Critical Analysis Of Currency Carry Trade Strategy

Theoretical Models

Critical analysis of the currency carry trade strategy employed in the financial world to leverage the high interest rate environment prevailing in a foreign market.
 

A Carry trade strategy is a way to make money by borrowing a currency with a rather low interest rate, convert that currency to another currency or bonds in another country, and make money on the basis that the bought currency will increase in value, or the bought bond gives a higher interest, and then convert back to the original currency. When buying strictly in terms of currency it is crucial to make sure the exchange rate between the two currencies doesn’t change in a negative way, because that way we could actually lose money.

In our FX Carry Trade portfolio we have chosen to focus on the foreign exchange rate of AUD/USD. Our work will be based on historical data as well as recent data collected in EIKON, official government websites and various news sites.

For the chosen foreign exchange rates we conducted a linear regression analysis. We collected historical data with a quarterly interval. The regression analysis is used by us because it demonstrates a simple way of analyzing data. We opted to input the inflation rates and interest rate of both Australia and The United States and analyzed on the percentage change. We refer to section 3 for the regressions. Our carry trade portfolio will be in the span of January 2016 – June 2016.

As the interest rate in Australia is a lot higher than in USA we expect that our carry trade strategy will involve investing in Australian dollars since we get a higher profit over time and when calculating the forward spot rate we don’t expect that number to have depreciated or appreciated enough to justify converting. In the latest year the AUD/USD foreign exchange rate has been up and down but on average on a stable level which also helps justifying the forward rate not being drastically changed at the end of our carry trade.

Theoretical Models

We have chosen the two commonly used theories which are the IRP and RPPP. These models are based on the interest rate of a country and the inflation rate of a country.

2.1 Relative Purchasing Power Parity

The absolute purchasing power parity states that spot exchange rate is determined by the relative prices of similar prices of similar basket of goods (the Big Mac Index example). However, the relative purchasing power theory makes the statement that the relative change in prices between two countries over a period of time determines that period’s change in the exchange rate. This is interesting because it takes the PPP theory a bit further. If spot exchange rate between two countries starts in equilibrium any change in the differential rate of inflation between the countries is often – over the long run – also equal, but opposite, change in the spot exchange rate.

Relative Purchasing Power Parity

2.2 Interest Rate Parity

The interest rate parity (IRP) is a theory that links the foreign exchange markets and the international money markets. It states that the difference in national interest rates for bonds, securities etc. of similar risk and maturity should be equal to the forward rate premium or discount for the foreign currency. Basicly if you have a currency with an interest yield of 4.0 % and another currency with an 8.0 % interest yield, the forward premium is 3.96 % (you subtract transaction costs). The theory is relevant for our carry trade strategy because it can determine whether a certain currency is better off being converted to another currency or simply invested with the country’s interest rate. 

3- E-views Result:

3.1- IRP:

Dependent Variable: EXCHANGE_RATE
Method: Least Squares
Date: 05/28/16   Time: 19:58
Sample (adjusted): 1995Q4 2015Q4
Included observations: 81 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
CASH_RATE	-0.033821	0.006191	-5.462718	0.0000
C	0.850723	0.020604	41.28826	0.0000
R-squared	0.274173	Mean dependent var	0.771084
Adjusted R-squared	0.264985	S.D. dependent var	0.152852
S.E. of regression	0.131044	Akaike info criterion	-1.202181
Sum squared resid	1.356636	Schwarz criterion	-1.143059
Log likelihood	50.68834	Hannan-Quinn criter.	-1.178461
F-statistic	29.84129	Durbin-Watson stat	0.184143
Prob(F-statistic)	0.000001

3.2- PPP:

Dependent Variable: EXCHANGE_RATE
Method: Least Squares
Date: 05/28/16   Time: 20:09
Sample (adjusted): 1995Q4 2015Q4
Included observations: 80 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
UNEMPLOYMENT	-0.030081	0.005825	-5.163990	0.0000
INFLATION_	0.013209	0.003831	3.447452	0.0009
C	0.838803	0.020943	40.05092	0.0000
R-squared	0.367131	Mean dependent var	0.771216
Adjusted R-squared	0.350693	S.D. dependent var	0.153811
S.E. of regression	0.123941	Akaike info criterion	-1.301251
Sum squared resid	1.182817	Schwarz criterion	-1.211925
Log likelihood	55.05004	Hannan-Quinn criter.	-1.265438
F-statistic	22.33407	Durbin-Watson stat	0.164007
Prob(F-statistic)	0.000000

3.3- The last model of combined : ( Inflation rate and Cash rate) 

Dependent Variable: EXCHANGE_RATE
Method: Least Squares
Date: 05/28/16   Time: 20:14
Sample (adjusted): 1995Q4 2015Q4
Included observations: 81 after adjustments
Variable	Coefficient	Std. Error	t-Statistic	Prob.
INFLATION_	0.012770	0.003723	3.430215	0.0010
CASH_RATE	-0.031008	0.005866	-5.286370	0.0000
C	0.835498	0.019832	42.12784	0.0000
R-squared	0.369312	Mean dependent var	0.771084
Adjusted R-squared	0.353141	S.D. dependent var	0.152852
S.E. of regression	0.122935	Akaike info criterion	-1.317992
Sum squared resid	1.178811	Schwarz criterion	-1.229308
Log likelihood	56.37866	Hannan-Quinn criter.	-1.282411
F-statistic	22.83726	Durbin-Watson stat	0.175242
Prob(F-statistic)	0.000000

The last regression is the best result because of adjusted R-squared is more confident because it explains 35% of the model errors.

more to discuss :

CR -coefficient 

Level of Significance: a = 0.05

P-value: 0.000 < 0.05

So, this is significant as it is less than the 95% confidence threshold of 0.05. 

3.4- Regression Equation: 

AUD/USD = b₀ +b₁ CASH RATE+b₂ INFLATION RATE +μ_t 

AUD/USD = (0.835)+( -0.031)+( 0.013) 

Where,

b₀= stationary = 0.835

b₁=coefficient of CASH RATE  : The exchange rate AUD/USD change by -3.1% in the same way for every unit change in the Cash Rate

b₂=coefficient of INFLATION RATE : The exchange rate AUD/USD change by 0.013 in the same way for every unit change in the INFLATION RATE. 

Various models have been formulated in order to ascertain the foreign exchange rate accurately. Regression model is used in this context due to its simplicity and ease of use. 

Cash rate and Inflation are two of the vital units which are used in the application of the model. These two units materially affect the determination of the exchange rate. Apart from this, there are other economic variables which are also considered in the model. These are variables such as the GDP of the nation and CPI rate that have a direct impact on the movement and determination of the exchange rate. For the purpose of this analysis, a set of data was taken to form the sample. 

The rate of changes in the exchange rate between AUD & USD was taken as the dependent variable whereas other variables such as cash rate, CPI, GDP and inflation rate were considered independent variables. The independent variables have been assumed in percentage. Thereafter, all the resultant data was assessed with the help of E-view program which involved the application of different tests. The level of data significance was analyzed using various tests such as T-test, F-test and other statistical analysis while the regression model was used to construct the model. 

Regression analysis was performed repeatedly on the data set in order to help eliminate all the irrelevant factors. The best model is the one which gives less forecasting error in comparison to other models. 

Conclusion

This report makes a critical analysis of the currency carry trade strategy employed in the financial world to leverage the high interest rate environment prevailing in a foreign market. As it is believed by certain financial experts that carry trade does not result in any significant gain since the return earned in a foreign currency is eroded through the interest rate parity. A critical analysis of this notion has been made with the help a regression model which predicts the exchange rate between AUD/USD at some time in future. For that a single equation has been derived and modelled with cash rate and interest rate as the independent variables with the exchange rate being the dependent variable. The results of the regression analysis gave the important parameters which needs to be used for a successful carry trade. Two relevant theoretical concepts have also been discussed in detail in this article namely Relative Purchasing Power Parity and Interest Rate Parity which form the basis of currency carry trade. A thorough evaluation of all these factors covered in this report has helped in establishing a successful trading strategy. This work can act as a guiding material for any investor who wants to take advantage of high interest rate environment in the emerging markets through currency carry trade.  

Reference Lists

Australian Financial Review 2016, ‘Australia dollar slammed by inflation outlook, bonds surge’, market, currency, viewed at 12 May 2016.

Engel, C 2015, ‘EXCHANGE RATES, INTEREST RATES, AND THE RISK PREMIUM’.

University of Wisconsin, NBER Working Paper No. 21042, viewed at 12 May 2016.

Scutt, D 2016, ‘Australia’s trade deficit was larger than anyone thought in February’, Business Insider Australia, Money & Market, viewed at 12 May 2016.