The influence of ECB’s conventional and unconventional
monetary policy announcements on daily European REIT
stock returns
Name:
Floris Pigeaud
University:
University of Amsterdam
Faculty:
Faculty of Economics and Business
Programme:
MSc Finance and Real Estate Finance
Document:
Master Thesis
Supervisor:
Marcel Theebe
Date:
July 2017
2
3
Abstract
Using hand-‐collected data for both announcements about conventional and unconventional monetary policy, in this thesis the effect of ECB’s conventional (i.e. changing the key interest rate) and unconventional monetary policy announcements on daily European REIT stock returns is analysed. The interest rate changes are separated into an expected and unexpected part using three-‐month EURIBOR futures data. Furthermore, the measure for unconventional monetary policy decisions is created by the change in the spread between German and Italian 10-‐year government bond yields. Results show that daily REIT stock returns are affected by both conventional and unconventional monetary policy. However, further research is needed to be more conclusive about the exact effects of monetary policy.
4
Table of Contents
List of tables and figures ... 5
List of abbreviations ... 6
1. Introduction ... 8
2. Literature review ... 11
2.1 Conventional and unconventional monetary policy ... 11
2.2 REIT valuation and the effect of interest rates ... 14
2.3 Effect of unconventional tools on REIT prices ... 18
2.4 Past research ... 18
2.4.1 Relationship between interest rates and stock returns ... 19
2.4.2 Unconventional tools and stock returns ... 20
2.4.3 Difference between general stock and REIT stocks ... 21
2.4.4 Interest rates and REITs ... 22
3. Hypotheses ... 25
4. Methodology ... 29
4.1 Disentangling the key interest rate change ... 30
4.2 Unconventional monetary policy surprise ... 32
4.3 Effect of the interest rate changes on the whole sample of REITs ... 32
4.3.1 Baseline regression ... 33
4.3.2 Sign of the interest rate change and policy inactions ... 36
4.3.3 Crisis ... 38
4.3.4 Asymmetric reaction crisis ... 39
4.3.5 Event of reversal ... 41
4.4 Individual regressions ... 43
5. Data ... 44
5.1 Sample of REITs ... 44
5.2 Time-‐period ... 44
5.3 Daily stock returns ... 44
5.4 (Un)expected interest rate changes and unconventional tool surprises ... 45
5.5 Policy meetings ECB ... 47
5.6 Announcements about (un)conventional monetary policy ... 47
5.7 Policy reversal ... 48
5.8 Sign of the interest rate and policy inactions ... 49
6. Results ... 49
6.1 Baseline regression ... 50
6.2 Asymmetric effect and policy inactions ... 52
6.3 Influence of the crisis ... 55
6.4 Monetary policy reversals ... 58
6.5 Individual regressions ... 59
7. Robustness check ... 61
8. Conclusion and Discussion ... 62
References ... 69
5
List of tables and figures
-‐‑ Table 1 Expected sign coefficients first regression
-‐‑ Table 2 Expected sign coefficients second regression
-‐‑ Table 3 Expected sign coefficients third regression
-‐‑ Table 4 Expected sign coefficients fourth regression
-‐‑ Table 5 Expected sign coefficients fifth regression
-‐‑ Table 6 Expected sign coefficients sixth regression
-‐‑ Table 7 Summary statistics
-‐‑ Table 8 Frequency of interest rate changes
-‐‑ Table 9 Policy reversals
-‐‑ Table 10 Summary positive/negative changes and policy inactions
-‐‑ Table 11 Results first regression: Baseline regression
-‐‑ Table 12 Results second regression: Sign and policy inactions
-‐‑ Table 13 Results third and fourth regression: Crisis/Asymmetry during the crisis
-‐‑ Table 14 Results fifth regression: Reversal effect
-‐‑ Table 15 Percentage changes individual regression
-‐‑ Table 16 Summary percentage changes individual regression
-‐‑ Table 17 Sample of REITS
-‐‑ Table 18 Announcements and corresponding policy rate changes
-‐‑ Table 19 Announcements of unconventional monetary policy
-‐‑ Table 20 Results individual regressions, full sample
-‐‑ Table 21 Results individual regressions, excluding possible outliers
-‐‑ Table 22 Crisis and asymmetry during the crisis (test for robustness)
-‐‑ Figure 1 𝑅"# and ∆𝑟
"&, full sample
-‐‑ Figure 2 𝑅"# and ∆𝑟
"&, excluding outliers
6
List of abbreviations
Abbreviation Explanation
∆𝑟"' The expected interest rate change
∆𝑟"'𝐶" Expected interest rate change during the crisis
∆𝑟"'𝐷"* Expected interest rate change for hikes in the Deposit
Facility rate
∆𝑟"'𝐷"*𝐶" Expected interest rate change for hikes in the Deposit
Facility rate during the crisis
∆𝑟"' 𝐷", Expected interest rate change at events of reversal
∆𝑟"&𝐷"- Unexpected interest rate change at events of
monetary policy inaction
∆𝑟"&𝐷"-𝐶" Unexpected interest rate change at events of
monetary policy inaction during the crisis ∆𝑟"& Unexpected interest rate change
∆𝑟"&𝐶" Unexpected interest rate changes during the crisis
∆𝑟"&𝐷"* Unexpected interest rate change for hikes in the
Deposit Facility rate
∆𝑟"&𝐷"*𝐶" Unexpected interest rate change for hikes in the
Deposit Facility rate during the crisis ∆𝑟"&𝐷
", Unexpected interest rate change at events of
reversal
𝐶" Crisis dummy, (1 during crisis, 0 otherwise) 𝐷"- Dummy variable, 1 when there was an policy
inactions
𝐷"* Dummy variable, 1 when there was a positive
interest rate change in the Deposit Facility 𝐷", Dummy variable, 1 at events of reversal
𝑓/,"12 Future rate at day t -‐ 1 𝑓/," Future rate at day t
𝑃" Share price at time t
𝑃"12# Price of security i at day t-‐1
𝑃"# Price of security i at day t
𝑟4 Risk-‐free rate
𝑅"# Daily stock return of security i at day t
𝑟5677 Weighted average cost of capital
𝑦/,"129 Yield of the 10-‐year German government bond at day
t-‐1
𝑦/,"12- Yield of the 10-‐year Italian government bond at day t-‐
1
𝑦/,"9 Yield of the 10-‐year German government bond at day
t
𝑦/,"- Yield of the 10-‐year Italian government bond at day t
% DF Interest rate of the Deposit Facility
% MLF Interest rate of the Marginal Lending Facility % MRO Fixed Fixed interest rate for the Main Refinancing
Operations
% MRO Variable Variable interest rate for the Main Refinancing Operations
%∆ DF Percentage change in the Deposit Facility rate %∆ MLF Percentage change in the Marginal Lending Facility
7
%∆ MRO Percentage change in the Main Refinancing rate ∆𝑔"; Measure of the surprise in unconventional monetary
policy
ABSP Asset-‐Backed Securities Programme
APP Asset Purchase Programme
BoJ Bank of Japan
BoE Bank of England
CBPP Covered Bond Purchase Programme
CET Central European Time
CSPP Corporate Sector Purchase Programme
DF Deposit Facility
DFCF-‐model Discounted Free Cash Flow model
EBIT Earnings Before Interest and Taxes
ECB European Central Bank
EURIBOR The Euro Interbank Offered Rate
Euro Stoxx The daily return of the Euro Stoxx 50
FED Federal Reserve
FFO Funds From Operations
FOCM Federal Open Market Committee
LTRO Long-‐Term Refinancing Operations
MLF Marginal Lending Facility
MRO Main Refinancing Facility
NAV Net Asset Value
NIY Net Initial Yield
OMT Outright Monetary Transactions
PSPP Public Sector Purchase Programme
QE Quantitative Easing
REIT Real Estate Investments Trust
S&P 100 Standard & Poor’s 100 S&P 500 Standard & Poor’s 500
SMP Securities Market Programme
US United States
Vt Firm value at time 0
Δ𝑟" Key interest rate change
8
1. Introduction
In 2007, the worst financial crisis started since The Great Depression in the 1930s. Before this financial crisis, price stability was the main aim of the ECB, i.e. to safeguard the value of the euro (ECB). Price stability is important for both job creation and economic growth, which are the two European Union’s objectives. Although the ECB cannot influence these goals directly, it can influence them indirectly using conventional and unconventional tools. The most direct effect of those tools (e.g. changing the key interest rate) is on the financial markets. By influencing the financial markets, policymakers try to change the economic behaviour of the market participants in such a way that it will help them to achieve their ultimate goals. After the financial crisis, the prevention of asset price bubbles became another task of the ECB. For example, Reinhart and Rogoff (2009) provide evidence that collapses in prices of commercial real estate markets, residential or collapses in both markets, are one of the major causes of financial crises. In most cases the collapses are caused by bubbles in the real estate prices, which appear to be associated with excessive availability of credit and loose monetary policy. When these bubbles burst the real economy and the financial sector are harmed, which the ECB is trying to avoid. Therefore, it is crucial to understand the effect of monetary policy on the real estate market. In this thesis, the effect of monetary policy is analysed by looking at the Real Estate Investment Trust (REIT) market in Europe. The REIT market is an interesting market to look at, because of their corporate structure and underlying portfolio. Due to this, there are multiple channels through which the financial, operating and share price performance can be influenced.
As we currently are in an environment in which the interest rates are at the zero lower bound, the ECB conducts, besides conventional monetary policy (i.e. changing the key interest rate), also unconventional monetary policy. Unconventional monetary policy is aimed at providing additional bank liquidity, to directly target liquidity shortages and to decrease the credit spreads in certain market areas. In general, unconventional tools are measures used to directly target the availability and the cost of external funds. The reason why unconventional monetary policy is conducted, is that the transmission mechanism of the conventional tools does not work properly when interest rates are at their zero lower bound. By using unconventional tools, the ECB tries to have additional influence on the economy. But, the
9 question remains how the economy is influenced by the unconventional monetary policy. And especially, what is the effect of unconventional monetary policy on the REIT market in Europe? In March this year, the FED decided to raise the interest rate from 0.75% to 1%. This raises not the question whether the ECB is also going to increase the key interest rate, but when this will happen. Institutions, investors and other market participants may question what will happen to the economy when the interest rates are increased after a period of being at the zero lower bound. Therefore, it is crucial to get more insight into the effects of monetary policy. This raises also the question how the REIT market reacts to conventional monetary policy.
This thesis is an empirical study, which tries to investigate the relationship between announcements of ECB’s conventional and unconventional monetary policy on daily REIT stock returns on ECB’s announcement days. I look at the effect on the announcement day of future monetary policy, because on that day news was transmitted to market participant. The response to interest rate changes is complicated, because it is likely that market participants only react to unexpected monetary policy changes on announcement days. When market participants do not expect decisions made by the ECB, then this new information will immediately be incorporated into the economy. Therefore, it is important to distinguish between expected and unexpected interest rate changes, because otherwise the independent variables would be wrongly specified. This would create an errors-‐in-‐variables problem which influences the results. The method suggested by Kuttner (2001), is used to conduct this analysis. The main idea behind this methodology is that the futures market reflects market expectations of future monetary policy. Changes in the price on the futures market could suggest that certain monetary policy decisions were not anticipated. Following Bernoth and von Haagen (2004), a reliable predictor for the policy rates of the ECB is the three-‐month EURIBOR future rates. Therefore, this future rate is used for the creation of the unexpected interest rate, which is thereafter used to create the expected interest rate.
As already said, this thesis also investigates the effect of unconventional monetary policy decisions. The methodology suggested by Rogers et al. (2014) is used to create a proxy for surprises in unconventional monetary policy. They use the change in the spread between German and Italian 10-‐year government bond yields on the day of the monetary policy
10 announcement. Following Rogers et al (2014), this can be used, because the goal of the ECB’s unconventional policy was to quite some extent aimed at decreasing the intra-‐euro area sovereign spreads. The spread will change, when market participants did not expect decisions about future unconventional monetary policy. Therefore, it is a measure for surprises in unconventional monetary policy. After the determination of the (un)expected interest rate changes and the proxy for unconventional monetary policy, regressions are performed using hand-‐collected announcement data.
This thesis has the following contributions to our current knowledge. First, the effect of conventional monetary policy changes on daily REIT stock returns is analysed by looking at the effect of (un)expected interest rate changes, both before and during the crisis. Second, the effect of unconventional monetary policy on daily European REIT stock returns is examined. To my knowledge this is the first study that investigates the effect of ECB’s conventional monetary policy using this methodology. Also, this is the first paper that investigates unconventional monetary policy changes on daily European REIT stock returns. Third, the possibility for asymmetry to positive and negative interest rate changes is analysed, both before and during the crisis. Fourth, the reversal effect of conventional monetary policy is analysed. Policy reversals are events at which a central bank changes the sign of the policy rate change, in comparison with one policy announcement prior to that announcement. Fifth, the reaction to monetary policy inactions is considered. Monetary policy inactions are announcements at which the ECB decide not to change the conventional monetary policy. Sixth, regressions are performed to see whether REITs react differently to conventional and unconventional monetary policy changes. To my knowledge, this is the first paper that examines these aspects for the REIT market. Besides these contributions, I also use hand-‐ collected data for announcements about conventional monetary policy. For the announcements of unconventional monetary policy, I manually updated the list created by Rogers et al. (2014).
First, I explain in my Literature Review, existing theories and past research. Using the information of my Literature Review, I create hypotheses, which I describe in the Hypotheses section. Third, I describe my data. Fourth, the methodology will be discussed. Hereafter, the
11 results of my regressions will be discussed and interpreted. Sixth, there will be a robustness check of my results. Last, there will be a conclusion with a discussion.
2. Literature review
In this section, I am going to discuss existing theory and past research. First, I am going discuss the conventional and unconventional policy of the ECB. Second, I am going to explain the relationship between ECB’s conventional tools and REIT stock prices, by discussing two REIT stock pricing models. Third, the effect of unconventional monetary policy on REIT stock returns is discussed. Fourth, I outline the past research.
2.1 Conventional and unconventional monetary policy
As already discussed, the central bank uses both conventional and unconventional tools to influence the economy. The conventional tools of a central bank, is setting a target for the interest rate in the interbank money market and the adjustment of the supply of money of that target by using open market operations. By changing the level of the key interest rate, the central bank effectively influences the liquidity conditions in the market and tries to achieve their main objective: price stability over the medium term (Smaghi 2009).
By altering the key interest rate, the ECB influences three interest rates: the interest rate for the Deposit Facility (DF), the interest rate for the Marginal Lending Facility (MLF) and the interest rate for the Main Refinancing Operations (MRO) (ECB). Following Haitsma, Unalmis en de Haan (2016), the most important interest rate is the rate for the MRO. The MRO rate is the interest rate, either variable or fixed, which is used for the main refinancing operations. The ECB provides through these operations liquidity to financial institutions in exchange for collateral. Before the crisis, it was a weekly operation with maturities of one week up to three months. By changing the interest rate of the MRO and by changing the amount of liquidity provided the ECB can influence both market interest rates and liquidity.
Another way of influencing the liquidity of banks is by altering the interest rate of the Deposit Facility. Overnight, banks can deposit an unlimited amount of money at the central bank at the Deposit Facility interest rate. A third way of affecting the liquidity is changing the interest rate of the Marginal Lending Facility. This enables banks to lend money at the Marginal Lending Facility rate. The two facilities are the standing facilities of the central bank and form the corridor for overnight interest rates (ECB).
12 During normal times, this policy has proved to be a reliable way to stimulate the economy in downturns and to depress the economy in upturns. However, during abnormal times, conventional monetary policy may be less effective, which means that the ECB needs other ways to achieve their ultimate objective.
There are two reasons why a central bank would need other measures: The first reason is when an economic shock is so severe, that the interest rate must be lowered to the zero lower bound. At this level, it is impossible to stimulate the economy by decreasing the interest rates. Additional monetary stimulus could be generated by conducting unconventional monetary policy, which is aimed at providing additional bank liquidity, to directly target liquidity shortages and to decrease the credit spreads in certain market areas. In general, unconventional tools are measures used to directly target the availability of external funds and the cost of external funds.
The second reason why it would be beneficial to use unconventional tools is when the monetary transmission process is less effective. This could be even when the interest rate is not at its zero lower bound. So, the central bank could use unconventional tools to have an additional influence on the economy. However, there exists a risk of hindering the functioning of the markets, when conducting unconventional monetary policy. The financing conditions could become too attractive, so that market participants become too dependent on operations conducted by the central bank. This could cause a situation in which it is hard to get back to normal market conditions.
After answering the question why it is important to conduct unconventional policy, it is important to explain what unconventional monetary policy is. Unconventional monetary policy consists of different tools. Besides the MRO’s, there are also long term refinancing operations (LTRO’s). These operations are used to finance euro zone banks at very low interest rates, with sovereign debt as collateral. The LTRO’s are offered on a monthly basis and have maturities of either three months, six months or one year. However, in December 2011 the ECB announced a three-‐year maturity LTRO, which increased the demand for these kind of loans. The LTRO causes two things: it increases bank liquidity and lowers sovereign debt yields. The bank liquidity is increased, because banks can borrow cheaper. This means that banks can lend more, and this may increase economic activity. It also enables banks to invest in higher
13 yield assets to earn a profit. The sovereign debt yields are lowered, because the banks can use their own sovereign debt as collateral. This raises demand for bonds and thus lowers yields. The LTRO’s are conducted using an auction. The ECB sets a fixed amount of liquidity and banks can bid up each other to get the available liquidity (ECB).
Another unconventional tool was the Securities Market Programme (SMP), which was initiated in May (2010). It was started, because there were tensions in the euro-‐area public and private sovereign debt markets. This programme intervenes in both the public and private securities markets to ensure depth and liquidity, especially in the markets that are less functioning (ECB). The goal of this programme is to restore the transmission mechanism of monetary policy. It reduced the liquidity premia and both the level and the volatility of European government bond yields. The influence of the operations is sterilized, by using other operations to re-‐absorb the injected liquidity, to ensure that the monetary policy stance is not influenced. In September 2012, the SMP was stopped, by the introduction of the Outright Monetary Transactions (OMT). This programme was introduced to buy an unlimited amount of government bonds (ECB).
The ECB also has an Asset Purchase Programme (APP). This programme consists of the Covered Bond Purchase Programme (CBPP), Asset-‐Backed Securities Programme (ABSP), Public Sector Purchase Programme (PSPP) and the Corporate Sector Purchase Programme (CSPP). In 2009, the ECB introduced the Covered Bond Purchase Programme (CBPP). This programme was aimed at reviving the covered bond market, which was an important source of funding for banks. The operations were, in contrast with the SMP not sterilized (ECB).
In 2014, the ECB started to buy asset-‐backed securities with their ABSP. The securities were bought on the primary and secondary markets to enhance the transmission mechanism of monetary policy, support the economy with liquidity. The main goal of this is to get closer to the ultimate goal: stable inflation (ECB).
In March 2015, the ECB introduced the Public Sector Purchase Programme. This allowed Euro system central banks to purchase government bonds on the secondary markets and debt securities of European agencies and institutions. Since March 2015 the Euro system has bought around €60 billion worth of securities every month. From March 2016 onwards, this number even increased to €80 billion. Another name, which is maybe far more common,
14 is quantitative easing (QE). The goal of QE is to decrease interest rates and so increase the economic activity in the Euro area (ECB).
Since 2016 the Euro system has also bought corporate bonds via their Corporate Sector Purchase Programme (CSPP). Six central banks were allowed to buy investment grade bonds on the primary and secondary market.
2.2 REIT valuation and the effect of interest rates
As discussed in the introduction, I am going to look at the effect of conventional and unconventional tools on the stock returns of REITs. I think it is worthwhile to explain theory about the pricing of REIT stock. This will give us the opportunity to discuss the effect of interest rate changes in the context of the real estate market.
One of the methods to value REIT stock, is the Net Asset Value (NAV) based valuation (Geltner and Miller 2014). This method is based on the fact that the assets of REITs are directly traded in the private property market. One could estimate the value of a REIT, by looking at the value of these assets. The first step of this valuation is to come up with a value of all the properties held by the REIT, as these properties would currently be valued in the private property market. The price of a property can be calculated using the Direct Capitalization method, which is shown in equation 1.
(1) 𝑃𝑟𝑜𝑝𝑒𝑟𝑡𝑦 𝑝𝑟𝑖𝑐𝑒 = IJ" KLJM6"#NO -N7PQJIJ" -N#"#6R S#JRT
The Net Operating Income is divided by the corresponding discount rate, the Net Initial Yield, to come up with a property price. After the valuation of the whole portfolio, one must subtract the liabilities of the REIT, to get the NAV. Dividing this number by all the shares outstanding, will give us the stock price of the REIT (See equation 2).
(2) 𝑅𝐸𝐼𝑇 𝑠𝑡𝑜𝑐𝑘 𝑝𝑟𝑖𝑐𝑒 =\]6MJ/ K;"/"6NT#NOIZ[
Another model to value REITs is the Discounted Free Cash Flow Model (DFCF model). This model uses the free cash flow of the firm to calculate the share price. The free cash flow is the
15 amount of money a firm generates, before any payments are made to the either debt or equity holders. For REIT valuation, the Free Cash Flow is also known as the Funds From Operations (FFO). It is a measure that shows the amount of cash that is generated from their operations, which is shown in equation (3).
3 𝐹𝐹𝑂 = 𝑁𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒 + 𝑑𝑒𝑝𝑟𝑒𝑐𝑖𝑎𝑡𝑖𝑜𝑛 𝑎𝑛𝑑 𝑎𝑚𝑜𝑟𝑡𝑖𝑧𝑎𝑡𝑖𝑜𝑛 + 𝑖𝑚𝑝𝑎𝑖𝑟𝑚𝑒𝑛𝑡 𝑐ℎ𝑎𝑟𝑔𝑒𝑠 + 𝑙𝑜𝑠𝑠𝑒𝑠 𝑓𝑟𝑜𝑚 𝑠𝑎𝑙𝑒 𝑜𝑓 𝑝𝑟𝑜𝑝𝑒𝑟𝑡𝑦 − 𝑔𝑎𝑖𝑛𝑠 𝑓𝑟𝑜𝑚 𝑠𝑎𝑙𝑒 𝑜𝑓 𝑝𝑟𝑜𝑝𝑒𝑟𝑡𝑦
The firm value, 𝑉l, is calculated by taking the present value of all future expected free cash
flows a firm will generate (see equation (4) and (5)). The DFCF model uses the weighted average cost of capital, 𝑟5677. The reason for this is, that the free cash flow is used to pay off
both debt and the equity holders. Therefore, the cost of equity cannot be used. The weighted average cost of equity is calculated by taking the weighted average of the cost of debt and equity.
(4) V0 = PV (Future Funds From Operations)
(5) 𝑉l =2*MmmKn opqq+ mmKr (2*Mopqq)r+ ⋯ + mmKv*[v (2*Mopqq)v
To come up with a share price, the value of the firm is, after the addition of cash and the subtraction of debt, divided by the total shares outstanding, which is shown in equation (6).
(6) 𝑃" =
( [w * x6/]w1yJz"w )
\]6MJ/ K;"/"6NT#NOw
Considering these two valuation methods the effects of interest rate changes will be discussed. Following Palmon and Yaari (1981), there are two theories which try to explain the relationship between interest rates and stock returns: The Expected Real Interest Rate theory and the Expected Inflation theory. Both theories make use of a theory of Fisher: the Fisher equation. The interest rate is split up into two parts, the real interest rate and the expected inflation, which is shown in equation (7).
16 (7) 𝑁𝑜𝑚𝑖𝑛𝑎𝑙 𝑖𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝑟𝑎𝑡𝑒 = 𝑟𝑒𝑎𝑙 𝑖𝑛𝑡𝑒𝑟𝑒𝑠𝑡 𝑟𝑎𝑡𝑒 + 𝑒𝑥𝑝𝑒𝑐𝑡𝑒𝑑 𝑖𝑛𝑓𝑙𝑎𝑡𝑖𝑜𝑛
First, the Expected Real Interest Rate theory. This theory assumes that the stock prices are affected by the first component of the interest rate, the real interest rate. This rate is the interest rate that an investor gets after correcting for inflation. There are two ways in which the real interest rate affects stock prices: directly and indirectly.
The direct effect is the effect on the discount rate. An increase in the nominal interest rate can be caused by an increase of the real interest rate. When the real interest rate increases, it is likely that the real discount rate, the rate which is used to discount the future expected cash flows, increases as well. Considering the two stock pricing models, an increase in the discount rates (NIY and 𝑟5677) will affect stock prices negatively. The size of this effect
relies upon the extent to which investors derive their required return on investment from the real interest rate.
The indirect effect is due to the negative correlation between real output and an increase in real interest rates. Akimov et al. (2015) say that when the interest rate changes are of sufficient magnitude, then the operating performance and cash flows to the firm are influenced. This could be caused by the impact on occupational demand and all related issues that accompany this. The related issues are: rental growth, occupancy and vacancy rates. Papers of DiPasquale and Wheaton (1992) and Fisher (1992) have demonstrated that the economic activity is directly linked to the property market. Bredin, O’Reilly and Stevenson (2011) explain the relationships of the related issues. The key issue here is that the underlying private real estate market is influenced by interest rate movements. Interest rate changes could alter overall economic activity. This could influence the occupational demand in the underlying real estate market, which could influence the obtainable rental values from properties. This influences the cash flow of the REITs, and therefore also the dividends. Considering the stock pricing models, when cash flows and dividends go down, stock prices go down too (ceteris paribus).
However, there are scientist that do not agree with the real interest rate theory. For instance, Bernanke and Kuttner (2005), say that the reaction of equity price is, for most part, not directly attributable to changes in the real interest rate. However, this does not mean that the interest rate does not affect stock prices at all. Following Bernanke and Kuttner (2005),
17 the stock returns are affected by either the expected future excess returns or the expected future dividends, which are influenced by the interest rate.
The second theory of Palmon and Yaari (1981), is the Expected Inflation theory. This theory is about the second component of the Fisher equation: the expected inflation. As already stated, an increase in the nominal interest rate could also be caused by an increase in inflation. When the inflation increases, it is likely that the inflation premium will go up too. The inflation premium is caused by a progressive tax system. Following Feldstein (1982), the tax system is based on nominal income, not on real income. The interaction between the tax system and the rate of inflation causes an increase in tax rates when inflation increases. An increase in tax rates, also increases the tax liabilities of market participants. So, this theory states that a higher inflation premium could lead to a decrease in the after tax real dividends. This means that market participants want to be compensated for their loss: they will hold securities for a lower price. Feldstein (1982) also argues that, when there is a higher inflation rate, the effective tax rate on equity earnings relative to other types of investment income, is raised too. Again, institutions and individuals will only hold stocks when they get compensated for this loss: they will hold stock for a lower price.
Also, the ECB explains how stock prices could be influenced by changes in the target interest rate. Following the ECB (2008), monetary policy exerts an immediate impact on the perceived riskiness of certain assets and the risk compensation required by investors. The increased riskiness could be caused by an interest rate increase. This could lead to an increase in funding cost and the weakening of a firms’ balance sheet, which could cause a decrease of asset prices. Also, the ECB argues that the risk appetite of markets’ participants could be changed, due to the change in the risk premia. This is another reason why interest rates affect stock prices.
Allen and Madura (2000), say that the general value of real estate can be influenced by the cost of financing. This cost affects the affordability and the demand for real estate. When interest rates go down cost of financing goes down as well. The affordability increases (ceteris paribus), which could lead to a price increase. An interest rate hike, could lead to an increase of the cost of debt financing. This could increase the debt expenses, which could lower the cash flow of a firm. According to the DFCF model, the value of a stock decreases.
18 However, Allen and Madura (2000) also point out that the relationship between interest rates and equity REITs may be questionable, because of the underlying forces that cause interest rate changes. Low inflationary expectations and weak economic conditions could result into low interest rates. An increase in the interest rate could mean that there is higher inflationary expectations and stronger economic growth, which would result in a stronger upward pressure on real estate prices. This may offset the effect of the hypothesized inverse relationship between property values and interest rates. My analysis will give us more insight whether this is true or not.
2.3 Effect of unconventional tools on REIT prices
In the last part, I explained the possible effects of conventional monetary policy on stock returns of REITs. In this part, I discuss the effects of unconventional monetary policy. As already mentioned, unconventional tools are used to increase the liquidity and decrease the cost of financing. The central bank not only influences the financial institutions, but also other institutions and individuals. When the ECB conducts expansionary unconventional monetary policy the financing cost becomes less expensive, because bond yields are decreased. As already discussed, when financing costs go down, demand for real estate properties can go up (ceteris paribus). Furthermore, when the financing costs go down, the discount rate, which is used to value properties, could also go down. This means that, considering the Direct Capitalization method, real estate values go up. When the ECB conducts expansionary unconventional policies, the real estate market will be positively affected.
2.4 Past research
The last sections described existing theory to have more insight into the effect interest rate changes might have on the stock returns. In this part, I outline the research that looked at the effect of interest rate changes on stock returns. First, I discuss the research about the relationship between interest rates and stock returns. Second, I describe the literature about the relationship between unconventional tools and stock returns. Hereafter, I explain why REITs would reaction differently to interest rate changes compared to general stock. Last, I outline the research that is done so far about the relationship between interest rate changes and REIT returns.
19 2.4.1 Relationship between interest rates and stock returns
There are a lot of studies that have investigated the relationship between stock returns and monetary policy. Researchers have found a negative relationship between stock returns and interest rate changes. This relationship is both investigated in the US and in Europe. However, it is examined in the US more extensively. In the US, a negative relationship has been found (Thorbecke (1997), Rigobon and Sack (2002) and Ehrmann and Fratzschwer (2004)). Also, in US the result is found that smaller firms are more affected by interest rate changes, which suggest a heterogeneous reaction to interest rate changes between firms (Thorbecke (1997). Another difference between the findings in the US and Europe is that a heterogeneous reaction between sectors is found in the US (Ehrmann and Fratzschwer (2004)), while within Europe the results are mixed. No heterogeneous reaction has been found by Bredin et al. (2009), yet a heterogeneous reaction has been found by Haitsma, Unalmis and de Haan (2016). A possible explanation for this is that the time-‐periods of both studies are different (1989-‐1999 and 1999-‐2015 respectively).
Authors argue that it is important to split up the interest rates into different parts (for example an expected and unexpected part), to eliminate the errors in variables problem. This problem arises when the independent variables are measured with error or a wrong specification is used. When the interest rate is split up into different parts results are mixed. In the US, a negative relationship between the unexpected interest rate (Aharony, Saunders and Swary (1986) and Bernanke and Kuttner (2005)) and a positive relationship between the expected interest rate and stock return has been found (Bernanke and Kuttner (2005)). Both papers make us of the method of Kuttner (2001), which will be more extensively discussed in the methodology. They make a distinction between an expected and unexpected interest by using the futures-‐market based proxy for the unexpected interest rates changes.
In contrast to the results found in the US, the results for Europe are mixed. Bohl, Siklos and Sondermann (2008) find a negative relationship for the unexpected interest rate change, whereas another paper find no relationship (Bredin et al. (2009)). A possible explanation for the different results is that both papers make use of a different methodology and time-‐period. The difference between the closed economy of the US and the open economy of Europe is also put forth to explain the difference between the relationship in the US and Europe.
20 Furthermore, results show that the reaction to monetary policy is different between crisis periods and non-‐crisis periods (Kontonikas et al. (2013) and Fiordelisi, Gallopo and Ricci (2014)). A negative relationship has been found before the crisis, while no negative relationship has been found during the crisis. An explanation for this is that interest rate decreases were a sign for future bad economic conditions. Another conclusion is that before the crisis, stock returns were affected by interest rate changes asymmetrically: stock returns reacted to a higher extent to positive interest rate changes (Kontinikas et al. (2013) and Chulia, Martens and van Dijk (2010)). Also, Lobo et al. (2000) and Koutmos (1998, 1999) find that stock prices are to a higher extent affected by unfavorable policy rate changes (i.e. interest rate hikes). Furthermore, monetary inactions have a negative effect on stock returns (Fiordelisi, Gallopo and Ricci (2014)).
Bernanke and Kuttner (2005)) and Chulia, Martens and van Dijk (2010)) also evaluate the effect of policy reversals. Policy reversals are events at which the central bank change the sign of the policy rate change, in comparison with one policy announcement before. Both papers find that when there is an event of reversal, the effect of interest rate changes is larger.
2.4.2 Unconventional tools and stock returns
Another important field of research is the effect of unconventional tools of central banks on asset prices. Haitsma, Unalmis and de Haan (2016), look at the impact of both ECB’s conventional and unconventional tools on the stock market between 1999 and 2015, using the event study suggested by Bernanke and Kuttner (2005). Their results show that especially unexpected interest rate changes affect the EURO STOXX 50 index. Furthermore, they find that expansionary monetary policy affect the Euro Stoxx 50 negatively. They create a proxy for the unconventional monetary policy by using future data about German and Italian government bonds. A decrease in the spread between the German and Italian bond yields, leads to an increase in stock returns. Surprisingly, they also find a negative effect to expected monetary policy. They point out that they do not have a proper explanation and that further research should point out why this relationship would hold. Furthermore, they conclude that there is a lot of heterogeneity in the reaction to interest rate changes between sectors.
Rogers, Scotti and Wright (2014) study the effect of unconventional monetary policy stock prices, bond yields and exchange prices for the FED, BoJ, BoE and the ECB. They conclude
21 that unconventional tools are good to ease broad financial conditions. Their results show that announcements of unconventional policies of the ECB increased stock returns during the crisis (Fratzscher et al. (2016) find identical results). However, Hosono and Isobe (2014) find that European stock markets are negatively affected to ECB unconventional monetary policy. The authors argue that an increase in unconventional policy in a crisis may signal an economic condition that is worse than the market participants realized. Therefore, the returns could decrease when the unconventional policies are increased.
2.4.3 Difference between general stock and REIT stocks
I am going to look at the effect of interest rate changes on stock returns of REITs. But first, it is important to explain why REITs would react differently to interest rate changes, compared to general stocks. When this is not the case, we could use the studies that are already conducted to see what the effect is of an interest rate change on the REIT stock returns. When there is a difference in reaction, it is also important to study solely the effect on REIT stock returns.
Stocks of REITs are different than other common stock (Xu and yang (2011)), because REITs must invest a certain percentage in real estate. Furthermore, REITs are obliged to pay out a certain percentage of their income as dividend to avoid double taxation. Bernanke and Kuttner (2005) argue that due to the high yield status of REITs, changes in the interest rate could have a larger effect on the present value of dividends. The reason for this is that REIT income is in general more influenced by interest rates compared the income of general stocks. The response to monetary policy surprises differ across sectors, because the interest-‐elasticity of demand for the products differ (Peersman and Smets (2005)). The real estate market has a high interest-‐elasticity, which makes it more influential to interest rate changes compared to sectors that have a low interest-‐elasticity. Another thing that they point out, is that the more a firm is dependent on bank funding, the more it will be affected by interest rate changes and thus the higher the interest-‐elasticity (this same argumentation has also been made by Ehrmann and Fratszscher (2004)).
Xu and Yang (2011) point out that the maturity and the relative size of the REIT sector may be another reason why the REITS could react differently, compared to the general stock