Jul 5, 2015 — die nicht mit den Positionen der Deutschen Bundesbank oder ihrer Mitarbeiterinnen und Mitarbeiter übereinstimmen müssen. Herausgeber:.
62 KB – 24 Pages
PAGE – 3 ============
In der Reihe —IWH Onlineﬁ erscheinen aktuelle Manuskripte der IWH -Wissenschaftlerinnen und -Wissenschaftler zeitnah online. Die Bände umfassen Gutachten, Studien, Analysen und Berichterstattungen. Kontakt : Professor Reint E. Gropp, Ph.D. Telefon: + 49 345 7753 700 Fax: + 49 345 77 53 820 E-Mail: reint.gropp @iwh -halle.de Bearbeiter: Geraldine Dany (IWH, Martin -Luther -Universität Halle -Wittenberg) Professor Reint E. Gropp, Ph.D. (Präsident des IWH, Otto -von-Guericke -Universität Magdeburg) Helge Littke (IWH) Dr. Gregor von Schweinitz (IWH, Martin -Luther -Universität Halle -Wittenberg, Deutsche Bundesbank) Die in dieser Studie geäußerten Einschätzungen und Meinungen stellen die persönlichen Ansichten der Autoren dar, die nicht mit den Positionen de r Deutschen Bundesbank oder ihrer Mitarbeiterinnen und Mitarbeiter übereinstimmen müssen. Herausgeber : LEIBNIZ -INSTITUT FÜR WIRTSCHAFTSFORSCHUNG HALLE Œ IWH Geschäftsführender Prof. Reint E. Gropp, Ph.D. Vorstand: Prof. Dr. Oliver Holtemöller Dr. Tankred Schuhmann Hausanschrift: Kleine Märkerstraße 8, D -06108 Halle (Saale) Postanschrift: Postfach 11 03 61, D -06017 Halle (Saale) Telefon: +49 345 7753 60 Telefax: +49 345 7753 820 Internetadresse: www.iwh -halle.de Alle Rechte vorbehalten Zitierhinweis : Leibniz -Institut für Wirtschaftsforschung Halle (IWH) (Hrsg.): German y‚s Benefit from the Greek Crisis . IWH Online 7/2015. Halle (Saale) 2015. ISSN 2195 -7169
PAGE – 8 ============
Germany™s Benefit from the Greek Crisis 2 (and a Grexit more likely), German government bond yields fell and each time an event increased the likelihood of an agreement on package, German government bond yields increased . Cumulatively, bad news for Greece resulted in a decline of German ten -year bu nd yields of more than 1.5%. The effect is symmetric: Good news for Greece resulted in increases in German bund yields of about equal ma gnitude .1 3 Counterfactual yields on German bunds without flight -to-safety The previous section provides convincing evidence that bad (good) news in Greece lowered (increased) German bund yields. However, in order to assess the overall effect on interest costs, one needs to simulate German government bond yields in the absence of a crisis. Hence, we provide two simple ways to calculate counterfactual yields for German bunds for 2010 to 2015. The first (naïve) approach uses the average bond yield between 2000 and 2007 as a benchmark, assuming that all deviations from such a yield c an be attributed to the crisis. The second approach relaxes this assumption and takes into account that some deviations from this normal value may be explained by the general macroeconomic environment in Germany. Hence, we calculate the counterfactual risk -free interest rate using a simple monetary policy rule (Taylor rule) for Germany. Both approac hes yield very similar results. 3.1 Benchmark: German bond yields German bond yields from the introduction of the Euro until 2007 were quite stable ( Figure 1 ). In that figure (as in the following), we use three different maturity bands: short -term bonds with a maturity of up to one year; medium -term bonds with maturities between one and five years , and long -term bonds with maturities of more than five years . Tabl e 1 reports the average yield from 2000 to 2007 and the yields for the subsequent years until 2015. The yields between 2000 and 2007 (differing by maturity) can be interpreted as equilibrium yields for Germany in the absence of a crisis situation . Yields f or all maturities fell to levels close to zero during the Great Financial Crisis and never recovered to their normal level afterwards despite the fact that the German economy fully recovered in 2009 and 2010. In this naïve approach, a ny difference between observed and ﬁnormalﬂ bond yields between 2010 and mid 2015 can therefore be attributed to the European debt crisis (which from 2010 onwards was mostly driven by events in Greece). That is, the Greek crisis created circumstances in which Germany was not only present as a safe haven, but actively sought as such by fleeing investors. Hence, a s our first set of counterfactual interest rates, we therefore use the normal average yields o bserved between 2000 and 2007 on secondary markets. 3.2 A German yield curve derived from a counterfactual monetary policy rule We are interested in the development of the German government bond yields for the counterfactual case that there had been no European debt crisis. Our second approach to this problem involves deriving hypothetical German policy rates by estimating three variants of central bank policy rules in the style of Taylor (1993), which assumes central bank decisions on interest rate are a function of two factors: the deviation of inflation from an inflation target and the deviation of output from potential 1 On the other hand, positive news from Greece increased yields on German benchmark bond yields on average, see Appendix. However, the direct effect of positive news was on average smaller, consistent with event studies on the effects of good and bad news (Afonso, Furceri and Gomes, 2012).
PAGE – 9 ============
Germany™s Benefit from the Greek Crisis 3 output. With the estimated German policy rates we then calculate the counterfactual German bond yields by assuming the slope of th e yield curve to be exogenous. For the policy rule estimation , we use quarterly data from 1980Q1 to 2015Q2. We obtain real output , the consumer pric e index , the FIBOR (one month, average of the month , 1990Q1 -1998Q4 ) and the Euribor (one month, average of the month , 1999Q1 -2015Q2 ) from the German Bundesbank statistics . We construct potential gross domestic product [ GDP ] by applying the Hodrick -Prescott filter to the GDP series. The inflation series is constructed by taking the quarterly average of year on year percentage change of the monthly inflation index. The policy rate is constructed by merging the FIBOR and the Euribor rate. We assume the real interest rate to be two percent as to approximate the long -run growth rate of the German economy during t he estimation period. As in Clarida et al. (1998), the inflation target is assumed to be 1.9 throughout the estimation. We estimate the following policy rule: = +(1)( ++()+()) where is the main measure for the conduct of monetary policy, is the inflation target of the central bank, is the long -run real interest rate , is the inflation rate , is the real output and is the potential output. We include interest rate smoothing into our specification. Thereby, we take into account that central bankers might prefer to change the policy rate in small steps. For the estimation procedure , we follow Cla -rida et al. (1998) and use GMM (Generalized Method of Moments ) estimatio n. We construct three variants of counterfactual German policy rates. The first two are estimations of the above specification for two subsamples 1990Q1 until 1998Q4 and 1990Q1 until 2007Q1. For the third variant , we simulate the rule specified above assum ing the parameters of Taylor (1993) for =1.5 and =0.5. The parameters of the first two estimations are in line with the literature. For the estimation of the subsample covering 1990 to 1998, we have =0.71 (0.084 ),=1.57 (0.2468 ) and =0.93 (0.5072 ) (standard deviations in brackets) . For the second subsample from 1990 to 2007, we find =0.74 (0.084 ),=1.62 (0.2468 ), =0.83 (0.5072 ). Previous to the introduction of the European Monetary Union (EMU) , the German Bundesbank officially followed monetary aggregate targeting, which implies changes to the interest rate when money growth deviates from its target value. Empirical evidence given by Bernanke (1997) showed , however , that at lea st the Bundesbank acted in favo ur of inflation targeting and that the interest rate policy of the Bundesbank can be well approximated by Taylor -type rules. With the introduction of the EMU, the European Central Bank took over interest rate settings for the whole currency un ion, targeting union™s average inflation deviations and average output gaps. In particular, interest rates are meant to be set to maintain actual EMU -wide inflation below but close to two percent and as to stabilize the business cycle. In reaction to the d ouble dip recession , ECB™s monetary policy has dominantly been expansive. This implies that the interest rate set by the ECB has been low as compared to an interest rate that would have been set by a central bank that bases its monetary policy decisions so lely on economic developments in Germany. Thus we find that the counterfactual interest rate in all the scenarios described above is well above the actual policy rate set by the ECB. Figure 2 displays the three counterfactual policy rates of Germany and th e actual realizations of the Euribor from the first quarter in 2000 until the second quarter of 2015. First notice that the dynamic forecasts of both estimated Taylor rules as well as the original Taylor rule variant closely follow the development of the E uribor from 2000 until the first quarter of 2007. Secondly, the differences between the estimate d policy rates and the Euribor rise from the first quarter of 2007 to the first quarter of 2009 as well as from the first quarter of 2010 onwards and start to d ecline slightly after 2012. Especially the counterfactual interest rate obtained from a Taylor rule based on pre -Euro data
PAGE – 10 ============
Germany™s Benefit from the Greek Crisis 4 gives credit to the assertion that the policy of the ECB was actually not much different from the one the Bundesbank would have chosen until 2007. The main reason for this similarity is the large weight of the German economy, combined with a comparably low concern of misalignments in the Euro area (Knedlik and von Schweinitz, 2012). When misalignments started to become apparent during the 2008/2009 f inancial crisis and subsequent sovereign debt crise s, the ECB opted for a more accommodative monetary policy than would have been appropriate for Germany alone . Figure 3 depicts the differences between each of the estimated policy rates and the Euribor, respectively. The differences peak on the heights of the financial and the sovereign debt crisis. Differences a re small until about 2007 (around 50 basis points, BP) and rise particularly during the European debt crisis. We use these estimated policy rates for Germany to calculate the counterfactual bond yields for the long -, medium – and short -run maturity by subtr acting the Euribor from the yields series and adding our estimated policy rates. The counterfactual and the actual yields are depict ed in Figure s 4 to 6. The counterfactuals indicate that in the absence of crisis , German government bonds yields would have been substantially higher. 4 The gains from the safe -haven effect We use information on actual bond auctions by the German government in order to calculate the overall gain to the budget. Table 2 presents the stru cture of public sector debt issuances from 2007 to today. T he variation in bond issuances of the German general government over time reflects the high government deficit during the Great Financial Crisis, and the slow consolidation of budgets from 2011 onwards. The auctioned amounts are obtained from the reports of the Federal Financing Agency on all auctions of newly issued bonds (including increased principal on outstanding bonds ) of the central government. However, the auctions from the Federal Financing Agency are not the only source of debt funding of the general German government. They do not include (a) debt by states or municipalities or (b) alternative debt financing sources like direct credit from banks. Therefore, auctions for different maturity b ands are only between 45% and 75% of total gross borrowing by the German state. Hence , the estimates on interest saving of the German government are necessarily only a lower bound on total gains from the Greek debt crisis. Germany issued betw een 297 (2007) and 676 billion E uro s (2010) each year. This is the relevant amount for our calculations, because interest savings will only accrue on newly issued debt, not on outstanding debt. Frequent rollovers allow Germany to ﬁcash inﬂ on reduced government bond yie lds. With this information in hand , we can proceed to calculate the annual interest savings that accrued to the German budget from the crisis. The difference between observed and the counterfactual interest rates provided in the previous section gives, for every point in time and every maturity group, a yield spread. If these yield spreads (observed minus counterfactual) are multiplied with newly issued debt, we obtain gains from the favorable interest environment. However, these gains do not only materialize in the year of issuance, but in all subsequent years until maturity (assuming normal coupon bonds). Assume a bond of size one billion with a maturity of ten years that is auctioned off in 2011 for an interest rate which is 4% below its counterfactual due to uncertainty on financial markets. Then the German state saves 4% of one billion (40 millions) in interest payments every year until 2020. For the
PAGE – 11 ============
Germany™s Benefit from the Greek Crisis 5 purposes of this analysis , we have limited ourselves to the gains that already accrued, and do not inclu de any future gains. 2 In the following, we assume that all German government bonds pay interest every year .3 That is, we use the maturity of bon ds in order to distribute interest gains for bonds issued between 2010 and today over the years following the issuance until 2015. Adding gains originating and materializing between 2010 and 2015 over different maturi ty bands allows us to get a feeling of total savings from bond auctions over this period, as shown in Figure 7. Using this conservative approach, we find savings in the ballpark of 100 billion Euro s, irrespective on how we specify the counterfactual. This should be viewed as a lower bound of the benefits accruing to the German government from the debt crisis . These gains are larger than the total Greek debt owed to Germany , (estimated by most accounts at 90 billion Euro s, including exposure from a still to be negotiated program) . That is, even in event that Greece defaulted on all its debt, the German central government alone would have benefit ed from the Greek crisis. The gains from other credit financing of the ge neral government (another 25% to 55% of total newly issued debt) are not even accounted for in this context. The gains from different counterfactual scenarios are remarkably similar, ranging from 93 billion for the normal yields scenario to 126 bi llion fro m the pre -Euro Taylor -rule scenario. The fact that very different assumptions yield quite similar results provid es a high degree of robustness. Concerning the future, we expect Germany to continue profiting from the current situation. If the situation calmed down suddenly, Germany would no longer be able to issue debt at depressed rates . However, sizeable amounts of medium – and long -term bonds issued in the past years are still far away from maturity, extending the period of German profits for some tim e to come. 2 Even in case interest rates r eturn immediately to their long -term average, there are still substantial future gain s from the long -term bonds is sued under the low interest environment. Rough estimates would suggest that these gains could as much as double the estimates on interest savings given below. 3 This assumption is somewhat conservative, since zero coupon bonds (where gains materialize in the first year in total) constitute around 1/3 of all newly issued bonds.
62 KB – 24 Pages