Thursday, August 9, 2012


What Explains the New Zealand Dollar?

 
The exchange rate is a very tough nut to crack.  The New Zealand dollar has been freely floating since 1985, but some small interventions took place in the past few years.  By contrast, the Bank did not intervene in the exchange rate market from 1985 to early 2000.  Freely floating currencies in general exhibit frequent swings, i.e., volatility, and that has been historically the case for the Kiwi and the Australian dollars.  There is an argument for that; freely floating exchange rates act as shock absorbers and insulate the real economy from shocks. High volatility makes them hard to explain by economic fundamentals , which have smaller variances.
We do not have a concrete explanation for exchange rate movements.  Many economists believe that exchange rates move around randomly, like other asset prices.  Models of the exchange rate from the 1970s and 1980s—such as the monetary model, the real interest rate differential model and the overshooting model— which include economic fundamentals as explanatory variables failed to predict the movements of the nominal and real exchange rates.[i]
Benigno et al. (2011) provide new answers and insights to exchange rate fluctuations by providing a general equilibrium theory. Their DSGE model includes an endogenous deviation from a UIP condition and that risk premium is driven by three uncertainties: monetary policy uncertainty, inflation target uncertainty and productivity shocks uncertainty.  Monetary policy uncertainty appreciates the currency in the medium-term through the hedging motive. Uncertainty improves hedging properties.  Agents search for relatively safer currencies in times when there is bad financial, economic or political news which increases demand for the currency and appreciates it.  A high volatility of TFP shocks depreciates the currency.[ii] 

The volatility of TFP shocks seems to explain large spikes in the TWI New Zealand exchange rate.[iii]  Volatility is the squared changes in TFP.  Figure 1 demonstrates.  The TWI is inverted for convenience (increase denotes depreciation) .  The spikes represent large depreciation.

Figure 1

For monetary policy shocks, the Reserve Bank of New Zealand changed its operating procedure in 2006. The description of this policy change is found in The Reserve Bank Bulletin article.[iv]  Essentially, the policy change increased the money supply.  The exchange rate depreciated first then significantly appreciated.  Then the international financial crisis hit. And the Kiwi dollar has been appreciating. Figure 2 plots volatility of the OCR (dotted black line) and the Overnight Interbank Cash Rate (solid black), and the level of the TWI (inverted, red dashed line).  These volatility measures are the monthly averages of daily squared changes in the interest rates. Again, there is an association between monetary policy volatility and the level of the TWI.  However, the volatility of interest rate is associated with large depreciation rather than appreciation as predicted by Benigno et al. (2011).  It may indicate net selling of New Zealand dollar dominated assets rather than net purchases. 

 
Figure 2

Although large spikes in the TWI are well captured by both monetary policy and TFP shocks volatilities, we do not have a robust explanation of the Kiwi dollar.

[i] See, among others, Obstfeld, M. and K. Rogoff, Foundation of International Macroeconomics, MIT Press, 1996.
[ii] Benigno, G., P. Benigno, and S. Nistico, Risk, Monetary Policy and the Exchange Rate, 2011.
[iii]  Data are quarterly The data source is the Reserve Bank of New Zealand and the International Labour Organization. TFP shocks are the squared change in the level of TFP.  TFP is simply equal to log GDP – share of capital* log capital – share of labour * labour . The share of capital, which I calculate from the National Income Account as the ratio of gross operating surplus / GDP ratio in 2011, and is fixed to 0.30 and the share of labour is assumed to be fixed at 0.70 such that the shares sum to unity.  There is no theoretical reason to assume this, but I just do so for convenience as it does not affect the results.  I calculate the initial stock of capital to be 2 times GDP in January 1999.  I assumed that capital grows according to the following law of motion, capital growth equal to the percentage of GDP invested in the economy* real GDP + the depreciation rate * current capital. I assumed that 20 percent of GDP is invested annually (0.05 per quarter); and the depreciation rate is 6 percent annually (0.015 per quarter).  And labour is working age population.  I use the ILO annual data, which I converted to quarterly data.  TFP is usually measured with a lot of assumptions.  The plot, however, is very robust to these assumptions.  I tried a lot of assumptions but the wrinkles remained the same.

Wednesday, August 8, 2012


New Zealand and Australia’s Short Term Interest Rate Volatility
I plot the daily New Zealand Official Cash Rate (OCR), which is the Reserve Bank policy rate, and the Overnight Interbak Cash Rate, which is the interest rate banks charge each others.  I also plot the difference between the two rates (OCR minus the Overnight Interbak Cash Rate).  These are averages of daily data by month from 1999 to 2012.

 
Figure 1
The two rates are equal until late 2005, when the OCR jumped significantly such that the difference jumped from zero to 0.42, followed by a significant decline in both the OCR and the market rate.  In 2006, both rates diverged significantly and the differences never recovered; the divergence is significant and prolonged.

The Reserve Bank of New Zealand changed its operating procedure in 2006. The description of this policy change is found in a Bulletin article.[i]  Essentially, the Reserve Bank increased market liquidity significantly.  It made more money available for banks.  When the supply of money increases, the Overnight Interbak Cash Rate should fall, but that clearly did not happen immediately because the difference in figure 1 is negative until February 20007, indicating a rise in the market rate above the cash rate.  These differences suggests that the market was uncertain about the policy change, and perhaps the fact that the New Zealand financial market is very thin.

In 2008, the market rate began to equal or be less than OCR, reflecting the expected response to the policy.  Differences are positive in 2010 indicating the falling of market interest rate below the OCR and reflecting the policy change in 2006.  It took a very long time for market interest rate to fall.  However, the data were also reflecting the international financial crisis.

Let us look at the same data for Australia.  Australia maintained the same operating procedure throughout.  In this case we expect to observe the OCR and the Overnigh
Interbak Cash Rate to be equal all the time unless there are some shocks pushing the two rates apart.  Figure 2 shows that differences between the policy rate and the market rate occurred in early 1999.  From 2002 to-date, and during the international financial crisis, Australia’s OCR and the Overnight Interbak Cash Rate are almost identical, with differences equal to zero. The Australian market has been adequately liquid.

Figure 2

Figure 3 shows New Zealand’s OCR and Overnight Interbank Cash Rate’s volatility (uncertainty).


Figure 3
New Zealand’s monetary policy volatility, measured by the monthly average squared changes in the OCR during the financial crisis, has been low except for a short period in 2009.  Monetary policy and market volatility were equal and only began to diverge in the days and months leading up to the crisis, during and after the crisis. The market’s volatility has been significantly higher than monetary policy’s and continues to be so.  These persistent differences in volatility and persistent volatility in the financial market are important research questions, which might have some real implications.
.
It gets more interesting when comparing the New Zealand figures to Austraia.  Figure 4 shows that New Zealand's volatility is significantly different in Australia.  For Australia, both monetary policy and market volatility were identical and small throughout the crisis, except for a very small spike in 2009.  During the international financial crisis Australia’s OCR had a small blip.


Figure 4

The next two graphs plot the New Zealand and Australia’s OCR and Overnight Interbank Cash Rate together.


Figure 5
Monetary policy volatility during the financial crisis is higher in New Zealand than Australia’s?


Figure 6

Similarly, financial market volatility in New Zealand, just before, during and after the financial crisis are significantly higher than Australia’s. 

The interest rate volatility might have some implications for the level exchange rate, which will be investigated in the next blog, see Benigno et al. (2011).[i] 



[i] Benigno, G., P. Benigno, and S. Nistico`, Risk, Monetary Policy and the Exchange Rate, 2011.







Tuesday, August 7, 2012


Austerity versus Growth


You must have heard the news media, newspapers articles and commentators talk about differences between the position of the German government of austerity and the French’s position of growth.  I could not find any written document that explains what the French president means by growth.  The confusion arises from using the two words, austerity and growth, at the same time.  It certainly suggests that the French president (and probably President Obama too) are not for austerity, and therefore are for increased spending.  But it does not necessarily and immediately follow that increased government spending causes higher economic growth.

The austerity argument is about reducing government expenditures, which could include reducing the size of the public sector by firing people, reducing and suspending some social programs, reducing the wage bill, salaries, pensions, perks, services and other expenditures.  Clearly, public sector employees do not like this approach, and who can blame them. But, what has this got to do with growth?  The government maybe able to induce a temporary increase in output, but that should not be confused with growth.

It is unclear what the growth program of the French president is. If it is meant to be the opposite of austerity, then all that is required is an increase in government spending.  What French President Hollande actually meant, and this should be explained, is that government expenditures could stimulate demand via spending, which would then get the economy out of recession, but it has nothing to do with economic growth either.

These economic thoughts are widespread, but confusing.  Some top economists also believe that government spending can resolve Europe’s economic problems and that government spending can cause economic growth, but careful empirical analysis by John B. Taylor and Valerie A. Ramey in the Journal of Economic Literature Vol. 49, No. 3, September 2011 shows that there is no credible evidence for this conjecture.  See a very interesting 5-page article by Edmund Phelps, the Nobel laureate, Keynes Vs. Hayek: Stumbling blocks to spending our way to prosperity, produced by Reuters, November 8, 2011.  You will get the message.

An increase in government expenditures, whether the level or the growth rate, cannot always cause real production of goods and services to grow faster. It could increase consumption and that would increase GDP level such that the economy gets out of the slump.  That’s not growth.  There is a significant difference in meaning between an increases in  the level of output and a sustained increase in output.  To affect the growth rate of output government spending must be directed at factors of production, or factors that cause technical progress such as education, R&D etc but that’s not as easy as politicians think and cannot be guaranteed.[1]

Growth of the production of goods and services is the speed at which the economy’s production increases from one year to another, and can only increases if technical progress grows.  It took economists decades of research to understand what that thing is and what might drive it.  Growth economists demonstrate that it is driven by knowledge and new ideas of producing new goods and services—innovations if you will.  These ideas do not come from thin air. They vary from one country to another.  They might depend on productive investments such as investments in human capital, investments in better quality of human capital or investments in research and development. 

No one would have believed in the 1950s that one could ever send a copy of a piece of paper (the modern fax) around the globe in a few seconds. This is knowledge producing a new commodity and creating markets.  Communication technology, which we are observing today like the iPad and smart phones, smart TVs, the internet, medical tools, fabrics and materials etc. are examples of technical progress leading to economic growth. 

Economists went further to suggest that there might be other variables that affect the growth of technical progress, such as geographical location and distance from big markets, the set of laws that regulate the markets and the businesses, the culture where some are savers and others are spenders, government policies, etc.  None of these  testable theoretical propositions include government expenditures.

Governments could invest in science, knowledge, human capita, etc. and these investments might affect the growth of technical progress in the long run.  They might be important in early stages of developments, but I am not sure this is what the French politicians, the news media and some top economists have in mind when they talk about growth policy, presumably, as an alternative to austerity

The major problem Europe needs to deal with is how to reduce unemployment, and this will not be solved by increasing government expenditures or the money supply.  It will not be reduced by inflating the economy.  I checked the IMF world economic outlook data set; the historical French aggregate time series data do not show a negative correlation between GDP growth and unemployment over the period 1980 to-date.  It means that an increase in GDP growth rate is not expected to be associated with a lower unemployment rate. 

Ed Prescott’s[2] explanation, which is confirmed by others such as Shimer and Nickell among others,[3] is that the Europeans are taxed relatively higher than Americans—presumably so that the government can finance social programs and the large public sectors—therefore, people work fewer hours.  Sadly, a lot of people will be unemployed for a long time unless the European governments, especially in France and the Southern European countries, reduce the size of their public sector, reduce taxes on labor and make their labor markets more flexible. But this won’t happen because people elected governments that promised just the opposite.


[1] Johnson, R., W A Razzak and Steve Stillman, “Has New Zealand Benefited from Its Investments in Research and Development?” Applied Economics, Volume 39, Issue 19, November 2007, 2425-2440, show that 40 years of R&D spending in New Zealand did not result in any spillover.
 [2] Prescott, E., Why Americans work so much more than European? Federal Reserve Bank of Minneapolis Quarterly Review, Vol. 28, No.1, July 2004, 2-13. 
[3] Shimer, R., Convergence in Macroeconomics: The Labor Wedge, American Economic Journal Macroeconomics, Vol. 1 No. 1 (January), 2009, 280-297.
Nickell, S.,Employment and Taxes, CESIFO Working Paper No.1109, 2003 (December).