Transaction Prices of German Closed-End Real Estate Funds in the Secondary Market and Their Influencing Factors

doi:10.21203/rs.3.rs-2113505/v1

Download PDF

Research Article

Transaction Prices of German Closed-End Real Estate Funds in the Secondary Market and Their Influencing Factors

https://doi.org/10.21203/rs.3.rs-2113505/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

This study is the first to assess and contribute to a better understanding of the factors affecting the prices of closed-end real estate funds (CEREFs) in Germany’s secondary market. The dataset is based on 16,718 observations of realized transaction prices from 412 different German CEREFs; the factors comprise macroeconomic, fund, property, and liquidity variables. We regress macroeconomic factors on benchmark variables, such as listed real estate assets and the stock market, to observe whether German CEREFs react differently from the overall economic state from a portfolio diversification perspective. The benchmark variables are homogeneously driven by stock market excess return, DAX volatility index (VDAX) growth, and, to a certain extent, the bond rate, while CEREFs are mainly affected by the unemployment rate and economic activity. Fund and property age have a significant impact on CEREFs. Additionally, they have optimal levels at which their respective impacts start to reverse. Initial fund leverage and rising liquidity in the secondary market also have a significant impact. However, the observed effects vary across the types of real estate held by CEREFs, and investors must consider this in their pricing policies. Finally, this study presents a comprehensive overview of the German secondary market environment for CEREFs and examines their development.

Finance

Real Estate

German Closed-End Funds

Secondary Market

Transaction Prices

Closed-end real estate funds (CEREFs) are mostly unlisted real estate investment vehicles that are part of the wider private equity investment sector (Nadler 2018). Unlike open-end real estate funds (OEREFs), real estate stocks (RESs), and real estate investment trusts (REITs), CEREFs are characterized by a limited time horizon with no option for the redemption of funds to the issuing company or for their sale in a liquid stock market (Mattarocci and Scimone 2020). Once the desired amount of nominal equity of a CEREF is collected and the issuance of additional shares is closed, the only possibility to dispose of the share lies in a secondary market transaction that is accessible to all and where the supply and demand of sellers and buyers meet and define the price. In rare cases, issuing companies voluntarily offer or accept a share redemption for a fixed price based on the internally calculated net asset value (NAV) as internally calculated by the fund manager.

Previous studies dealing with CEREF performance have focused on the US market and have relied on a relatively small number of self-reported NAVs from fund managers. However, this is the first study to assess and contribute to a better understanding of the German secondary market for CEREFs and determine the factors driving transaction prices. Transaction price is the most important factor in the secondary market for CEREF investors, as previous deals mostly serve as a benchmark for their own investment decisions. This study provides an overview of the corresponding market environment and its development. For factor analyses, we use a dataset based on 412 CEREFs and 16,718 observations of realized transaction prices in the secondary market. We apply a stepwise buildup approach by testing the effects of several macroeconomic, fund, property, and liquidity variables on transaction prices. Additionally, from a portfolio diversification perspective, we examine whether the listed real estate assets, namely DAX subsector real estate (DAXRE) and REITs, and the stock market, namely CDAX, react differently to the overall economy.

The remainder of this paper is organized as follows. Section 2 presents a literature review on real estate private equity performance and factors impacting it; Section 3 depicts the secondary market in Germany and its development; Section 4 describes the underlying data for our analyses; Section 5 provides an overview of the applied methodology; Section 6 presents our results; and Section 7 presents the main conclusions of the study.

Following Kaplan and Schoar’s (2005) comprehensive study on overall private equity performance, there is increasing scientific interest on the performance of real estate private equity funds (Nadler 2018). However, the literature on the factors affecting CEREF performance and thus share value is scarce and mainly focused on global performance (Arnold, Ling, and Naranjo 2021, 2019; Aarts and Baum 2016; Krautz and Fuerst 2015; Alcock et al. 2013; Tomperi 2010; Anson and Hudson-Wilson 2003) and European performance (Morri, Perini, and Anconetani 2021; Delfim and Hoesli 2016; Kiehelä and Falkenbach 2015; Fuerst, Lim, and Matysiak 2014, 2013; Fuerst and Matysiak 2013; Brounen, Veld, and Raitio 2007). National studies exist only on the US market (Arnold, Ling, and Naranjo 2021; Farrelly and Stevenson 2019; Barker, Seah, and Shilling 2019; Arnold, Ling, and Naranjo 2017; Fisher and Hartzell 2016; Case 2015; Hahn, Geltner, and Gerardo-Lietz 2005; Pagliari, Scherer, and Monopoli 2005) and to a lower extent on the UK market (Farrelly and Matysiak 2012; Bond and Mitchell 2010; Baum and Farrelly 2009). To the best of our knowledge, the only studies analyzing CEREFs in the German market and their performance are Nadler (2018) and Delfim and Hoesli (2016). Nadler (2018) compares the ex-ante and ex-post performance disclosures of German CEREFs and finds substantial problems due to the overestimation bias of many fund managers. Delfim and Hoesli (2016) compare five European countries using NAVs from 2001 to 2014, inferring that the impacts of macroeconomic factors are country dependent and, as posited by Farrelly and Stevenson (2019) and Fuerst, Lim, and Matysiak (2014, 2013), depend on whether the fund is open-ended or closed-ended. Furthermore, they compare the reactions of non-listed funds, listed RESs, and direct real estate funds to several fund characteristics and conclude that fund characteristics such as investment style, vehicle structure, size, and leverage affect all three fund types; the latter two factors exhibit optimal levels at which their respective impacts reverse, while fund age only affects CEREFs.

Studies assessing the German market for OEREFs, the other part of private equity real estate funds, are scarce; however, there are more studies related to OEREFs than CEREFs. Gerlach and Maurer (2020) assess the secondary market activities of OEREFs and find high trading levels when the fund management company has suspended the issue or redemption of shares. Moreover, they find that shares trade at a discount when the fund management company suspends their redemption, whereas they trade at a premium when the fund management company suspends their issue. Finally, they find that there has been an increase in secondary market trading activities since German regulations introduced a minimum holding period of 24 months and mandatory notice period of 12 months for share redemptions. Schnejdar et al. (2020) assess the impact factors of the NAV spread on the secondary market for OEREFs, based on the difference between transaction prices on the secondary market and the indicated NAV. They find that leverage, liquidity ratio, extraordinary payouts, the share of institutional investors, and the policy uncertainty index affect the NAV spread. Stein (2014, 2013) provides an overview of the effects of the global financial crisis on the risk and return characteristics of OEREFs with regard to NAV and transaction prices in the secondary market. Kurzrock, Gläsner, and Wilke (2009) assess whether significant performance differences occur between retail and institutional funds and find that the asset allocation of funds mainly drive the mean relative returns and that institutional investors are better-off than retail investors.

Most studies that focus on real estate private equity fund performance compute the returns on NAVs provided by industry associations and other third-party data providers, such as INREV, Burgiss, Cambridge Associates, Investment Property Databank, NCREIF, and Preqin. This is particularly problematic given that the available information from industry associations and other third-party data providers is voluntarily provided by the respective fund managers, as the issuing companies are not obliged by the federal or state regulations to publish performance data or other fund information (Arnold, Ling, and Naranjo 2019). Additionally, in almost all cases, NAVs are computed for non-liquidized funds, thus increasing the potential reporting bias; this has caused much controversy regarding the reliability of performance data. Phalippou and Gottschalg (2009) find that private equity fund performance in previous research is overstated and largely driven by the inflated accounting valuation of ongoing investments. Harris, Jenkinson, and Kaplan (2014) and Jenkinson, Sousa, and Stucke (2013), conversely, find that fund valuations are conservative, with exceptions during the period when follow-on funds are raised. Nadauld et al. (2019) assess a global sample of transaction prices in the secondary market for private equity investments and find that most transactions occur at a discount on the NAV. They strongly emphasize that substantial discretion is required when computing NAVs and point out that the extent to which NAVs fairly represent the present value of a fund’s future cash flows is unclear. However, our data are not based on hypothetical NAVs provided by fund managers but on realized transaction prices in the secondary market, where the supply and demand of the market meet and define the price of a CEREF share.

Another issue with the performance data of CEREFs with regard to industry associations and other third-party data providers is that the number of underlying funds is relatively small compared to the total number of available funds on the market. For example, INREV includes only 38 funds from 10 different issuing companies, and 19 funds stem from one issuing company for the German CEREFs market. However, our data are based on 412 CEREFs from 131 issuing companies, thus reflecting a large sample and market diversity.

Factors affecting the price or performance of real estate assets have been extensively discussed in the literature, but few studies have investigated private equity real estate funds, and even fewer have focused on CEREFs. Fuerst and Matysiak (2013), Fuerst, Lim, and Matysiak (2014, 2013) and Fairchild, MacKinnon, and Rodrigues (2011) find that for samples of private equity real estate funds, fund size and fund age respectively have positive and negative impacts on price, based on European or US samples. However, most of the literature is related to direct real estate or listed real estate companies. Kohlert (2010) shows that, in the UK direct real estate market, unemployment negatively affects asset performance. According to Schätz and Sebastian (2009), the German market displays the same relationship between direct real estate funds and OEREFs; additionally, they find that government bonds have a positive impact. Adams and Füss (2010) showed the positive impact of economic activity on a global data sample of direct real estate.

To the best of our knowledge, property-related factors, such as property size or age, have not yet been analyzed in the context of listed or private equity real estate markets. A major reason for this is the broad portfolio of various properties within one vehicle, making it difficult to summarize their status in concrete figures. While German CEREFs are, for the most part, single-property funds, for which property-related figures are relatively easy to specify, listed real estate assets or German OEREFs are multi-property vehicles. For example, German OEREFs must follow the principle of risk mixing through diversification such that no property has a share of more than 15% in the portfolio at the time of purchase. However, Esrig, Hudgins, and Cerreta (2011) show that large properties have historically outperformed smaller properties on an absolute and risk-adjusted basis while controlling for sectors: of office, multifamily, retail, and the top six metropolitan areas Boston, New York, Washington DC, Chicago, San Francisco, and Los Angeles. Additionally, they state that previous studies have generally failed to control for property type and, thus, the academic literature on the performance of large properties is inconclusive. However, this study controls for property types by generating subsamples. In general, for most properties we can assume that the older a property, the lower the transaction price, as the future costs of renovation work increase with time.

Transaction-related factors such as the number or volume of transactions in stock or secondary markets are often used as simple proxies for liquidity (Ametefe, Devaney, and Marcato 2016). However, no previous study has assessed CEREFs using this proxy, as most of them are based on NAVs and not on secondary market transactions. Because of the overall illiquid characteristic of the secondary market for CEREFs, we assume that higher liquidity in the market reflects higher demand, causing an increase in prices.

CEREFs are issued in the primary market and set up for long-term holding periods; investors cannot redeem their capital prior to the contracted date. This is one of the key differences between CEREFs and OEREFs. However, the establishment of functioning secondary markets enables investors to sell their shares prior to the contract date at a price that meets the current supply and demand of the market. While trade volumes in the secondary market are continuously increasing, nominal equity in the primary market dropped in 2013 (see Fig. 1). This is because of the Alternative Investment Fund Managers Directive (AIFMD), which was implemented in Germany in 2013 via the German Capital Investment Code (KAGB), and also because of ongoing effects of damage to investor trust during the global financial crisis, as multiple CEREFs failed to deliver returns or were liquidated (Nadler, 2018).

Trades on secondary markets occur on either open or internal trading platforms of issuing companies; thus, trading platforms are either independent of issuing companies or not. The largest trading platform for CEREFs in Germany’s secondary market is Fondsbörse Deutschland, which is an open trading platform operated and monitored by the stock exchanges of Düsseldorf, Hamburg, and Hanover. According to Fondsbörse Deutschland, this trading platform accounts for approximately 90% of all transactions in the secondary market, with over 3,500 transactions per year, and a yearly average nominal volume of circa €140 m in the last three years of the review period. Comparing the growth of the average transaction price in the CEREF market with the listed real estate markets, namely DAXRE and REITs, shows that the overall development is comparable but exhibits slightly higher volatility (see Fig. 2). The latter phenomenon results mainly from the fact that the average transaction price includes different CEREFs every month, while the composition of the listed real estate indices barely changes.

Contrary to other trading platforms, Fondsbörse Deutschland has a closed-order book, so investors’ bids do not influence each other, which encourages fair pricing. One disadvantage of closed-order books is insufficient transparency; however, due to monitoring by the stock exchanges of Düsseldorf, Hamburg, and Hanover and the non-interdependent bidding aspect, this issue is relatively small. The pricing on Fondsbörse Deutschland results from a so-called “Einheitskurs” process and not from an auction or fixed pricing. Therefore, for each CEREF, all bids and asks are collected, and then, for each CEREF, one price per day is assessed, leading to the highest transaction volume with the lowest excess supply and demand. This process steadies the price and reduces outlier bids and asks. Furthermore, it benefits both the seller and buyer, as the assessed price is averaged between the bid and ask prices, leading to higher and lower prices for the seller and buyer both.

Table 1

Einheitskurs vs. Auction Price
One bid price at 80% of €100k of nominal volume	Einheitskurs	Auction price
One ask price at 90% (for €100k)	85%	80%
Two ask prices at 90% and 95% (each for €100k)	92,5%	91%
Two ask prices at 90% and 95% (each for €50k)	90%	80%
This table presents the resulting prices and respective advantages of sellers and buyers when different pricing procedures (Einheitskurs vs. auction price) are adopted.

The seller side on the secondary market for CEREFs consists of private individuals selling their shares; the buyer side is more heterogeneous. Before the global financial crisis, institutional investors accounted for over half of all transactions, which declined to approximately a quarter in 2009 and was rebalanced at a level of about a third (see figure 3). However, private buyers in the secondary market for CEREFs can be considered institutional given their experience. Hence, a strict categorization of buyers as private or institutional buyers is not feasible.

Fondsbörse Deutschland Beteiligungsmakler AG provided us with all the CEREF transaction data on their trading platform between May 2, 2007, and October 23, 2020, comprising 33,440 transactions from 1,092 different funds, with information for each transaction on the name of the fund and issuing company, transaction date, nominal volume, transaction price as a percent of nominal volume, and currency.

For technical and IT policy reasons, Fondsbörse Deutschland does not provide fund or property information; we manually collected data from their trading platform. Fund information includes the issue date of the fund prospectus, the total fund volume, and initial leverage, the latter being calculated by dividing the debt level at the end of the placing phase by the collected nominal equity. Property information includes the number of properties, property type (office, retail, hotel, residential, healthcare, shopping center, etc.), property address, construction year of the property, and gross lettable area in m².[1] However, since fund and property information are not always complete or obtainable, we analyze CEREFs prospectuses, financial statements, and other reports. Thus, we have complete information on 18,153 transactions from 480 CEREFs. To control for potential distorting biases or interacting effects of different properties within a CEREF, we consider only CEREFs with one property, slightly reducing our sample to 16,718 observations on transactions from 412 different funds.

Additionally, we generated two CEREF subsamples, mitigating potential selection bias within the dataset and testing for further robustness of the results. The tested sample comprises German CEREFs with one property, located anywhere in the world; thus, we test for domestic effects by only including the properties located in Germany in our first subsample (sample 2). Hence, sample 2 included 10,626 transactions from 322 different funds. Controlling for property types, and since offices have been the preferred real estate type for CEREF structures over the years, the second subsample (sample 3) includes only properties in Germany that are solely used as offices. Hence, sample 3 included 4,418 transactions from 95 different funds.[2]

To control for the overall economic state under which the CEREFs are invested, we consider several macroeconomic variables sourced from Federal Reserve Economic Data and the German Central Bank. We include GDP growth, inflation growth, M1 money supply growth, the German 10-year government bond rate, stock market excess return (defined as the return difference between the CDAX and DAXRE, subtracting the German 10-year government bond rate as a risk-free rate), unemployment rate, VDAX-NEW index growth, and activity index growth.[3] To understand whether the German CEREF market reacts differently to macroeconomic effects compared to listed real estate assets and the stock market, we regress the macroeconomic effects on the DAXRE, RX REIT All Share, and CDAX indices. As the RX REIT All Share index is available only from November 2007, the starting point of the analysis is January 2008.[4]

[1] As a robustness check of our estimation results, we consider for the German sample in a binary format whether the properties are located in a Top 7 city (Berlin, Hamburg, Munich, Frankfurt, Stuttgart, Cologne, Düsseldorf) and within a Top 7 city whether they are at a 1A location according to Jones Lang LaSalle’s definition. Both factors have no significant impact on the price and do not change our estimation results.

[2] Further property types, like retail, hotel, residential, care facility, shopping center, and others, are also included for the observed transactions. However, we do not generate subsamples for them as the respective transactions are not observed for several months, and no consistent monthly time series data can be generated for them.

[3] The activity index is only available on a weekly basis. Hence, the closest value to a month-end is considered as the monthly value.

[4] The change in the period under review from May 2007 to January 2008 does not change our estimation results.

Compared to listed real estate assets, for example, CEREFs have fewer transactions in the secondary market and are not regularly traded on a monthly basis.[5] To generate the time series model and assess which factors affect the transaction prices of German CEREFs in the secondary market, we first grouped all transaction prices by month and weighted them by their nominal volume to calculate for every month a weighted average transaction price. Second, we apply the same procedure to all property and fund variables. Every monthly calculated variable in the period under review includes an average of 212 CEREF transactions, with minimum and maximum levels of 55 and 363 transactions in April 2009 and March 2018, respectively. By contrast, we use no performance or growth information for the variables, as their compositions change substantially on a monthly basis, thus making performance information inconclusive. We have 6 dependent and 15 independent variables, with 154 monthly observations each. Table 2 presents summary statistics for the observed variables.

Table 2 Descriptive statistics

Type	Variables	Mean	Std. Dev.	Min	Max
Dependent variables	Sample 1: All	0.6825	0.1881	0.3158	1.2743
	Sample 2: Germany	0.6876	0.2404	0.2646	1.6152
	Sample 3: German offices	0.6273	0.2077	0.2286	1.2961
	DAXRE	0.0049	0.0728	-0.2964	0.3884
	REIT return	0.0061	0.0732	-0.3992	0.2880
	MSCI World Index	0.0035	0.0478	-0.1904	0.1090
Macroeconomic variables	Inflation growth	0.0131	0.0079	-0.0050	0.0332
	Unemployment rate	0.0733	0.0117	0.0530	0.0970
	M1 growth	0.0063	0.0053	-0.0070	0.0314
	Bond rate	0.0138	0.0139	-0.0070	0.0463
	Stock excess return	-0.0089	0.0596	-0.3865	0.1628
	GDP growth	-0.0003	0.0064	-0.0502	0.0242
	VDAX growth	0.0253	0.2220	-0.3656	1.2803
	Activity growth	-0.0655	0.8904	-5.3075	3.6525
Fund variables	Fund size (in mEUR)	142.4	45.1	72.8	511.3
	Fund age (in months)	188.9	62.6	20.9	402.2
	Initial leverage	0.8382	0.1922	0.3632	1.5688
Property variables	Property size (in m²)	35,365	6,053	22,335	64,070
Property variables	Property age (in years)	21.1	8.7	8.6	79.8
Liquidity variables	Volume (in mEUR)	4.2	2.2	0.8	10.5
Liquidity variables	Number	107.1	50.5	21.0	230.0

Note: This table presents the summary statistics of all the variables used in the analyses. The data are sourced from Federal Reserve Economic Data, German Central Bank, Fondsbörse Deutschland Beteiligungsmakler AG, annual reports of the fund managers, and Thomson

The average transaction price in our sample was approximately 68% of the nominal volume, with a slightly lower average of 63% for office buildings. To test the effects of macroeconomic, fund, property, and liquidity variables on CEREF prices, we adopt a build-up approach by testing the effects of the aforementioned variables using a stepwise method:

$${p}_{it}= {Macro}_{it}{\beta }_{i}+{u}_{it}$$

$${p}_{it}= {Macro}_{it}{\beta }_{i}+{Fund}_{it}{\gamma }_{i}+{u}_{it}$$

$${p}_{it}= {Macro}_{it}{\beta }_{i}+{Fund}_{it}{\gamma }_{i}+{Property}_{it}{\delta }_{i}+{u}_{it}$$

$${p}_{it}= {Macro}_{it}{\beta }_{i}+{Fund}_{it}{\gamma }_{i}+{Property}_{it}{\delta }_{i}+{Liquidity}_{it}{\epsilon }_{i}+{u}_{it}$$

where ${p}_{it}$ is the ${n}_{i} x 1$ vector of monthly prices for sample $i$ at time $t$, ${Macro}_{it}$ is a ${n}_{i} x 8$ matrix of macroeconomic variables for sample $i$ at time $t$, ${\beta }_{i}$ is the $8 x 1$ vector of coefficients for sample $i$, ${Fund}_{it}$ is a ${n}_{i} x 3$ matrix of fund variables for sample $i$ at time $t$, and ${\gamma }_{i}$ is the $3 x 1$ vector of coefficients for sample $i$. Following the same logic, ${Property}_{it}$ is considered a ${n}_{i} x 2$ matrix of property variables for sample $i$ at time $t$, ${\delta }_{i}$ is the $2 x 1$ vector of coefficients for sample $i$, ${Liquidity}_{it}$ is a ${n}_{i} x 2$ matrix of the transaction variables for sample $i$ at time $t$, and ${\epsilon }_{i}$ is the $2 x 1$ vector of coefficients for sample $i$. While Eq. 1 can be applied to all dependent variables, equations 2 through 4 can only be applied to the sample and its respective subsamples; this is because property and fund variables are only considered for CEREFs. We integrate the squared values of funds and property age to test for the turning points of their respective factors.

Table 3 presents the correlation coefficients among the observed variables and shows several sets of correlations. Overall, the correlation coefficients are not of great magnitude, except for the correlations among liquidity variables, such as the number of transactions and unemployment rate, bond rate, fund age, and the sample itself. The number of transactions (.59) and the respective volume (.64) reflect the demand in the secondary market and have positive correlations with the average transaction price. The positive coefficient between the number of transactions and respective volume (.78), as well as fund age (.68), is as expected; it can be assumed that the volume of transactions increases with an increasing number of transactions and fund age. The correlation between fund size and property size (.70) is positive, as we consider only CEREFs with one property; usually, with increasing property size, there is an increase in necessary funds. The negative correlations of bond rate and unemployment rate with fund age (-.66, -.64), number of transactions (-.82, -.89), and the respective volume (-.59, -.67), respectively, result from the overall positive economic development with decreasing bond and unemployment rates during the period under review. Similar results are observed in the case of the positive coefficient between the bond rate and unemployment rate (.86), as well as the increasing transaction price and decreasing unemployment rate (-.61). However, to ensure that our econometric analyses and results are reliable, we conduct variance inflation factor analyses to confirm that multicollinearity does not exist in our models.

Testing all the equations for heteroscedasticity using the Breusch–Pagan and Cook–Weisberg tests revealed that sample 1 and sample 2 were homoscedastic, while sample 3 slightly exceeds the critical threshold for homoscedasticity. Hence, to obtain unbiased standard errors of OLS coefficients under heteroscedasticity, we use robust standard errors in all the regressions. Additionally, we test for stationarity in our models using the Dickey–Fuller and Philipps–Perron unit root tests, as well as the Kwiatkowski–Phillips–Schmidt–Shin test of stationarity, and find that all the time series models are stationary.

[5] In the period under review, our dataset includes on average every month 43 and 7 CEREFs with two and five transactions per month, respectively. Hence, a panel structure, even an unbalanced panel structure, does not exist.

In Table 4, we summarize our stepwise regression results and shows that the models’ explanatory power is good, with an adjusted R² of at least 35.5% in equation 1, comprising only macroeconomic variables, and up to 75.3% in equation 4, when fund, property, and liquidity variables are considered. Based on the Akaike and Bayesian information criteria, there is no evidence of overfitting of parameters, thus strengthening the results of the stepwise approach. Sample 2 has a higher R² than samples 1 and 3 in all equations, except for equation 4. This is due to the interaction between the chosen macroeconomic variables, which reflects overall economic development in Germany, and the composition of the samples. While sample 2 only considers properties located in Germany, sample 1 includes properties located in other parts of the world, and sample 3 focuses on the office property niche within the German real estate market. A comparison of the overall significance structure among the three samples shows that only samples 1 and 2 are comparable. This indicates that different real estate types within the CEREF set are driven differently, and that sellers and buyers should not adopt a “one valuation model fits all” approach when trading in the secondary market.

Comparing R² between sample 2 and the benchmark variables shows that RESs and the stock market are better explained and driven differently by macroeconomic variables than by CEREFs. In turn, REITs are substantially poorer explained than CEREFs. While CEREFs are driven by the unemployment rate and economic activity, the performance of the benchmark variables is homogenously driven by stock market excess returns, VDAX growth, and in the case of RESs, also by bond rate. The signs of bond rate and VDAX growth are as expected. Lower bond rates raise available funds, and thus overall demand and performance. An increase in VDAX growth reflects the increasing uncertainty of stock market participants and is thus negatively correlated with performance. However, excess stock market returns have a negative impact on the performance of listed real estate assets but a positive impact on stock markets. This relationship might be due to the anticyclical characteristic of real estate assets, in which investors tend to stay on the stock market in bullish times and move their focus toward real estate assets in bearish times. At first glance, CEREFs may be adequate for investors to further diversify their existing listed real estate and stock portfolios. However, as we focus on the impact factors of the transaction prices of German CEREFs and not on the portfolio diversification possibilities between CEREFs and different real estate asset types, we do not perform further portfolio diversification analyses.

Table 5 Regression results for benchmark variables

	DAXRE	REIT	CDAX
Inflation growth	-0.6444	-0.8916	-0.2024
Unempl. rate	0.9935	0.6576	1.0014
M1 growth	-0.5144	-0.9387	-0.4074
Bond rate	-1.7264**	-0.6451	-1.0123
Stock excess return	-0.8398***	-0.4837***	0.1538**
GDP growth	1.2550	1.3785	1.1589
VDAX growth	-0.1601***	-0.1342***	-0.1558***
Activity growth	-0.0048	-0.0004	-0.0059
Constant	-0.0358	-0.0162	-0.0472
R²	0.7191	0.3500	0.5493
Std. Error	Robust	Robust	Robust

Note: This table presents the regression results for the benchmark variables in equation 1. ***, ** and * denote the 1%, 5%, and 10% significance levels, respectively.

In accordance with the literature, the results suggest, across all three samples and thus regardless of geographic or asset class effects, strong negative relationships between CEREF transaction prices and the unemployment rate, except in equation 4. Increasing unemployment negatively affects purchasing power and consumption behavior within an economy (Schätz and Sebastian 2009), leading to lower demand and, thus, lower transaction prices. Positive relationships also exist for activity and VDAX growth in equation 1 for sample 2 and 3, respectively. Increasing VDAX growth reflects increasing uncertainty of stock market participants, and the positive sign indicates stock investors’ shift toward safe havens, such as German office properties, when stock market volatility increases. However, in the other equations, when additional factors were considered, activity and VDAX growth had no significant impact. In equation 4, the bond rate and stock market excess return become significant, similar to the benchmark variables. For stock market excess returns, the anticyclical characteristic of real estate assets in the form of a negative relationship becomes apparent, but the opposite, positive relationship between bond rate and transaction price cannot be explained. All other macroeconomic factors, namely inflation growth, M1 growth, and GDP growth, have no significant impact on the transaction prices of the German CEREFs.

Size has a weak significant impact on both the fund and property levels in a few samples and seems not to be an explanatory factor for transaction prices of German CEREFs on the secondary market.

Fund age, in contrast, has a highly significant negative impact on all equations for samples 1 and 2, and the squared age factor is positive. This reveals that older funds generate lower transaction prices; however, there is a turning point beyond which the transaction price starts to increase. The tax and return benefits of CEREFs fade over time, thereby leading to lower transaction prices and termination after a certain time. However, successful CEREFs in terms of returns have higher transaction prices and are usually perpetuated, leading to higher fund ages.

Initial fund leverage has a highly significant positive impact on all equations, contradicting the literature. For example, Case (2015), Baum, Fear, and Colley (2011), and Giacomini, Ling, and Naranjo (2015) report a negative effect of leverage on prices and returns. Van der Spek and Hoorenman (2011) find that leverage has a positive impact until a leverage level of approximately 40% but is negative at higher levels. However, our dataset does not contain yearly changing leverage levels but the initial fund leverage at the end of the placing phase. Hence, we find a highly significant positive relationship with trust. Prior to the implementation of the KAGB in 2013, issuing companies were subject to lower constraints on leverage and applied, when possible, high levels of leverage in their funds. Usually, only issuing companies with a good market reputation, who had proved in the past to be successful and thus had tendentially higher prices on the secondary market for their CEREFs, were able to collect higher amounts of debt.

A quadratic relationship exists between property age and transaction price for all equations for sample 2. The property age coefficient is positive, and the squared age factor is negative. This cluster shows that transaction prices increase in the first years but decrease after a turning point. The underlying reason might be that real estate investors prefer stable long-term letting contracts in the first years and thus property prices initially increase; however, the older a property becomes, the more investors worry about renovation costs, and when the turning point is exceeded investors start to negatively factor it into the price.[6]

The liquidity proxies, namely transaction volume and the respective number, are both significantly positive only in the office sample, and thus increase the adjusted R². Therefore, the more trade that occurs in the secondary market, the higher the transaction prices of the German CEREFs invested in offices. Arguably, this results from an increasing number of bidders, the size of the seller side being relatively constant, which increases the prices.

This study provides evidence on the importance of controlling for macroeconomic, fund, property, and market liquidity factors when sellers and buyers trade in the secondary market for German CEREFs. As regards macroeconomic factors, CEREFs and benchmark variables are driven differently and provide real estate investors with a scope for portfolio diversification. CEREFs are mainly driven by the unemployment rate and economic activity, while the performance of the benchmark variables is homogenously driven by stock market excess return and VDAX growth, and in the case of the DAXRE, by bond rate. Fund and property ages have a significant impact and optimal age levels at which the respective factors start to exert an opposite effect on the transaction price. When the respective turning points are exceeded, fund and property ages exhibit negative and positive impacts, respectively. Initial fund leverage, interpreted as a trust indicator, has a significant positive impact; the liquidity variables result from an increasing number of market participants in the secondary market, especially bidders, who push prices up. However, the abovementioned impacts vary with regard to real estate types, which investors must consider while formulating their pricing policies.

The perception of the secondary market for CEREFs has fundamentally changed from a pure liquidity provider market for sellers to a market for portfolio reallocation and the partial recognition of profits. The possibility of constructing positions in specific CEREFs and markets, as well as the existence of specialized secondary funds, show the increasing acceptance of buyers for CEREFs as an investment strategy and an established instrument of active portfolio management that generates returns. Hence, understanding the factors that impact fund pricing is indispensable.

Acknowledgements We would like to thank Fondsbörse Deutschland Beteiligungsmakler AG for making the data available. We are solely responsible for any errors.

Data availability Not applicable.

Conflict of Interest None

Aarts, Siem, and Andrew Baum. 2016. “Performance persistence in real estate private equity.” Journal of Property Research 33 (3): 1–16.
Adams, Zeno, and Roland Füss. 2010. “Macroeconomic Determinants of International Housing Markets.” Journal of Housing Economics 19 (1): 38–50.
Alcock, Jamie, Andrew Baum, Nicholas Colley, and Eva Steiner. 2013. “The Role of Financial Leverage in the Performance of Private Equity Real Estate Funds.” The Journal of Private Equity 17 (1): 80–91.
Ametefe, Frank, Steven Devaney, and Gianluca Marcato. 2016. “Liquidity: A review of dimensions, causes, measures, and empirical applications in real estate markets.” Journal of Real Estate Literature 24 (1): 1–29.
Anson, Mark J., and Susan Hudson-Wilson. 2003. “Should One Use Leverage in a Private Equity Real Estate Portfolio?” The Journal of Portfolio Management 29 (5): 54–61.
Arnold, Thomas R., David C. Ling, and Andy Naranjo. 2019. “Private Equity Real Estate Funds: Returns, Risk Exposures, and Persistence.” The Journal of Portfolio Management 45 (7): 24–42.
Arnold, Thomas R., David C. Ling, and Andy Naranjo. 2017. “Waiting to Be Called: The Impact of Manager Discretion and Dry Powder on Private Equity Real Estate Returns.” The Journal of Portfolio Management 46 (6): 23–43.
Arnold, Thomas R., David C. Ling, and Andy Naranjo. 2021. “Private Equity Real Estate Fund Performance: A Comparison to REITs and Open-End Core Funds.” The Journal of Portfolio Management 47 (10): 107–26.
Barker, David, Kiat Y. Seah, and James D. Shilling. 2019. “How Big of a Lemons Market Is the Secondary Market for Private Equity Real Estate Limited Partnerships?” The Journal of Real Estate Finance and Economics 59 (3): 391–418.
Baum, Andrew, and Kieran Farrelly. 2009. “Sources of alpha and beta in property funds: a case study.” Journal of European Real Estate Research 2 (3): 218–34.
Baum, Andrew, Jane Fear, and Nick Colley. 2011. “Have property funds performed.” Urban Land Institute Europe Policy and Practice Committee Report.
Bond, Shaun A., and Paul Mitchell. 2010. “Alpha and Persistence in Real Estate Fund Performance.” The Journal of Real Estate Finance and Economics 41: 53–79.
Brounen, Dirk, Hans O. Veld, and Ville Raitio. 2007. “Transparency in the European Non-listed Real Estate Funds Market.” Journal of Real Estate Portfolio Management 13 (2): 107–17.
Case, Brad. 2015. “What Have 25 Years of Performance Data Taught Us About Private Equity Real Estate?” The Journal of Real Estate Portfolio Management 21 (1): 1–20.
Delfim, Jean-Christophe, and Martin Hoesli. 2016. “Risk factors of European non-listed real estate fund returns.” Journal of Property Research 33 (3): 190–213.
Esrig, Dave, Michael C. Hudgins, and Luigi Cerreta. 2011. “Revisiting the impact of large assets on real estate portfolio returns.” The Journal of Portfolio Management 37 (5): 125–36.
Fairchild, Simon, Greg MacKinnon, and John Rodrigues. 2011. “Are All Open-End Core Funds Created Equal?” The Journal of Portfolio Management 37 (5): 51–67.
Farrelly, Kieran, and George A. Matysiak. 2012. “Performance Drivers of UK Unlisted Property Funds.” Working Papers in Real Estate & Planning. 05/12. Working Paper. University of Reading, Reading. http://centaur.reading.ac.uk/29401/1/0512.pdf.
Farrelly, Kieran, and Simon Stevenson. 2019. “The Risk and Return of Private Equity Real Estate Funds.” Global Finance Journal 42 (1).
Fisher, Lynn M., and David J. Hartzell. 2016. “Class Differences in Real Estate Private Equity Fund Performance.” The Journal of Real Estate Finance and Economics 52 (4): 327–46.
Fuerst, Franz, Wayne Lim, and George Matysiak. 2013. “Non-Listed Real Estate Funds: Leverage and Macroeconomic Effects.” The Narodowy Bank Polski Workshop: Recent Trends in the Real Estate Market And its Analysis. SSRN: https://ssrn.com/abstract=2841716 or http://dx.doi.org/10.2139/ssrn.2841716.
Fuerst, Franz, Wayne T. Lim, and George A. Matysiak. 2014. “Private Equity Real Estate: Asymmetric Effects of Gearing and Drivers of Performance.” SSRN: https://ssrn.com/abstract=2519794 or http://dx.doi.org/10.2139/ssrn.2519794.
Fuerst, Franz, and George A. Matysiak. 2013. “Analysing the Performance of Nonlisted Real Estate Funds: a Panel Data Analysis.” Applied Economics 45 (14): 1777–88.
Gerlach, Philipp, and Raimond Maurer. 2020. “The Growing Importance of Secondary Market Activities for Open-end Real Estate Fund Shares in Germany.” Schmalenbach Business Review 72: 65–106.
Giacomini, Emanuela, David C. Ling, and Andy Naranjo. 2015. “Leverage and returns: A cross-country analysis of public real estate markets.” The Journal of Real Estate Finance and Economics 51 (2): 125–59.
Hahn, Thea C., David Geltner, and Nori Gerardo-Lietz. 2005. “Real Estate Opportunity Funds.” The Journal of Portfolio Management 31 (5): 143–53.
Harris, Robert S., Tim Jenkinson, and Steven N. Kaplan. 2014. “Private Equity Performance: What Do We Know?” The Journal of Finance 69 (5): 1851–82.
Jenkinson, Tim, Miguel Sousa, and Rüdiger Stucke. 2013. “How Fair are the Valuations of Private Equity Funds?” SSRN: https://ssrn.com/abstract=2229547 or http://dx.doi.org/10.2139/ssrn.2229547.
Kaplan, Steven N., and Antoinette Schoar. 2005. “Private Equity Performance: Returns, Persistence, and Capital Flows.” The Journal of Finance 60 (4): 1791–1823.
Kiehelä, Sami, and Heidi Falkenbach. 2015. “Performance of Non-Core Private Equity Real Estate Funds: A European View.” The Journal of Portfolio Management 41 (6): 62–72.
Kohlert, Daniel M. 2010. “The Determinants of Regional Real Estate Returns in the United Kingdom: a Vector Error Correction Approach.” Journal of Property Research 27 (1): 87–117.
Krautz, Sebastian, and Franz Fuerst. 2015. “Size Signals Success: Evidence from Real Estate Private Equity.” The Journal of Portfolio Management 41 (6): 73–81.
Kurzrock, Björn-Martin, Sebastian Gläsner, and Elaine Wilke. 2009. “Performance of open‐ended real estate funds in Germany: Are institutional investors better off than retail investors?” Journal of European Real Estate Research 2 (2): 110–31.
Mattarocci, Gianluca, and Xenia Scimone. 2020. “Closed End Real Estate Funds and Real Estate Portfolio Management: Evidence from the Italian Retail Market.” Journal of Real Estate Portfolio Management 26 (1): 27–36.
Morri, Giacomo, Ugo Perini, and Rachele Anconetani. 2021. “Performance determinants of European private equity real estate funds.” Journal of European Real Estate Research 14 (2): 209–28.
Nadauld, Taylor D., Berk A. Sensoy, Keith Vorkink, and Michael S. Weisbach. 2019. “The liquidity cost of private equity investments: Evidence from secondary market transactions.” Journal of Financial Economics 132 (3): 158–81.
Nadler, Michael. 2018. “Performance vs. prospectus = transparency in German closed-ended real estate funds?” Journal of Property Investment & Finance 36 (2): 158–70.
Pagliari, Joseph L., Kevin A. Scherer, and Richard T. Monopoli. 2005. “Public Versus Private Real Estate Equities; a More Refined, Long-Term Comparison.” Real Estate Economics 33 (1): 147–87.
Phalippou, Ludovic, and Oliver Gottschalg. 2009. “The Performance of Private Equity Funds.” The Review of Financial Studies 22 (4): 1747–76.
Schätz, Alexander, and Steffen Sebastian. 2009. “The links between property and the economy – evidence from the British and German markets.” Journal of Property Research 26 (2): 171–91.
Schnejdar, Sebastian, Michael Heinrich, René-Ojas Woltering, and Steffen Sebastian. 2020. “The Discount to NAV of Distressed Open-End Real Estate Funds.” The Journal of Real Estate Finance and Economics 61: 80–114.
Stein, Michael. 2013. “German Open Ended Real Estate Fund Performance: the Impact of Liquidity.” Credit and Capital Markets 46 (1): 119–51.
Stein, Michael. 2014. “German real estate funds: changes in return distributions and portfolio favourability.” Journal of European Real Estate Research 7 (1): 87–111.
Tomperi, Ilkka. 2010. “Performance of private equity real estate funds.” Journal of European Real Estate Research 3 (2): 96–116.
van der Spek, Maarten, and Chris Hoorenman. 2011. “Leverage: please use responsibly.” Journal of Real Estate Portfolio Management 17 (2): 75–88.

Tables 3 and 4 are available in the Supplementary Files section.

Tables3and4.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Transaction Prices of German Closed-End Real Estate Funds in the Secondary Market and Their Influencing Factors

Status:

Version 1

Abstract

Figures

1 Introduction

2 Literature Review

3 The Secondary Market For Cerefs In Germany

4 Data

5 Methodology

6 Results

7 Conclusion

Declarations

References

Tables 3 and 4

Supplementary Files

Status:

Version 1