I’ve seen a lot of presentations on macroeconomics and international finance over the past week. I’m particularly interested in the paper on interest rate parity.
Pablo Cuba Borda, Federal Reserve Board of Governors
Martin Bodenstein, Federal Reserve Board of Governors
Niels Gonemann, Federal Reserve Board of Governors
Ignacio Fresno, Federal Reserve Board of Governors
We propose a costly international financial intermediation model that links exchange rate movements to changes in demand for domestically produced goods relative to demand for imported goods (trade rebalancing). Our model is consistent with stylized facts about exchange rate dynamics, including those related to trade balances, which are generally overlooked in the literature on exchange rate determination. Quantitative estimates suggest that trade rebalancing explains about 50% of exchange rate fluctuations over the business cycle, while exogenous bias in the unincorporated interest rate parity, the main source of exchange rate fluctuations in the literature, accounts for just over 20%. Using data on trade flows or trade balances is important for properly identifying the determinants of exchange rates. Our model thus overcomes the sharp dichotomy between real exchange rates and macroeconomics inherent in other models of exchange rate determination.
Discussant:
Federica Romei, University of Oxford
In this first paper, I pointed out that the paper was trying to explain the Fama result (the forward premium ratio does not correctly predict exchange rate depreciation), but that this is not a constant. BCCF(IMF) emergency room 2022).
Figure 1: EUR/USD 3-month ex-post annualized depreciation (vertical axis) and 3-month foreign US-EUR interest differential (horizontal axis) for 1999M01-2005M04 (top panel), 2005M05-2017M04 (middle panel), and 2017M05-2021M06 (bottom panel). Since the samples are related to interest rate data, the 1999M01-2005M04 sample includes 3-month ex-post depreciation through 2005M07.
Tarek Alexander Hassan, Boston University and NBER
Thomas Mertens, Federal Reserve Bank of San Francisco
Wang Jingye, Renmin University of China
The formal long-run risk and habit model reconciles high equity premiums with smooth risk-free rates by inducing an inverse relationship between the variance and mean of the probabilistic discount factor. We show that this highly successful solution to the closed-economy asset-pricing puzzle is fundamentally flawed when applied to open economies with complete markets. This requires that differences in currency returns arise almost entirely from predictable increases in interest rates, rather than from differences in interest rates. On the other hand, in our data, exchange rates are largely unpredictable, and currency returns differ because interest rates vary widely across currencies. We show that the currency risk premium arising from the formal long-run risk and habit model cannot be consistent with this fact. We argue that this tension between formal asset pricing and international macroeconomic models is a major reason why researchers have had difficulty reconciling the observed behavior of exchange rates, interest rates, and capital flows across countries. The absence of such a unified model is a major obstacle to understanding the effects of risk premia in international markets.
Discussant:
Hanno Lustig, Stanford University and NBER
Seunghoon Na, Purdue University
Yinxi Xie, Bank of Canada
This paper investigates exchange rate dynamics and their forecast errors by incorporating bounded rationality into a small-scale open economy New Keynesian model. Decision makers have limited forward-looking ability and can only plan finitely into the future. This leads to dynamic overshoots in real exchange rate forecast errors over various time horizons. We also distinguish between short-run and long-run expectation formation, where the law of repeated expectations is violated. This framework provides a microscopic basis for understanding the temporal and forecast horizon variability in the uncovered interest valuation (UIP) puzzle. The model predictions for these UIP violations are qualitatively and quantitatively consistent with empirical estimates.
Discussant:
Rosen Balchev, Boston College and NBER
I found this paper particularly compelling because, together with Guy Meredith, we found relationships consistent with this model. Chin and Meredith (1998).
Figure 2: Panel beta coefficients at various horizons. Note: Panel estimates for six currencies up to 12 months, US dollar, euro deposit rates, 1980Q1-2000Q4; 3-year results zero coupon yield, 1976Q1-1999Q2; 5-year and 10-year, constant yield to maturity, 1980Q1-2000Q4 and 1983Q1-2000Q4 (last observation corresponds to exchange rate data). Source: Chin (2006).
Some updated results are included in the 2022 post.
