators for both cash and credit lines however this does not lead to different conclusions. Furthermore I have shown that the used KZ-index differentiate well between constraint REITs. However Lamont, Polk, and Saa-Requejo (2001) suggest that using lagged values when calculating the KZ-index yield more accurate results. The regressions run with the lagged KZ-index show only marginally different coefficients and again do not lead to different conclusions.

The results from section 6 show that the market valuation of liquidity is dependent on both the REIT’s level of financial constraints and its investment opportunities. However these factors could influence each other. A REIT might be financially constrained but if they have no investment opportunities holding cash would still decrease value. Hence hypotheses 2 and 3 should be tested simultaneously. Using the method of Pinkowitz et al. (2006) I create nine dummy variables based on two dimensions called dual interaction dummies. I first use P/FFO to split the sample into high, middle and low investment opportunities. The second step is to further segment these groups based on financial constraints determined by the KZ-index. In this way the data is divided into 9 different subgroups. In Table 7 the coefficient for high growth and high constraints group are shown and compared with the coefficients from the high growth and low constraints group. In this way the

(22)

- 22 -

growth options remain constant while the financial constraints vary. Only the credit line coefficient of the fixed estimator shows a significant difference in valuation. It shows that if growth options are held constant credit lines are only valuable if the firm is highly financially constrained. While having high growth options but no constraints actually decreases value. A possible explanation is that since credit lines are costly and if they can be profited from with external capital (low constraints), having (excess) credit lines would actually decrease firm value because the costs are unnecessary. However the results are not conclusive and further research focusing on the interaction effect of multiple firm characteristics is needed to make definite conclusions.

Some limitations with the study have to be mentioned. The dataset only contains REITs that where active at the time of data collection (2013Q3). This means that REITs who did not survive the 2008 crisis and might have had different liquidity valuations, were not included in the dataset which creates a survivorship bias. Furthermore the high/low groups determined by the KZ-index show a large difference in absolute size in terms of total assets: highly constrained REITs are larger than low constrained REITs. Since size can be a proxy for information asymmetries as shown by Diamond and Verrecchia (1991) the increased value of cash holdings in highly constrained REITs can also be the result of decreased information asymmetries which has been shown to increase value (Drobetz et al., 2010).

Section 8: Conclusion

In this thesis I estimate the market value of liquidity for equity REITs by using a Fama French valuation model. The results show that on average the market values cash holdings by REITs at a premium. The average market value of one dollar held in cash by a REIT is ranging from $1.108 to $1.244 depending on the estimation method used. Unused Credit lines are less valuable but still increase firm value. Valuations range from $0.353 to $0.781 for every dollar in Unused Credit Lines. This premium on cash valuations is consistent with a transparent market structure, reduced agency problems and the relative importance of liquidity in the REIT industry. Credit lines and cash are, at least partially substitutes. Cash decreases in value if more credit lines are available which supports prior theory. It also explains the voluntary low cash holdings since REITs can depend on credit lines to finance their investments.

(23)

- 23 -

The value the market places on liquidity is dependent on a REIT’s investment opportunities and access to funding. REITs with many growth opportunities have their liquidity valued higher compared to REIT with no such growth options. Furthermore liquidity is substantially more valuable in financially constraint REITs. This is due to either firm characteristics or outside market effects. Furthermore the access to funding has a larger impact on the value of liquidity than available investment opportunities, which is contrary to earlier findings for regular firms. However this result is only applicable to credit lines and further future research is needed. REIT managers should take these results into account when setting levels for cash holdings and credit lines. Overall the factors affecting the market value of liquidity correspond with the determinants of liquidity. Contrary to expectations, having a credit rating seems to increase liquidity valuations although results are mixed. Overall the results provide evidence that support the financial slack arguments by Myers and Majluf (1984) and Jensen (1986) showing that shareholders weigh the benefits of liquidity against the agency costs associated with it.

Bibliography

Adam, T., & Goyal, V. K. (2008). The investment opportunity set and its proxy variables. Journal of Financial

Research, 31(1), 41-63.

Almeida, H., Campello, M., & Weisbach, M. S. (2004). The cash flow sensitivity of cash. The Journal of Finance,

59(4), 1777-1804.

Almeida, H., Campello, M., & Weisbach, M. S. (2011). Corporate financial and investment policies when future financing is not frictionless. Journal of Corporate Finance, 17(3), 675-693.

Brown, D. T., & Riddiough, T. J. (2003). Financing choice and liability structure of real estate investment trusts.

Real Estate Economics, 31(3), 313-346.

Byoun, S. (2011). Financial flexibility and capital structure decision. Available at SSRN 1108850.

Campbell, R. D., Devos, E., & Spieler, A. C. (2008). The information content of equity REIT bank credit facility announcements. Journal of Real Estate Portfolio Management, 14(1), 1-6.

Denis, D. J., & Sibilkov, V. (2010). Financial constraints, investment, and the value of cash holdings. Review of

Financial Studies, 23(1), 247-269.

Diamond, D. W., & Verrecchia, R. E. (1991). Disclosure, liquidity, and the cost of capital. The Journal of Finance,

46(4), 1325-1359.

Dittmar, A., & Mahrt-Smith, J. (2007). Corporate governance and the value of cash holdings. Journal of

(24)

- 24 -

Drobetz, W., Grüninger, M. C., & Hirschvogl, S. (2010). Information asymmetry and the value of cash. Journal

of Banking & Finance, 34(9), 2168-2184.

Fama, E. F., & French, K. R. (1998). Taxes, financing decisions, and firm value. The Journal of Finance, 53(3), 819-843.

Faulkender, M., Milbourn, T., & Thakor, A. (2006). Capital structure and dividend policy: two sides of the same coin. Washington University in St. Louis, working paper.

Faulkender, M., & Petersen, M. A. (2006). Does the source of capital affect capital structure? Review of

Financial Studies, 19(1), 45-79.

Faulkender, M., & Wang, R. (2006). Corporate financial policy and the value of cash. The Journal of Finance,

61(4), 1957-1990.

Fazzari, S., Hubbard, R. G., & Petersen, B. C. (1988). Financing constraints and corporate investment: National Bureau of Economic Research Cambridge, Mass., USA.

Feng, Z., Ghosh, C., & Sirmans, C. (2007). On the capital structure of real estate investment trusts (REITs). The

Journal of Real Estate Finance and Economics, 34(1), 81-105.

Gamba, A., & Triantis, A. (2008). The value of financial flexibility. The Journal of Finance, 63(5), 2263-2296. Giambona, E., Harding, J. P., & Sirmans, C. (2008). Explaining the variation in REIT capital structure: the role of asset liquidation value. Real Estate Economics, 36(1), 111-137.

Hardin III, W. G., Highfield, M. J., Hill, M. D., & Kelly, G. W. (2009). The determinants of REIT cash holdings. The

Journal of Real Estate Finance and Economics, 39(1), 39-57.

Hill, M. D., Kelly, G. W., & Hardin III, W. G. (2012). Market value of REIT liquidity. The Journal of Real Estate

Finance and Economics, 45(2), 383-401.

Hovakimian, A., Hovakimian, G., & Tehranian, H. (2004). Determinants of target capital structure: The case of dual debt and equity issues. Journal of financial economics, 71(3), 517-540.

Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. The American economic

review, 76(2), 323-329.

Kaplan, S. N., & Zingales, L. (1995). Do financing constraints explain why investment is correlated with cash flow? : National Bureau of Economic Research.

Kaplan, S. N., & Zingales, L. (1997). Do investment-cash flow sensitivities provide useful measures of financing constraints? The Quarterly Journal of Economics, 112(1), 169-215.

Kim, C.-S., Mauer, D. C., & Sherman, A. E. (1998). The determinants of corporate liquidity: Theory and evidence. Journal of Financial and Quantitative Analysis, 33(03), 335-359.

Kothari, S. P., & Shanken, J. (1992). Stock return variation and expected dividends: A time-series and cross-sectional analysis. Journal of financial economics, 31(2), 177-210.

(25)

- 25 -

Kraus, A., & Litzenberger, R. H. (1973). A STATE‐PREFERENCE MODEL OF OPTIMAL FINANCIAL LEVERAGE. The

Journal of Finance, 28(4), 911-922.

Lamont, O., Polk, C., & Saa-Requejo, J. (2001). Financial constraints and stock returns. Review of Financial

Studies, 14(2), 529-554.

Miller, M. H., & Orr, D. (1966). A Model of the Demand for Money by Firms. The Quarterly Journal of

Economics, 80(3), 413-435.

Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporation finance and the theory of investment.

The American economic review, 261-297.

Myers, S. C., & Majluf, N. S. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of financial economics, 13(2), 187-221.

NAREIT. (2014). What is a REIT? Retrieved 21-02-2014, 2014, from http://www.reit.com/nareit/policy-issues/federal-tax-legislation/reit-modernization-act-rma-1999

Opler, T., Pinkowitz, L., Stulz, R., & Williamson, R. (1999). The determinants and implications of corporate cash holdings. Journal of financial economics, 52(1), 3-46.

Ott, S. H., Riddiough, T. J., & Yi, H. C. (2005). Finance, investment and investment performance: Evidence from the REIT sector. Real Estate Economics, 33(1), 203-235.

Pinkowitz, L., Stulz, R., & Williamson, R. (2006). Does the contribution of corporate cash holdings and dividends to firm value depend on governance? A cross‐country analysis. The Journal of Finance, 61(6), 2725-2751. Pinkowitz, L., & Williamson, R. (2002). What is a dollar worth? The market value of cash holdings. The Market

Value of Cash Holdings (October 2002).

Riddiough, T. J., & Wu, Z. (2009). Financial constraints, liquidity management and investment. Real Estate

Economics, 37(3), 447-481.

Sufi, A. (2009). Bank lines of credit in corporate finance: An empirical analysis. Review of Financial Studies,

22(3), 1057-1088.

Yun, H. (2009). The choice of corporate liquidity and corporate governance. Review of Financial Studies, 22(4), 1447-1475.

(26)

- 26 -

Appendix

A.1 Tables

Table 1 Summary of Hypotheses and Predictions

Hypothesis Prediction for Cash Prediction for Unused

Credit Lines

Empirical proxies (expected sign) H1: Market

value of Cash and Credit Lines

Market values an additional dollar in cash

at a premium

Market values credit lines

positively none

H2: Investment

Opportunities + +

Price FFO Ratio (+) FFO Growth (+)

H3: Financial

Constraints + +

KZ-Index (+) Available Credit Lines (-) Crisis (+)

(27)

- 27 -

Table 2 Descriptive Statistics

Mean 1st Quartile Median 3rd Quartile Standard

Deviation Mrktval t 132.9% 108.9% 125.1% 147.9% 37.5% FFO t 1.40% 1.04% 1.43% 1.77% 1.06% dFFO t 0.07% -0.09% 0.06% 0.24% 1.10% dFFO t+2 -0.08% -0.26% -0.07% 0.09% 1.13% dNA t 4.44% -0.83% 2.08% 7.20% 10.6% dNA t+2 -6.45% -7.75% -2.13% 0.82% 18.6% Intexp t 0.76% 0.59% 0.74% 0.92% 0.31% dIntexp t 0.02% -0.03% 0.01% 0.07% 0.14% dIntexp t+2 -0.04% -0.07% -0.01% 0.03% 0.19% Div t 1.18% 0.78% 1.08% 1.37% 1.35% dDiv t 0.05% 0.00% 0.03% 0.11% 1.65% dDiv t+2 -0.06% -0.12% -0.03% 0.00% 1.66% dMrktval t+2 -10.5% -17.7% -7.2% 1.37% 25.0% Cash t 2.45% 0.40% 0.96% 2.54% 4.54% Clun t 9.56% 4.58% 8.52% 13.07% 6.94%

Table 2 Xt is the level of variable X in year t divided by the level of net assets in year t. dX t is the change in the

level of X from year t-2 to year t divided by net assets in quarter t ((Xt-2 - X t)/NA t). dX t+2 is the change in the level of X from quarter t+2 to quarter t divided by net assets in quarter t ((X t+2 -X t)/NA t). MrktVal is market value of equity + preferred equity + total liabilities. FFO is funds from operations. NA is net assets, which is defined as total assets - cash and cash equivalents. IntExp is interest expense. Div is total dividends paid. Cash is cash and cash equivalents. Clun is credit lines available - credit lines drawn.

Table 3 Pearson Correlation Coefficients

Mrktval NA FFO Cash Div Int Exp

NA 0.081c 1 FFO 0.393c -0.032b 1 Cash 0.226c -0.030b 0.054c 1 Div 0.251c -0.094c 0.240c 0.112c 1 Intexp -0.097c -0.030b -0.198c -0.019 -0.121c 1 Clun 0.215c -0.099c 0.164c -0.058c 0.124c -0.245c

Table 3 shows the Pearson correlation coefficients for the data used in the analysis. MrktVal is market value of

equity + preferred equity + total liabilities. FFO is funds from operations. NA is net assets, which is defined as total assets - cash and cash equivalents. IntExp is interest expense. Div is total dividends paid. Cash is cash and cash equivalents. Clun is credit lines available - credit lines drawn. a, b, c denote statistical significance at the 10% ,5%, 1% levels, respectively.

(28)

- 28 -

Table 4 Descriptive Statistics based on Financial constraints (KZ index)

High Financial Constraints Low Financial Constraints

Mean Standard Deviation Mean Standard Deviation

NA 3,944,426 6,208,915 1,939,706 2,367,419 Mrktval 141% 45% 140% 46% FFO 1.08% 1.40% 1.96% 1.14% Cash 2.74% 5.53% 3.99% 6.85% Div 0.91% 2.05% 1.77% 1.72% Clun 6.84% 6.09% 12.73% 9.41% Leverage 71 % 8.79% 29.07% 14.36% Rating 42.55% - 48.54% - Invgrade 16.76% - 35.67% -

Table 4 Descriptive statistics based on two subsamples. The REITs with the 20% highest and lowest levels of

(29)

- 29 -

Table 5 Regression results of Fama Fench valuation model of US REITs.

Fama Macbeth Fixed Effects OLS

Cash t 1.244c 1.108c 1.399c (3.91) (4.93) (10.78) Clun t 0.353b 0.781c 0.554c (2.86) (5.33) (7.36) Div t 16.109c 4.499c 8.807c (6.44) (3.74) (7.19) FFO t 19.904c 10.078c 19.663c (7.92) (5.55) (15.44) Intexp t 29.203c 14.747b 23.508c (11.94) (2.17) (12.47) dDiv t -4.405a -1.158b -3.399c (1.87) (2.54) (3.96) dDiv t+2 -7.587c -2.104c -2.977c (3.22) (2.77) (3.48) dFFO t -8.676c -2.494c -6.552c (3.66) (3.82) (6.21) dFFO t+2 -4.921b -4.628c -7.417c (2.2) (4.93) (6.83) dIntexp t -10.298b -6.267 -7.610a (2.19) (0.89) (1.89) dIntexp t+2 -5.378 20.765b -2.06 (0.85) (2.4) (0.56) dNA t 0.340c 0.311c 0.185c (5.9) (3.33) (3.92) dNA t+2 -0.187a -0.524c -0.089 (2.04) (4.64) (1.63) dMrktval t+2 0.163a 0.256c 0.122c (1.88) (2.92) (3.24)

Firm fixed effects No Yes No

Property sector dummies Yes No Yes

N 5085 5085 5085

R-squared 0.343 0.256 0.545

Table 5 Dependent variable is Mrktval which is market value of equity + preferred equity + total liabilities over

Net assets. This table shows the estimation results of the regression in equation (1). Xt is the level of variable X

in year t divided by the level of net assets in year t. dX t is the change in the level of X from year t-2 to year t divided by net assets in quarter t ((Xt-2 - X t)/NA t). dX t+2 is the change in the level of X from quarter t+2 to quarter t divided by net assets in quarter t ((X t+2 -X t)/NA t). MrktVal is market value of equity + preferred equity + total liabilities. FFO is funds from operations. NA is net assets, which is defined as total assets - cash and cash equivalents. Intexp is interest expense. Div is total dividends paid. Cash is cash and cash equivalents. Clun is credit lines available - credit lines drawn. The year dummies used in the OLS regression are left out to save space. With the Fama/MacBeth methodology each year cross-sectional regressions are run and the reported coefficients are the means of the time series of regression coefficients. T-statistics (absolute values) are in parentheses. The R2 of the fixed effects regression refers to the within-dimension. The Fama/Macbeth and OLS regressions contain property sector dummies but these are omitted due to space constraints. Constants were included in the regressions but are not reported as well. The reported R2 for the Fama Macbeth regression is the average of the R-square’s from the yearly cross-sectional regressions. Statistical significance is indicated by a (10%), b (5%), and c (1%) letter.

(30)

- 30 -

Table 6 Cash and Credit lines coefficients of various regressions

Fama Macbeth Fixed Effect

Cash Credit Lines Cash Credit Lines

FFO Growth

Lowest FFO Growth 1.305 0.302 1.787 0.737

(0.53) (0.45) (0.27) (0.72)

Highest FFO Growth 0.792 0.387 1.071 0.690

(0.45) (0.44) (0.68) (1.07) Price/FFO Lowest P/FFO 0.013 -0.051 0.792 0.706 (2.21)b (2.84)b (1.71)a (0.07) Highest P/FFO 1.351** 0.227 0.912 0.690 (1.37) (0.72) (1.52) (0.02)

Avg. Real Estate Investment Growth

Lowest RE Inv. Growth 1.697 0.604 1.601 0.660

(0.68) (1.61)b (1.00) (0.63)

Highest RE Inv. Growth 0.906 0.030 0.087 0.853

(0.49) (0.89) (3.52)a (0.03)

Financial constraints KZ Index

Lowest Financial Constraints 0.05 0.162 0.768 0.630

(1.43) (1.82)a (2.84)c (0.46)

Highest Financial Constraints 3.245*** 1.025** 1.478 1.775***

(4.73)c (1.78) (3.78)c (4.02)c

Credit lines Available

Lowest Credit Lines Available 1.403 - 1.133 -

(1.28) (0.72)

Highest Credit Lines Available -0.687** - 0.491 -

(2.04)a (1.18) Credit Rating No Credit rating 0.901 0.488 0.829 0.520 (2.10)b (2.69)c (4.76)c (3.56)c Credit Rating 1.437 0.251 1.509* 1.141*** (1.26) (1.33) (1.85)a (2.71)c Crisis Crisis 1.604 1.306 1.357 0.791 (0.41) (2.22)b (1.38) (0.08) No Crisis 1.740 0.852** 1.071 0.780 (12.1)c (9.21)b (4.42)c (5.17)c

Table 6. For the regressions run using the methodology of Fama and MacBeth each year cross-sectional

regressions are run. The reported coefficients are the means of the time series of regression coefficients. The absolute T-values are in parentheses. The coefficients shown are the interaction between cash to net assets or unused credit lines to net assets and a dummy variable indicating the firm characteristic. The coefficients represent the value the market places on one dollar of cash or unused credit line. The dummy variables take the value of 1 if the firm has that characteristic in that year. Lowest and Highest refer to the quintiles of the variable which are re-ranked each year. a, b, c denote statistical significance and *, **, and *** indicate that the coefficient is larger than the other category at the 10, 5, and 1 percent levels respectively.

Financial constraints, investment opportunities and the market value of REIT liquidity.