The impact of interest rate fluctuations on
stock returns of financial companies traded
on the AEX and AMX
An empirical research on the effect of interest rate fluctuations on stock returns of the
biggest financial companies publicly traded in the Netherlands.
Author: J.A. Scheper
Student number: 10351949
Thesis supervisor: D.W. Van Dijk
Finish date: February 2016
I.Abstract
This thesis describes the effect of interest rate fluctuations on the stock returns of companies traded on the AEX and AMX. The main focus is on the return of the 5 biggest financial companies traded in the Netherlands. The thesis consists of background information of theories on the effect of interest rate on stock returns and includes earlier empirical research in this area. A panel data regression shows that over the entire period the effect of the interest rate is not significant. However, in the period before the crisis, an increase in the interest rate has a negative effect on the financial company’s stock return, as expected in the hypothesis.
Keywords:
Interest rate, financial companies, stock return, AEX. JEL-classification: G12
Verklaring eigen werk
Hierbij verklaar ik, Joel Scheper, dat ik deze scriptie zelf geschreven heb en dat ik de volledige verantwoordelijkheid op me neem voor de inhoud ervan.
Ik bevestig dat de tekst en het werk dat in deze scriptie gepresenteerd wordt origineel is en dat ik geen gebruik heb gemaakt van andere bronnen dan die welke in de tekst en in de referenties worden genoemd.
II. Table of Contents
I. Abstract 2
II. Table of Contents 3
1. Introduction 4 2. Literature Review 6 3. Data 10 4. Model 12 5. Results 14 6. Robustness Checks 18 7. Conclusion 21 8. Reference List 23 9. Appendix 25
1. Introduction
The last few years fluctuations of the risk-free interest rate were massive, with rates even going close to 0% in the last year. These fluctuations were, among others, caused by the financial crisis of 2008 and the following macro-economic adjustments set by the Dutch government. During the crisis investors also lost their trust in the stock market. One important measure for investors to check whether to buy or not to buy stocks from a company is the return on stock. A high return on stock could persuade investors to buy the stock, as it indicates high profitability and a strong financial position. However, higher return could also indicate higher risk.
Ways in which interest rate fluctuations can influence return on stock are among others explained by the nominal contracting hypothesis and the idea of maturity mismatch. Another reason is that stocks can be seen as a simple call options, which are also influenced by changes in the interest rate.
The impact of these fluctuations of the interest rate on the stock return have been subject to empirical research. However, the results of these researches differ and the research has not been done before with data from the Netherlands. Fama & Schwert (1977) find that the fluctuations affect the stock returns, arguing that expected and unexpected interest rate fluctuations and stock return are negatively related , while Chance & Lane (1980) find that both financial and non-financial stock returns are not affected by anticipated interest rate changes. The research question is as follows: What is the effect of interest rate fluctuations on the stock returns of the financial companies traded on the AEX and the AMX?
The financial companies traded at the AEX and the AMX have been chosen for an empirical study as it has not been done before and it narrows the research question to a more feasible one. This study has scientific value as it adds another country to the list of countries where this has been done. It is for the first since a long time that the Netherlands have witnessed a financial crisis, so it is important to analyze the market as much as possible to give investors and researchers a better understanding of the Dutch financial market. This thesis will contribute to existing literature in two ways: To start, adding a first in-depth analysis of the effect of interest rate fluctuations on the stock returns of the financial companies traded on the AEX and the AMX. The time period used covers a large dataset, including several developments on the economic and financial markets which were not seen for a long time, or even unique, in the Netherlands. Second it uses two econometric approaches which have not
been combined before to research a topic like this, namely a panel data OLS-regression and a time series OLS-regression. The panel data analysis will enhance the analysis by filtering out company fixed effects.
Financial companies are examined as they have the most to do with interest. As a financial company’s assets and liabilities are sensitive to interest rates, equilibrium returns for those companies are based on certain expectations about interest rates. Those equilibrium returns in their turn affect the stock returns of those financial companies. The AEX consists of the 25 most traded companies of the NYSE Euronext Amsterdam. As there are only three financial companies traded on the AEX, namely Aegon, Delta Lloyd Group and ING Group1, the research was expanded with the financial companies traded on the AMX. The AMX consists of the next 25 most traded companies of the NYSE Euronext Amsterdam. The financial companies on the AMX are as follows: Binckbank and for most of the period analyzed SNS Reaal. This will bring the total of financial companies analyzed to a number of five.
The research can be seen as a case study to earlier research of Fama (1977) and as a follow up to the work of Dinenis (1998) and Kasman (2011), who did researches likewise in the United Kingdom and Turkey. A combination of their methods will be followed to examine the relationship between interest rate fluctuations and stock returns of financial companies.
1
2. Literature review
Flannery and James (1984) say that to examine the relationship of interest rate changes and stock returns, it is helpful to approach the nominal return of a company’s stock as consisting of two parts: the return on nominal assets and the return on real or physical assets (liabilities, leaving out equity, can be accounted for by making them negative assets). Nominal assets are assets with cash flows that are fixed in nominal terms, like most contracts, accounts receivable and debt. Physical assets cash flow however, fluctuate corresponding to changes of the price level. The nominal contracting hypothesis states that a company’s holdings of nominal assets has a lot of influence in explaining the behavior of stock returns. This is done through the redistributive effects of unanticipated change of the interest rate and unanticipated changes in the expected interest rate (French, Ruback and Schwert, 1983).
Since unanticipated changes of interest rate, ceteris paribus, affect the real value of nominal but not physical assets, stockholders of companies with fewer nominal assets than nominal liabilities should benefit from unexpected rise of the interest rate. When nominal liabilities are less than nominal assets, the effect is vice versa: stockholders of these companies are harmed by a unexpected rise of the interest rate (Grove, 1974).
Unexpected changes in expected interest rate will affect the nominal value of nominal contracts. The change in the interest rate causes the nominal interest rate used to discount the cash flows to differ. The effect of unexpected changes in expected interest rate on the value of nominal assets and liabilities will depend on the maturity of the nominal contract. The effect on the company’s value of equity will be greater the longer the maturity of the company’s nominal assets relative to its nominal liabilities (Grove, 1974). According to Flanery and James (1990) the nominal contracting hypothesis implies a relationship between company’s stock return and interest rate changes: the higher the proportion of the net nominal assets and the longer the maturity of net nominal assets held, the more sensitive should be the company's stock returns to interest rate changes. For example, with an interest rates change, the stock returns of an all-equity financed company with only nominal assets should behave identical to a bond with a maturity equal to the average maturity of the firm's assets.
There are a few other factors suggested why interest rates should have a direct effect on stock returns. As Stone (1974) suggested, with risky debt instruments, the market portfolio should not solely be approximated by an equity index, but by a combination of risky equity and debt return. A second factor is based on the idea of maturity mismatch (Flannery and James, 1984).
Financial company’s assets and liabilities are affected by interest rates and returns for those companies are based on certain expectations about interest rates. When interest rates change unexpectedly, and assets and liabilities are not matched by maturity or more general by duration, equity returns will be affected (French, Ruback and Schwert, 1983). As a third factor, stocks can be viewed as a call option on the assets of the firm (Galai and Masulis, 1976) and consequently their stock returns should be affected by changes in interest rates, just as with a normal call option. The stocks being viewed as a call option can be explained in the following way: The owner of a call option has claim to a part of a stock’s price distribution to the right of the exercise price at maturity date T. In the same way, company’s stockholders have claim to the part of the company’s price distribution to the right of the face value of the company’s debt at its maturity date (Galai and Masulis, 1976).
Financial companies are more sensitive for interest rate changes than non-financial companies as they own more assets and liabilities that are affected by interest rates than non-financial companies. Another reason why financial companies are more sensitive for interest rate changes than non-financial companies is mentioned by Chance and Lane in 1980. They say that financial stocks are subject to such comparatively modest changes in investor expectations that the interest rate factor is influential almost all of the time. This is not the case with non-financial companies stock.
Understanding the channels through which interest rates affect portfolios of stocks is important for performance measurement, asset pricing, portfolio selection and risk management. Hedging a portfolio of stocks by selling stock index futures, example given, will only protect the portfolio from changes in interest rate that are reflected in the market index but will let the portfolio exposed to interest rate risk that operates directly on the financial company’s stock return.
Another factor that is important during this research is the financial crisis. The financial crisis could have an effect on the stock returns of both financial and non-financial companies. According to Hartmann (2004), stocks act way more correlated during a crisis. Also the interest rate became extremely low during the crisis, which could influence the effect we are interested in.
Previous studies on the effect of interest rates on financial company’s stock return obtain contrasting results. Chance and Lane (1980) say that their evidence strongly suggests that neither security analysts nor portfolio managers should be worried with the effect of interest
rates on the returns of stock of financial institutions. According to them any association is apparently captured by the relationships between those stock and market-wide factors. Other studies suggest that there is a relationship between interest rate changes and returns of stock of financial institutions. Among those is the research of Lynge and Zumwalt (1980) who find a negative effect. However, Saunders and Yourougou (1990) find on average a weak significant positive relation, but only for a sub period of their sample. Flannery and James (1984) and Booth and Officer (1985) also find a negative relationship. Stock returns for non-financial companies, on the other hand, are either not affected by interest rate changes (Chance and Lane, 1980), or exhibit lower sensitivity as found by Lynge and Zumwalt (1980).
The difference between these results might come from different treatments of interest rate changes. Chance and Lane (1980) and Lynge and Zumwalt (1980) use current interest rate which we will also do in this thesis. Flannery and James (1984) however, estimate and use unanticipated interest changes, just like Saunders and Yourougou (1990) do. Booth and Officer (1985) use both current and unanticipated interest changes, but they use a different expectation process than Flannery and James. The choice of the interest rate variable (short, intermediate or long term) does not appear to affect the significance of the results and the interest change rate effect is bigger throughout periods of high interest rate volatility (Yourougou, 1990).
Empirical research in this field has been done by Dinenis (1998) in the United Kingdom and more recently by Kasman (2011) in Turkey. Dinenis (1998) did an empirical research on the financial companies in the United Kingdom, while at the same time using 95, randomly selected, industrial and commercial companies as a control sample. Dinenis (1998) used a simple two-factor model, with weekly return on the market portfolio and the weekly rate of a default free debt index as proxy of interest rates as the two regressors to estimate the effect of interest rate changes. To avoid collinearity auxiliary a sub equation was used. Collinearity would affect the standard errors.
The results found by Dinenis (1998) when using current interest rates are as following: the sensitivity of individual financial groups was statistically different from that of the commercial companies. There was a negative relationship between an increase in the interest rate and the return on financial company’s stock. In total, the use of a longer term interest rate has increased the interest rate sensitivity of the portfolios. This result was also found by Bae
(1990) who concluded that the impact of interest rate sensitivity on stock returns was greater the longer the maturity of the interest rate.
The findings of Dinenis (1998) with respect to unanticipated interest changes are among others: first, unanticipated changes in interest rates, irrespective of the interest rate variable used, have a statistically significant negative effect on financial company’s returns. Second, the effect of unanticipated changes in interest rates on non-financial institutions is also statistically significant, but substantially lower than the corresponding effect on financial firms.
Kasman (2011) uses a different approach for his empirical research in Turkey. Kasmans (2011) research is on the effect of interest and exchange rate changes on stock return of commercial banks listed on the Istanbul Stock Exchange. Other financial companies are not analyzed. A three-factor model is used for the OLS-regression and a GARCH-model was added. The GARCH-model is used to capture time-varying risk properties in the data.
The results found in this research by the OLS-regression are not efficiently estimated because of the presence of residual autocorrelation. The results found by the GARCH-model indicate that interest rate and exchange rate changes have a negative and significant impact on the conditional bank stock return. Another result is that bank stock return sensitivities are found to be stronger for market return than interest rates and exchange rates, implying that market return plays an important role in determining the dynamics of conditional return of bank stocks. The results further suggest that interest rate and exchange rate volatility are found to be a major determinant in the conditional bank stock return volatility.
3. Data
For the interest rate we use the rate given by the Dutch government on a 10 year government bond as it comes closest to the local risk-free interest rate of the AEX and the AMX. The interest rate given by a 10 year government bond changed from 3.62% on 2 January 2005 to 0.68% on 1 January 2015, a decrease of 2.94%-point or 81.2%.
The idea is to examine the relationship between the fluctuations and stock returns by using a panel data regression (OLS). Also a time-series OLS regression will be used as a check for robustness. As Kasman (2011) warned us for residual autocorrelation we will also estimate a GARCH-model. Data for these regressions are derived from stock prices from Datastream, and processed using Stata. Data will come from the period from 1 January 2005 until 1 January 2015, so we have a 10-year period. The returns used will be weekly observations, bringing the total of observations to 521 observations per company.
Furthermore, the total period will be divided in a pre-crisis period and the rest of the period to check if there is a difference in different economic times. The crisis will start at 15 September 2008, the day of the fall of Lehman Brothers. Adjusted closing prices will be used to calculate returns. Adjusted closing prices are used as they take in account among others dividend payments.
In addition the 23 non-financial companies traded on the AEX are used as a control sample subject .
The sample period varies per company, as some entered the AEX/AMX later than 2005 and some left earlier than 2015. Delta Lloyd Group entered the AEX in 2009 as it went public that year. SNS Reaal entered the AMX in 2006 as it went public that year. SNS Reaal left the AMX in 2013 as it was brought under state control. SNS was brought under state control due to poor financial results. This could affect the quality of the estimated results and might be a reason to leave SNS out as a robustness check. TNT entered the AMX in 2011 as it went public that year. Tom Tom entered the AEX in June 2005 as it went public.
Table 3.1 Sample Periods used per bank and market. (FC) denotes a financial company.
Company Period
Aegon(FC) 03-01-2005/29-12-2014
Binckbank(FC) 03-01-2005/29-12-2014
Delta Lloyd Group(FC) 09-11-2009/29-12-2014
ING Group(FC) 03-01-2005/29-12-2014
SNS Reaal(FC) 29-05-2006/04-02-2013
TNT 06-06-2011/29-12-2014
Tom Tom 06-06-2005/29-12-2014
All Other non-financial AEX companies 03-01-2005/29-12-2014
Market Period
AEX 03-01-2005/29-12-2014
Interest rate per week is calculated by the formula shown below.2
The weekly returns of the financial companies and AEX-index, and the yearly return of a 10-year Dutch government bond can be seen in tables 10.1-10.7. In the case of Aegon, ING, SNS Reaal and the AEX index we see increased volatility during the period of the economic crisis (2008-2010). SNS Reaal, Delta Lloyd group and the AEX index show increased volatility around 2012. Binckbank does not have the increased volatility during the period of the economic crisis or in 2012, but has increased volatility at the start of the period observed (2005-2006). The annual return on a 10 year Dutch government bond increases during the crisis and around 2011, but eventually ends in 2015 at 0.68%, a decrease of 2.94%-point or 81.2% compared to the starting point in 2005 which was a 3.62%
The summary of the panel data can be seen in Table 9.8.
2
𝑟 =((1+YTM)^(1/n))-1
Where YTM= Yield to Maturity, n=number of weeks in observati
Bond Period
4. Model
Panel data analysis is used as it allows us to control for variables that cannot be observed or measured, like company culture. Panel data analysis basically accounts for individual heterogeneity when using company fixed effects.
Panel data OLS-regression equation;
Returnit = 𝛽0 +𝛽1MRKt +𝛽2INTt +𝛽3FCi +𝛽4FCi * INTt + 𝜀t
Where Returnit is the return of the stock and the dependent variable, INTt is the interest rate and the main explanatory variable, MRKt is the return of the AEX and is a control variable, FC is a dummy variable with 1 indicating a financial company, FCi * INTt is an interaction variable between the dummy variable and the main explanatory variable and 𝜀t is the error term.
Hypotheses;
H0: 𝛽2=0 meaning interest changes do not have an effect on the stock return of companies traded on the AEX and the AMX. 𝛽4=0 meaning interest changes do not have an extra effect on the stock return of financial companies traded on the AEX and AMX.
H1: H0 not true meaning interest changes do have an effect on the stock return of companies traded on the AEX and the AMX and/or interest changes do have an extra effect on the stock return of financial companies compared to the effect on stock return of all companies traded on the AEX and AMX.
A rejection of the H0 will answer the research question as it proves interest rate has an effect on the stock return of financial companies. A negative 𝛽2 and/or 𝛽4 implies a negative relation between an increase of the interest rate and the stock returns of the (financial) companies, whereas a positive 𝛽2 and/or 𝛽4 implies a positive relation between an increase of the interest rate and the stock returns of the (financial) companies. The expectation is that a negative relationship is found, just as Kasman (2011) found in Turkey and Dinenis (1998) found in the UK.
If H0 is not rejected, it will answer the research question as it proves there is no significant relation between an increase of interest rate and stock returns of financial companies.
In the literature review the following factors were named that could affect the return on stock: The perceived risk has a positive relationship with the return on stock, as with higher risk investors want higher possible returns for their exposure. Also, the higher the proportion of the net nominal assets and the longer the maturity of net nominal assets held, the higher the effect of a change in interest rate. A higher proportion means more exposure to changes in interest rate, and this relationship is negative. The last channel mentioned was that the stocks can be seen as a call option. Higher interest implies a lower return on the call option, meaning that this relationship is also negative.
5. Results
Doing a panel data regression on the entire period without fixed effect gives the results as seen in Table 5.1 (1). As you can see, the effect of the AEX return is significant and positive on the return of companies. It is close to 1, which was also expected as the AEX return reflects economy-wide factors. The coefficient means that the average beta of the selected companies is 0.89. The effect of the interest rate on the return of companies is also positive, but is not significant. Being a financial company also affects the return positively, however this result is also not significant. The same applies to the interaction variable, it is not significant. It is the only one to imply a negative relationship with the return of financial companies. In Table 5.2 (2) the results of a panel data regression with company fixed effects are shown. In comparison to Table 5.2 (1) the results differ slightly, but the results are still nog significant except for the effect of the AEX return.
Table 5.1 Panel data regression results.
(1) (2)
Return Coeff. Std. Err. Coeff. Std. Err.
MRK 0.89347***3 0.01346 0.89332*** 0.01346 INT 0.39346 0.24077 0.36476 0.24355 FC 0.00170 0.00350 0.00055 0.0045 FC * INT -0.44553 0.58611 -0.16248 0.60958 Constant -0.00133 0.00145 Adj. R2 0.26780 0.26750 N 12066 12066
Fixed Effects No Company
Through the entire thesis the abbreviations mean the following: MRK=Market return on AEX, INT=Interest rate, FC=Financial company dummy, FC *
INT=Interaction variable of FC and INT, Adj. R2=Adjusted R-squared, N=number of observations, Coeff.=Coefficient and Std. Err.=Standard Error.
3 Through the entire thesis:
*** =significant at the 1% level ** = significant at the 5% level * =significant at the 10% level
Table 5.2 (1) shows the panel data regression without fixed effects before the crisis. The effect of the market return, the beta, is now 0.97. The effect on returns of a being a financial company is now positive and significant. This means that a 1%-point rise of the interest in a week increases the return of financial company’s stock with 3.33% more than the return of a non-financial company. The interaction variable is now negative as expected in the hypothesis and significant. This means that a financial company’s stock will see a decrease in return of 481% (the coefficient of INT and FC * INT combined) if the interest rate rises 1%-point in a week. This results appears to be extreme, but a 1%-point drop of the interest rate in a week is also very extreme and unlikely. This shows that the interest rate has a negative relationship with return of financial companies before the crisis. Table 5.2 (2) shows the panel data regression with company fixed effects before the crisis. The values of the coefficients stay close to those of the previous regression, with a small drop of the effect of the interaction variable. The coefficient of the Financial Company dummy variable is lower now and marginally significant at the 10% level.
Table 5.2 Panel data regression before the crisis.
(1) (2)
Return Coeff. Std. Err. Coeff. Std. Err.
MRK 0.96991*** 0.02543 0.96995*** 0.02544 INT -0.29136 0.78618 -0.27979 0.78662 FC 0.03330** 0.01519 0.02967* 0.01587 FC * INT -4.51416** 2.02884 -4.25663** 2.01855 Constant 0.00351 0.00583 Adj. R2 0.2643 0.2638 N 4165 4165
Fixed Effects No Company
The results during and after the crisis can be seen in table 5.3 (1). The only variable which has a significant effect on the return is the AEX return. During and after the crisis the effect of the interaction variable is not significant anymore. It could be that during a crisis the effect of the interest rate is undermined by more important factors. A few possible explanations are found in literature: according to Hartmann (2004), stocks act more correlated during a crisis. Another possible explanation could be that the interest rates became so low during the crisis, that the effect disappeared. Campello (2010) says that financial companies can become credit-restrained. Being credit-restrained affects the financial companies stock return as the
companies were forced to burn a sizeable portion of their cash savings during the crisis and to cut more deeply their planned dividend distributions. Table 5.3 (2) shows the panel data regression with company fixed effects after 2008. Not much has changed in comparison to Table 5.7, except for the fact that the effect of the financial company dummy is now significant at the 10% level. This means that being a financial company decreases your return by 11% if the interest rises with 1%-point in a week in the period during and after the crisis.
Table 5.3 Panel data regression during and after the crisis.
(1) (2)
Return Coeff. Std. Err. Coeff. Std. Err.
MRK 0.87218*** 0.01615 0.87218*** 0.01653 INT 0.57447 0.35921 0.54497 0.36190 FC 0.00159 0.00443 -0.01082* 0.00640 FC * INT -0.5392 0.8627 -0.07321 0.88158 Constant -0.0021 0.00183 Adj. R2 0.27 0.2696 N 7901 7901
Fixed Effects No Company
6.
Robustness checks
A time series OLS-regression is used as a robustness check.
Time-series OLS-Regression equation: Returnt = 𝛽0 +𝛽1MRKt +𝛽2INTt + 𝜀t
Where Returnt is the return of a company and the dependent variable, INTt is the interest rate and the main explanatory variable, MRKt is the AEX return and a control variable and 𝜀t is the error term.
Hypotheses;
H0: 𝛽2=0 meaning interest changes do not have an effect on the stock return of financial companies traded on the AEX and the AMX.
H1: 𝛽2≠0 meaning interest changes do have an effect on the stock return of financial companies traded on the AEX and the AMX.
In Table 6.1-6.5 the time series regressions for the individual financial companies are shown. In all cases, the effect of interest rate fluctuations on the return of the financial company is not significant. Another thing seen in all cases is that the effect of AEX return on the return of the financial company is positive and around 1. A possible reason for the insignificant results is mentioned by Kasman (2011). He says that a normal time series regression will give autocorrelation in the residuals. A possible solution proposed by Kasman (2011) is the estimation of a GARCH-model.
Table 6.1 Regression results time series of Aegon.
Return Coeff. Std. Err.
MRK 1.32952*** 0.08147
INT -0.44196 1.33527
Constant 0.00292 0.00808 Adj. R2 0.3516
Table 6.2 Regression results time series of BinckBank.
Return Coeff. Std. Err.
MRK 0.82823*** 0.07947
INT 182.070 1.30241
Constant -0.00671 0.00789
Adj. R2 0.1791
N 492
Table 6.3 Regression results time series of ING Group.
Return Coeff. Std. Err.
MRK 1.49833*** 0.08236
INT -0.50254 1.34973
Constant 0.00293 0.00817
Adj. R2 0.4030
N 492
Table 6.4 Regression results time series of Delta Lloyd Group.
Return Coeff. Std. Err.
MRK 0.94944*** 0.08662
INT -0.53721 1.61754
Constant 0.00304 0.00737
Adj. R2 0.3167
N 257
Table 6.5 Regression results time series of SNS Reaal.
Return Coeff. Std. Err.
MRK 1.27825*** 0.09577
INT 2.63023 0.19700
Constant -0.02370 0.01333
Adj. R2 0.3470
N 333
A GARCH-model is estimated to get rid of the residual autocorrelation. The GARCH-model is used to capture time-varying risk properties. The results of the GARCH-estimations per company are shown below. All the coefficients are still not significant, except for the effect of the return of the AEX.
Table 6.6 GARCH model estimation Aegon.
Return Coeff. Std. Err. MRK 1.32952*** 0.05211 INT -0.44196 1.89013 Constant 0.00292 0.00809 Wald-Chi2 660.61 N 492
Table 6.7 GARCH model estimation Binckbank.
Return Coeff. Std. Err. MRK 0.82823*** 0.05782 INT 182.070 1.54012 Constant -0.00671 0.01000 Wald-Chi2 201.47 N 492
Table 6.8 GARCH model estimation ING Group.
Return Coeff. Std. Err. MRK 1.49833*** 0.05128 INT -0.50254 1.81105 Constant 0.00293 0.01186 Wald-Chi2 881.14 N 492
Table 6.8 GARCH model estimation Delta Lloyd Group.
Return Coeff. Std. Err. MRK 0.94944*** 0.07402 INT -0.53721 1.80939 Constant 0.00304 0.00828 Wald-Chi2 176.81 N 257
Table 6.8 GARCH model estimation SNS Reaal.
Return Coeff. Std. Err. MRK 1.27825*** 0.07379 INT 2.63023 2.03066 Constant -0.02370 0.12728 Wald-Chi2 301.57 N 333
7. Conclusion
A research to the effect of interest rate fluctuations on the stock return of financial companies has been of extra importance last years as a consequence of the crisis and extremely low interest rates. Therefore, this thesis examines the previous named effect with a time-series OLS-regression and a panel data regression, a combination which wasn’t used before. Due to residual autocorrelation with the time-series OLS-regression, a GARCH-model was also estimated. The markets analyzed are the Dutch AEX and the financial companies of the AMX.
Ways in which the interest rate could affect the stock return were among others explained by the nominal contracting hypothesis and the idea of maturity mismatch. Also the perceived risk plays a part and stocks can be seen as normal call options, which are also affected by changes in interest rate.
The results of this thesis indicate that there is a negative and significant relationship between interest rate and stock return of financial companies traded on the AEX and AMX. However, this effect is only significant in the period before the crisis. The starting point of the crisis is taken as the fall of Lehmann Brothers on 15 September 2008. In the period during and after the crisis the effect is not significant. One possible explanation is that companies act more correlated during a crisis. Another possible explanation could be that the interest rates became so low during the crisis, that the effect disappeared. A third possible explanation is that being capital-restrained could affect the stock returns. The effect of interest rate on the stock return of non-financial companies is not significant.
Other significant results found are at first: being a financial company positively influences the return before the crisis. Contrary, during and after the crisis being a financial company influences the return negatively. Second: In almost all cases the return of the stock is positively correlated with the return on AEX.
The results of the time-series OLS-regression aren’t significant due to residual autocorrelation. The GARCH-model is used to filter the time-varying volatility, but the results of the GARCH-model are still not significant.
To fully comprehend and explain the results it must be said that there are some limitations to this research. We only use the current interest rate and do not estimate and use the unanticipated interest rate. This could affect the estimates. Furthermore, time fixed effect are
not filtered while doing the panel data regression as they correlate to much with the other variables.
The findings of this paper provide important information for investors, particularly on the Dutch market, for bank managers involving in risk management and for policy makers in setting monetary policies. The results suggest that investors should not only follow the financial markets, but should also follow the monetary policies when making an investment. Changes in the interest rate should lead to rearrangement of the portfolio’s hold by investors.
Overall, further research is needed to fully understand why the effect of the interest rate on the stock return of the financial companies is significant before the crisis, but isn’t significant anymore during and after the crisis.
8.
Reference List
Bae, S.C., 1990, Interest rate changes and common stock returns of financial institutions: Revisited, Journal of Financial Research, vol.9, no.1 pp. 71-79.
Booth, J.R. and D.T. Officer, 1985, Expectations, interest rates, and commercial bank stock, Journal of Financial Research, Volume 8, Issue 1, pp. 51–58.
Campello, M., Graham, J.R. and Harvey, C.R., 2010, Journal of Financial Economics, Volume 97, Issue 3, pp. 470–487.
Chance, D.M. and W.L. Lane, 1980, A re-examination in the common stock of financial institutions, Journal of Financial Research, Vol. 3 Issue 1, pp. 49-56
Dinenis, E. and S.K. Staikouras, 1998, Interest rate changes and common stock returns of financial institutions: evidence from the UK, The European Journal of Finance, Volume 4, Issue 2, pp 113-127.
Fama, E.F. and W.G. Schwert , 1977, Asset Returns and Inflation, Journal of Financial Economics, pp. 115-146.
Flannery, M.J and C.M. James, 1984, The Effect of Interest Rate Changes on the Common Stock Returns of Financial Institutions, The Journal of Finance, Vol. 39, No. 4, pp. 1141-1153.
Fogler, H.R., K. John and J.Tipton, 1981, Three Factors, Interest Rate Differentials and Stock Groups, The Journal of Finance, Vol. 36, No. 2, pp. 323-335.
French, K.R., R.S. Ruback and G.W. Schwert, 1983, Effects of Nominal Contracting on Stock Returns, Journal of Political Economy, Vol. 91, No. 1, pp. 70-96.
Grove, M.A., 1974, On "Duration" and the Optimal Maturity Structure of the Balance Sheet, The Bell Journal of Economics and Management Science, Vol. 5, No. 2, pp. 696-709.
Hartmann, P., S. Straetmans and C. G. de Vries, 2004, Asset Market Linkages in Crisis Periods, The Review of Economics and Statistics, Vol. 86, No. 1, pp. 313-326.
Kasman, S., G. Vardar and G.Tunc, 2011, The impact of interest rate and exchange rate volatility on banks' stock returns and volatility: Evidence from Turkey, Economic Modelling, vol. 28, no. 3, pp. 1328-1334.
Lynge, M.J. and J.K. Zumwalt, 1980, An empirical study of the interest rate sensitivity of commercial bank returns: A multi-index approach, The Journal of Financial and Quantitative Analysis, vol. 15, No. 3, pp. 731-742.
Saunders, A. and Yourougou, P.,1990, Are banks special? The seperation of banking from commerce and interest rate risk, Journal of Economics and Business, 42, 171–182.
Stone, B.K., 1974, Systematic interest rate risk in a two-index model of returns, Journal of Financial and Quantitative Analysis, vol. 9, No. 5, pp.709–721.
Yourougou, P., 1990, Interest-rate risk and the pricing of depository financial intermediary common stock: Empirical evidence, Journal of Banking & Finance, Volume 14, Issue 4, pp. 803-820.
9. Appendix
Table 9.1 Weekly return Aegon in % .
Table 9.2 Weekly return Binckbank in %.
-30 -20 -10 0 10 20 30 40 50 -30 -20 -10 0 10 20 30 40 50
Table 9.3 Weekly return ING Group in %.
Table 9.4 Weekly return Delta Lloyd Group in %.
-50 -40 -30 -20 -10 0 10 20 30 40 50 60 -25 -20 -15 -10 -5 0 5 10 15
Table 9.5 Weekly return SNS Reaal in %.
Table 9.6 Annual return on 10year Dutch government bond in 1000*%.
-40 -30 -20 -10 0 10 20 30 40 50 0 1 2 3 4 5 6
Table 9.7 Weekly return on AEX-Index in %.
Table 9.8 Summary of the panel data.
Variable Obs Mean Std. Dev Min Max
Return 12768 0.00167 0.05207 -0.47258 1.04082 MRKt 12792 0.00059 0.03058 -0.16811 0.14545 Interest 13572 0.00579 0.00185 0.00130 0.00905 -20 -15 -10 -5 0 5 10 15 20
