FRIDAY 22 NOVEMBER

Economic Determinants of the Term Structure of Nominal Interests Rates.
Martin Eichenbaum (North Western University)
Charles Evans (Federal Reserve Bank of Chicago)
David Marshall (Federal Reserve Bank of Chicago)

Discussants: Enrique Santana (CEMFI, Madrid)
Etsuro Shioji (Universitat Pompeu Fabra, Barcelona)

Dissecting the Predictive Content of the Yield Curve for Economic Activity: The Case of the United States and Germany.
Frank Smets (Bank for International Settlements and CEPR)
Kostas Tsatsaronis (Bank for International Settlements)

Discussants: Fernando Restoy (European Monetari Institute)
Michael Wickens (University of York and CEPR)

Long Term Debt and the Political Support for a Monetary Union.
Harald Uhlig (CentER, Tilburg University and CEPR).

Discussants: Ramon Caminal (Institut d’Anàlisi Econòmica/CSIC)
Gilles Saint-Paul (DELTA, Paris and CEPR)

SATURDAY 23 NOVEMBER

Asset Price Fluctuations in a Lifecycle Economy: Do Collateral Constraints Matter?
François Ortalo-Magné (London School of Economics)

Discussants: Anton Braun (CEMFI, Madrid)
Vicenzo Quadrini (Universitat Pompeu Fabra, Barcelona)

Borrowing Constraints and the Financing of Household Capital in Economies with Banking.
Javier Díaz-Giménez (Universidad Carlos III de Madrid)
Luis A. Puch (Universidad Carlos III de Madrid)

Discussants: Xavier Freixas (Universitat Pompeu Fabra, Barcelona)
Albert Marcet (Universitat Pompeu Fabra, Barcelona)

Growth Enhancing Bubbles.
Jacques Olivier (Hautes Études Commerciales, Jouy-en-Josas)

Discussants: Graziella Bertocchi (Universitá di Modena)
Angel de la Fuente (Institut d’Anàlisi Econòmica/CSIC, Barcelona and CEPR)

Time Varying Risk Premia in General Equlibrium with Production.
Emilio Domínguez (Universidad Complutense, Madrid)
Alfonso Novales (Universidad Complutense, Madrid)

Discussants: Tryphon Kollintzas (Athens University of Economics and Business and CEPR)
Robert Waldmann (European University Institute)

Consumption and Credit Constraints: International Evidence.
Philippe Bacchetta (Studienzentrum, Gerzensee, Université de Lausanne and CEPR)
Stefan Gerlach (Bank for International Settlements and CEPR)

Discussants: Morten Ravn (University of Southampton and CEPR)
Andrew Scott (London Business School and CEPR)

ORGANIZERS
Fabio Canova (Universitat Pompeu Fabra, Barcelona and CEPR)
Rafael Repullo (CEMFI, Madrid and CEPR)
Philippe Weil (ECARÉ, Université Libre de Bruxelles, and CEPR)

Economic Determinants of the Term Structure of Nominal Interest Rates

Martin Eichenbaum (North Western University)
Charles Evans (Federal Reserve Bank of Chicago)
David Marshall (Federal Reserve Bank of Chicago)

Abstract

The paper explores the economic determinants of the term structure of nominal interest rates with emphasis on the role of monetary policy. We measure the exogenous shocks to monetary policy as orthogonalized shocks to the federal funds rate. We find that shocks to monetary policy have a pronounced but transitory impact on short-term interest rates with almost no effect on long terms rates. As a result, the effect of monetary policy shocks on the term structure is to increase the level, flatten the slope, and decrease the curvature of the yield curve. These effects are short-lived dissipating in about 12 months. In contrast, orthogonalized shocks to employment and to a measure of the change in key input prices have a longer-lived effect on interest rates across the maturity spectrum inducing a parallel shift in the yield curve that persists for at least two years. These latter shocks account for a large fraction of the unconditional variance of interest rates. We tentatively associate these shocks with impulses that shift aggregate demand. A more definitive interpretation of these shocks must await an explicit general equilibrium model of the economic environment.

Discussions

Enrique Sentana (CEMFI, Madrid) pointed out that central bankers would agree with the results presented here since they claim that they know only how to move the low end of the term structure. He also stressed five important points. First, it matters a lot when the Federal Funds Rate the authors use is measured since different timing may distort the correlation with other variables. Second, although it is common practice to used revised data, it is much more reasonable for studies like this to use variables measured at the time when decisions are taken. Third, the use of forward rates instead of yield curves could give more clear information about how long and short term rates react. Fourth, although the procedure used by the authors is reasonable, it is probably wasting information and it would be better to use all yields in one large VAR. Finally, what is missing from the analysis is a clear discussion of which economic factors explain comovements of yields. Etsuro Shijoi (Universitat Pompeu Fabra, Barcelona) suggested that the sample may be too long to be stable and suggest dropping the pre-1975 period from the analysis. He also pointed out that there may be some problems in interpreting the identified shocks as monetary policy shocks especially because total reserves respond positively to a tightening of monetary policy. Finally, he wondered if it is correct to leave long term interest rates out of the policy equation. Gordon and Leeper (JPE, 1994) found them important. However, in studies conducted by the discussant, it seems appropriate to leave them out. Harald Uhlig (CentER, Tilburg University and CEPR).suggested that the results presented propose a new and different puzzle, i.e. that monetary policy shocks may have a sizable and prolonged long terms effects on real rates. Michael Wickens (University of York and CEPR) indicates that the Federal Funds Rate responds primarily in the 1979-82 period when the Fed was using money supply targeting.

Dissecting the Predictive Content of the Yield Curve for Economic Activity: the Case of the United States and Germany

Frank Smets (Bank for International Settlements and CEPR)
Kostas Tatsaronis (Bank for International Settlements)

Abstract

In this paper we investigate the economic determinants of the leading indicator property of the slope of the yield curve for future economic activity in Germany and the United States. We use a structural VAR framework which identifies aggregate supply and demand as well as monetary policy and risk premia shocks. We then analyze the effects of these innovations on the real, inflation, and risk premium components of the term spread, and find that the first shocks are responsible for most of the yield curve dynamics for both counties as well as for the predictive content of the spread. Monetary policy shocks have some importance for Germany while risk premium shocks are important in the United States. We also compare the impact of the same innovations on the term structure with respect to the predictions of the expectations hypothesis, and find that these responses are more consistent with the hypothesis in Germany than in the United States.

Discussions

Fernando Restoy (European Monetary Institute) argued that the questions posed by the paper, i.e., what are the determinants of the yield spread and what makes the spread able to forecast output, are interesting and crucial to policymakers. The main problems with the paper is that it is assumed that Central Banks do not react to exogenous changes in expectations, while clearly this is not true in the real world. He also pointed out that simply computing covariances between the current spread and future output may not be so useful unless we also take into account the information contained in past changes in output. Michael Wickens (University of York and CEPR) suggested that what the authors call an innovation in the risk premium, it is really a linear combination of various shocks at different points in time so that there may be aggregations problems that the paper does not take into account. Moreover he suggests that the assumption that the term premium are uncorrelated with monetary policy shocks is implausible since the term premium is a function of the spread and of future interest rates.

Long Term Debt and the Political Support for a Monetary Union

Harald Uhlig (CentER, Tilburg University and CEPR)

Abstract

This paper examines the role of long term debt for the political support of a monetary union. The central idea is that the decision about membership in the union leads to a redistribution between debtors and creditors if they are holding long term debt with a nominally fixed interest rate. For example, if joining the union means a decrease in the inflation rate, creditors should favor joining while debtors should be against it. A government of a high inflation country might strategically try to exploit this effect by selling more long term debt denominated in its own currency rather than in a foreign currency to its citizens. We show that the effect on political support is unclear. While the “creditor effect” of increasing the number of agents holding long-term nominal debt helps generating support for joining the union the “tax effect” of having to raise more taxes in order to pay for the increased real debt payments after a successful monetary union works in the opposite way. The paper then studies a number of special cases and ramifications. In the conclusion this author voices the opinion that issuing long term debt denominated in the domestic currency is a bad idea if one seeks political support for membership enhancing actions.

Discussions

Ramon Caminal (Institut d’Anàlisi Econòmica/CSIC) indicated that the focus of the paper is too narrow and the author could give a much more interesting outlook at the issue by incorporating some game theoretical issue. He also suggests that while the paper claims that taxpayers are worse off with a monetary union, this not unique to joining the EMU and can be obtained in any model where government reduce inflation. Moreover, the model seems to suffer from agency problems, in the sense that the government can take an action independently of voters’ preferences on debt management. On a more casual ground he also wonders if it is really true that voters anticipate that the EMU will imply a capital loss for public debt holders. Gilles Saint-Paul (DELTA, Paris and CEPR) stressed that in the real world governments never issued different types of debt strategically to get political support and this lack of empirical support is a major drawback of the model. He also argues that the model is about credibility and not monetary union. Michael Wickens (University of York and CEPR) suggests that the government could issue debt at a discount to avoid costs of issuance in a monetary union.

Asset Price Fluctuations in a Lifecycle Economy: Do Collateral Constraints Matter?

François Ortaló-Magné (London School of Economics)

Abstract

This paper demonstrates that collateral constrains generate persistent and cyclical asset price and output fluctuations in response to temporary shocks. Following a permanent shock, asset price and output first overshoots the new steady state, then converge to it in cyclical oscillations. To quantitatively assess the effects of the collateral requirement, the model is applied to the US dramatic farmland price boom bust cycle of the past thirty years. The collateral requirement is found to be key in explaining changes in debt leverage, volume of transactions, reasons for sale, and farmer’s entry and exit, but not land price fluctuations.

Discussions

Anton Braun (CEMFI, Madrid) wonders if there is any way to increase the response of land prices to shocks by introducing further frictions in the model. He also suggests that the process for expectations is very ad-hoc and this is a problem for the empirical examination of the model. Further, he suggests that it would be very interesting to study how the demographic features of the model change in response to shocks. Vincenzo Quadrini (Universitat Pompeu Fabra, Barcelona) asked why do we need credit constraints to generate cycles in land prices and he suggests that a simple Lucas-tree model can generate this kind of feature. He is left wondering what are the crucial aspects of the paper that drive the results. Tryphon Kollintzas (Athens University of Economics and Business and CEPR) argues that land is not a divisible commodity and once this is taken into account it is certainly possible to increase the volatility of land prices in response to shocks.

Borrowing Constraints and the Financing of Household Capital in Economies with Banking

Javier Díaz-Giménez (Universidad Carlos III de Madrid)
Luis Puch (Universidad Carlos III de Madrid)

Abstract

In this paper we quantify the individual, aggregate and welfare consequences of imposing different levels of borrowing constraints in a model economy that includes uninsured household-specific risk, household capital accumulation and an explicit banking sector. We calibrate the model economy to Spanish data and we compare the steady state of a model economy where households can borrow up to the resale value of their household capital with the one that obtains in a model economy with no credit. Some of our findings are the following: i.) output is 10% larger in the model economy with credit and most of this increase is accounted for by the imputed rent of owner occupied housing, ii.) the stock of household capital is 47% larger in the economy with credit, iii.) the stock of household financial assets is 50% smaller in the model economy with credit, and iv.) switching from the economy with no credit to the economy with credit brings about an increase in welfare which is equivalent to 8.71% of total wealth of the economy with no credit.

Discussions

Xavier Freixas (Universitat Pompeu Fabra, Barcelona) indicates that the assumption that the price of land is exogenous is problematic since it implies an infinite supply of residential housing. Also the restriction on the amount of housing is not very credible (usually housing is assumed to increase with wealth). Finally, he stressed that the paper does not make very clear what do we gain in going from a standard partial equilibrium to the general equilibrium setup used by the authors. Albert Marcet (Universitat Pompeu Fabra, Barcelona) points out that the model is very complex and the details of the financial sector very complex so that it is hard to understand exactly what is the additional contribution of each of the features to the results. He also suggests that the author should do more sensitivity analysis on the parameters in order to assess the robustness of the results because the financial sector in Spain has changed a lot over the last 10-15 years and the parameters used may not correspond to neither the pre nor the post liberalization situation.

Growth-Enhancing Bubbles

Jacques Olivier (Hautes Études Commerciales, Jouy-en-Josas)

Abstract

This paper shows that speculation on stock markets can increase the long run growth rate of an economy by raising the market value of firms. In a world where agents have the choice between supplying their labor on the labor market and creating new firms, a (price-increasing) bubble on the value of firms reduces labor supply and raises investment. Consequently, more firms are being created at the equilibrium and long-run growth is permanently increased. A testable implication of the model is that financial markets have a positive impact on growth even when investment productivity is unchanged. This is consistent with recent empirical evidence brought forward by Harris (1994). Finally, some regulatory implications of the model are briefly discussed.

Discussions

Graziella Bertocchi (Universitá di Modena) suggests that with an endogenous growth model, the claim that bubbles always decrease welfare is only partially correct because it takes a sufficiently strong externality to insure underaccumulation. For example, in endogenous growth models of the type analyzed by Azariadis and Reichlin (1994), bubbles may increase welfare. Also the policy implications that the paper draws are not completely convincing since negative bubbles can occur and this may depress growth. Angel de la Fuente (Institut d’Anàlisi Econòmica/CSIC, Barcelona and CEPR) indicates that bubble equilibria have unattractive features and look less plausible than non-bubble equilibria. This makes the thesis of the paper not very convincing since it argues that financial markets can engine growth and this occurs only in the presence of bubbles. He also stresses the fact that the empirical relevance of this argument is hard to assess: there is no reason to believe that more developed stock markets are more prone to bubbles.

Time Varying Term Premia in General Equilibrium with Production

Emilio Domínguez (Universidad Complutense, Madrid)
Alfonso Novales (Universidad Complutense, Madrid)

Abstract

Endowment economies have generally been considered when trying to reproduce the empirical rejection of the expectation hypothesis as an implication of equilibrium asset pricing models. Previous attempts have not been successful: large risk aversion parameters are needed to produce sizable term premia and even then, the expectation hypothesis is not rejected. We present an economy with a time-to-build technology, in which consumption is subject to cash-in-advance constraints. In it, the expectations hypothesis of the term structure does not hold. Monetary shocks are much more important than real demand or supply shocks in producing the result.

Discussions

Tryphon Kollintzas (Athens University of Economics and Business and CEPR) indicates that the paper is undersold by the authors and that, potentially, the model can deliver many more implications that those presented. Moreover, the paper should use more statistics to examine the validity of the model. For example, the paper can be used to examine what the yields curve has to say about macro variables, both in terms of explanatory and forecasting power. Robert Waldmann (European University Institute) claims that is not really clear what drives the model. Results appears to emerge only because the cash-in-advance constraints works like a consumption tax (you need cash to pay for consumption but not for investment). Also he suggests that the frequency of the model is not well chosen since over quarters the money to output ratio is not constant. He also examines whether the alternative explanation of agents not having rational expectations can be entertained to explain the behavior of the yield curve and concludes that, quantitatively speaking, such an approach will fall short of achieving the task. Michael Wickens (University of York and CEPR) indicates that tests of the expectations hypothesis of the term structure assuming zero risk premium do not make any sense and should not be undertaken. Fabio Canova (Universitat Pompeu Fabra, Barcelona and CEPR) argues that it is still reasonable to test the validity of a model (not of an hypothesis) using a somewhat misspecified statistics. David Marshall (Federal Reserve Bank of Chicago) suggests that it is not clear from the paper what makes the intertemporal marginal rate of substitution conditional heteroskedastic in the model.

Consumption and Credit Constraints: International Evidence

Philippe Bacchetta (Université de Lausanne and CEPR)
Stefan Gerlach (Bank for International Settlements and CEPR)

Abstract

In this paper we show that the presence of credit constrained consumers has two important implications for aggregate consumption. First, consumption displays “excess sensitivity” to income and, more interestingly, to credit and the wedge between borrowing and lending rates. Second, the excess sensitivity varies over time in response to changes in severity of the credit constraints. Our empirical work supports these propositions. Using data for the United States, Canada, United Kingdom, Japan and France, we find a substantial impact of credit aggregates on consumption in all countries considered, and in the United States and Canada, an impact of the borrowing-lending wedge. Furthermore, using time-varying coefficient techniques, we show that there is some tendency for the excess sensitivity to decline over time, particularly in the United States.

Discussions

Morten Ravn (University of Southampton and CEPR) suggests that the paper is faced with an aggregation problem: tests are undertaken with aggregate data while the theory is at micro-level and this, as shown by Attanasio and Weber (1996), may bias the results. He also suggests that credit constraints do not make much sense from a theoretical point of view and he suggests that collateral type of constraints are more appropriate. Andrew Scott (London Business School and CEPR) says that it is not clear from the paper how to interpret the results since there are serious aggregation problems. He is also unsatisfied with the econometric approach and he claims that the paper ignores one important restrictions which is contrary to the liquidity constraints theory of consumption, i.e. that consumption is seasonal while income is not seasonal. Moreover the claim that monetary policy will have effects on consumption when there are liquidity constraints is correct only if we include durables into the measure of consumption we use, which is not the case here.