Tuesday, March 25, 2014

The Kiwi Dollar and Monetary Policy

The empirical exchange rate models of the seventies and the eighties, where the exchange rate depends on the relative changes in market fundamentals (e.g., output, money, interest rate, and inflation differentials between the home and the foreign country), could not outperform the random walk model in and out of sample. The fundamentals vary much less than the exchange rate; hence the correlation between them is too small. As a result, there is a general agreement among economists that the exchange rate is very difficult to predict.[1]

On March 13, 2014, the Reserve Bank of New Zealand increased the Official Cash Rate (OCR) by a quarter of a percent for the first time since 2011.  People understand that a rise in interest rate today and the near future means that the RBNZ is predicting increasing inflationary pressures. The question is to whether we can predict the exchange rate from interest rate differentials?  

The Uncovered Interest Rate Parity condition (UIP) says that risk-neutral investors would be indifferent to interest rates in the home and the foreign countries because the exchange rate between the two countries is expected to adjust  such that the dollar returns on dollar deposits is equal to the dollar return on foreign deposits, thereby eliminating the potential for an uncovered interest rate arbitrage profits. My own research at the Reserve Bank of New Zealand, more than a decade ago, showed that the UIP condition does not hold in a number of currencies Vis-a'-vis the U.S. dollar, but holds much better in the case of the New Zealand – Australian currency.

I plot the 90-day interest rate differential (i – i*) between New Zealand and the U.S., where asterisk denotes the U.S. effective federal fund rate, against the expected depreciation rate, and the same for New Zealand and Australia using monthly data from 2001. The expected depreciation rate is the spot nominal exchange rate minus the sample’s average as a proxy for the expected exchange rate.[2]

Figure 1

There correlation between the interest rate differential and the exchange rate (USD-NZD) is weak, and it breaks down completely in 2009.[3] New Zealand and the U.S. short term interest rates (policy rates) remained unchanged for a long period of time since the global financial crisis, so the change in interest rate differential remained constant while the exchange rate varied, which explains why the correlation between the interest rate differential and the exchange rate depreciation is low.[4]  

Monetary policy affects output and inflation. The increase in interest rate today reduces future inflation and output growth. Lower output growth relative to the U.S. depreciates the Kiwi dollar and a lower inflation relative to the U.S. appreciates the Kiwi dollar. The correlations between output growth differential and inflation differential between New Zealand and the U.S., and the depreciation rate are weak.[5] Based on these figures, it is unclear how the USD-NZD dollar would move in the future.[6]

Figure 2

I do not plot the Australian-U.S. UIP condition because it is just as bad as the New Zealand-U.S. case. However, contrary to the two cases which involve the U.S. dollar, the UIP for the (AUD-NZD) seems to hold. The exchange rate moved in tandem with the interest rate differential over the sample, and recent appreciation of the Kiwi against the Australian dollar is pretty much a result of monetary policy conditions summarized by the interest rate differential.[7] The correlation between the interest rate differential and the exchange rate depreciation rate is reasonably high, and significantly higher than the U.S. dollar UIP.[8] And since output growth and inflation are very close in New Zealand and Australia, they are not good predictors of the currency.[9] Thus, the short-term interest rate differential is a better predictor of the exchange rate in this case.

Figures 3 and 4 plot the interest rates and output growth rates of the three countries. Clearly Australia’s interest rate is different from that of the U.S. and New Zealand. Australia’s output growth was much higher than that of the U.S. and New Zealand during the recent global financial crisis, which probably explains the different responses of monetary policy. The interest rate differential between New Zealand and Australia varies a lot more than that between New Zealand and the U.S. And that this variation explains why the UIP holds better between New Zealand and Australia compared to the New Zealand - U.S. and Australia-U.S. cases.

Figure 3

Figure 4


[1] Meese, R. and K. Rogoff, 1983a, Empirical Exchange Rate Models of the Seventies: Do They Fit Out of Sample?, Journal of International Economics, 14, 3-24.

Meese, R. and K. Rogoff, 1983b, The Out-of-Sample Failure of Empirical Exchange Rate Models: Sampling Error or Misspecification? In Jacob Frankel, ed., Exchange Rates and International Macroeconomics, University of Chicago Press. 

Flood, Robert P. and Rose, Andrew K., 1995, Fixing Exchange Rates A Virtual Quest for Fundamentals, Journal of Monetary Economics, Elsevier, vol. 36(1), pages 3-37, August. 

[2] The New Zealand data are taken from the Reserve Bank of New Zealand, the Australian data are from the Reserve Bank of Australia and the U.S. data are from the Federal Reserve Bank of St Louis.

[3] The correlation is 0.008.

[4] The standard deviation of the interest rate differential is 0.01 while the standard deviation of the exchange rate is high, 0.20.

[5] It is 37 percent, and the correlation between the inflation differential and the depreciation rate is -41 percent.

[6] Future oil price contracts (Brent, NY light, Oman crude etc.) have been falling steadily over time. This trend might be a good predictor of the future trend of the U.S. dollar. Hence, the Kiwi dollar might depreciate against the U.S. dollar in the coming year.

[7] The standard deviation of the interest rate differential is 0.01 and that of the exchange rate is 0.05, which are closer than those in the case of the U.S. dollar.

[8] The correlation coefficient between the interest rate differential and the depreciation rate is 74 percent. Further, regressing the New Zealand 90-day interest rate (i) on a constant term, trend, and the Australian 90-day interest rate plus the depreciation rate (i*+d log (s)), where (d) is the difference operator, using Fully Modified OLS method, gives a slope coefficient that is insignificantly different from unity, and statistically insignificant constant term and trend.

[9] The correlation between output growth differential and the depreciation rate is 13 percent and that between the inflation rate differential and the depreciation is zero.

Sunday, March 16, 2014

Regulations and Corruption

"That government is best which governs least." 
 Thomas Jefferson

The recent global financial crisis ignited a new wave of debate about the need for more regulations and government interventions. More regulations might very well be necessary. But it’s unclear how much regulations are needed and where. And there are always questions of over and under regulation, their efficacy, and the effects on the working of the markets. It’s quite conceivable that more regulations can lead to more corruptions and more corrupt countries can in fact have more regulations already. Here are some empirical cross-sectional correlations, which seem alarming.

Measuring corruption

Transparency International publishes a Bribe Paying Index. It involves twenty eight of the world's largest economies - they represent 80 percent of the world total outflow of goods, services and investments. It is supposed to measure the perceived likelihood of companies in these countries to pay bribes abroad. The scale is 0 to 10, where the maximum score corresponds with the view that the companies from that country never pays bribes, and zero corresponds with the view that they always do. We do not have data on bribes paid or perceived to be paid in own country. I use this index as a proxy measure of corruption. The index was published for the years 1999, 2002, 2008 and 2011. 

Measuring regulations

The World Bank publishes data on the ease (difficulty) of doing business. I chose The Number of Procedures required to establishing a business, the Number of Procedures to Deal with Constructions and the Number of Procedures to Enforce Contracts as measures of regulations. The assumption is that the longer these procedures are the more regulations the country has. There is no particular reason for me to choose these indices and I could have easily chosen other data such as getting credit, paying taxes etc.

Both indexes are for the year 2011 because the bribe index is not available for more recent years.

Putting them together

I invert the bribe paying index before I plot it, separately, against the three numbers of procedures above, which measure regulations. Figures 1, 2 and 3 show a clear positive correlation between regulations and corruption. Obviously I cannot pin down the direction of causality. But I would argue that causality is not really important. There could be bi-directional causality running from regulations to corruption or vice versa. But what matters the most is the close association between these two indices, which are constructed by two different and independent institutions.

Countries which have relatively less regulations have companies which do not anticipate paying bribes abroad. It would not be very wrong to argue that that they are also less likely to pay bribes in their own countries. Countries which are relatively highly regulated have companies, which have a propensity to paying bribes abroad, and most likely they would do so in their own countries. Further, it may well be the case that relatively more corrupt countries are also countries that probably tend to regulate more to extract more bribes.

I conjecture that a country which is characterized to be a difficult one to do business in, difficult to register a property and difficult to enforce a contract is also a country were bribes are anticipated to be paid. While countries like Canada and Australia are ones which are relatively less regulated and relatively less corrupt, countries like China, Russia, the UAE and Argentina represent the other polar side. Majority of countries are in between.

I would ague that one should be very wary of over-regulation as much as being wary of under-regulation. Too many regulations could be just as bad as no or fewer regulations. Striking the right amount is important, such as the marginal benefit must equal the marginal cost.

Figure 1

Figure 2

Figure 3