Best Currency Positions For December 2017

|
Includes: AYT, CEW, CNY, CYB, DBMX, DRR, EIS, EPOL, ERO, ERUS, EUFX, EUO, EWW, FXC, FXCH, FXE, FXY, HEWW, INR, ISL, ISRA, ITEQ, JEM, JYN, MEXX, MXE, MXF, PGD, PLND, RSX, RUSL, RUSS, UDN, ULE, URR, USDU, UUP, YCL, YCS
by: Ruerd Heeg
Summary

Some currencies went up against the USD, others went down. Most European currencies were up, petro-currencies went down and emerging market currencies were mostly up.

I have added 4 currencies to my quantitative analysis: TRY, CNY, INR, IDR.

Overvalued: ILS, PLN, CZK and EUR. Undervalued: CNY, RUB, JPY, MXN and CAD.

Also quantitatively undervalued but without my qualitative comments: IDR and INR.

I invest based on statistics. For more on investing in statistically undervalued stocks see here. But you can also short overvalued currencies and use the proceeds to buy undervalued currencies. To keep track of currency valuations I publish articles like this one every month. See for example my article from last month and this article. This month I added 4 currencies to my quantitative analysis: the Chinese yuan, the Indian rupee, the Indonesian rupiah and the Turkish lira.

In my qualitative discussion of longs and shorts I will focus on currencies that can directly be traded with Interactive Brokers. Among the currencies in my quantitative analysis the Turkish lira, the Indian rupee, the Indonesian rupiah and the Brazilian real cannot be traded with Interactive Brokers.

4 currency trading strategies

Like last month I value currencies based on 4 statistical currency strategies:

  1. Changes in purchasing power relative to changes in purchasing power of other currencies. In other words: suppose the 5-year difference in inflation between 2 currencies is not compensated by a 5-year decrease in the value of the currency with the most inflation. Then a long position in that inflationary currency and a short position in the other currency is a statistically favorable bet.

  2. The term spread. This is the difference between long term interest rates and short term interest rates. Currencies with inverted or flat yield curves have better returns, at least on average. I use the difference between the 10-year yield and the 1-year yield.

  3. The 1-month change in the 10-year yield. The larger this change the better the statistical return of that currency.

  4. Momentum: I use 6-month raw price momentum.

For each of these 4 basic strategies I compute for each currency a rank number. Low rank numbers predict low, or negative, returns and high rank numbers predict high (positive) returns, at least on a statistical basis. Average rank numbers are computed for 4 combinations of currency strategies. I combine the following strategies:

1. Changes in purchasing power with the term spread strategy

2. Changes in purchasing power with 1-month changes in the 10-year yield

3. Momentum with the term spread strategy

4. Momentum with 1-month changes in the 10-year yield.

From the paper Value and Momentum Everywhere we know the correlation between changes in purchasing power and momentum is low. From this yield-curve paper we know the correlation between the term spread strategy and 1-month changes in the 10-year yield is also low. Therefore it does not make sense to consider other combinations apart from these 4.

Basic currency data

First I will present basic data I have used to make the rankings. I have added data for the Brazilian real. In the table below data is presented for each currency from November 23, 2017. The arrow behind the currency symbols shows whether it went up or down compared to last month. The column price is the exchange rate relative to the USD with the USD being the base currency in the currency pair. Currencies are sorted using the term spread, which is the currency strategy with the highest Sharpe ratio. The higher the term spread the lower the statistical return. The column "Changes in purchasing power" is the difference between the left hand side and the right hand side in the second formula of this wiki-article. I compute this difference using 5-year inflation data and the 5-year change in the exchange rate. Positive differences indicate undervaluation relative to the USD while negative differences signal overvaluation.

Ranking

&

Symbol

Price

[USD.XXX]

Term
spread
(%)

Changes in

purchasing

power

1-month

Δ 10Y yield

(%)

6-month

momentum
(%)

1. TRY

3.92↓

5.41

2.50

1.15

-9.03

2. BRL

3.23↑

3.16

0.31

0.38

1.13

3. HUF

263.64↓

2.04

0.20

-0.48

4.04

4. PLN

3.55↑

2.02

0.14

-0.05

4.65

5. ZAR

13.89↓

2.01

0.39

0.51

-6.86

6. ILS

3.51↓

1.63

-0.02

-0.05

2.03

7. EUR

0.84↑

1.57

0.12

-0.09

5.88

8. SEK

8.30↑

1.46

0.28

-0.18

4.77

9. CZK

21.46↑

1.43

0.10

0.13

9.98

10. NOK

8.13↓

1.11

0.35

-0.13

2.96

11. DKK

6.28↑

0.98

0.11

-0.14

5.58

12. NZD

1.45=

0.95

0.20

-0.22

-1.98

13. HKD

7.81↓

0.92

-0.07

-0.04

-0.29

14. GBP

0.75↑

0.85

0.18

-0.08

2.77

15. AUD

1.31↓

0.81

0.33

-0.28

2.01

16. SGD

1.35↑

0.79

0.13

-0.14

3.13

17. USD

1.00

0.73

0.00

-0.06

0.00

18. KRW

1084↑

0.66

-0.02

0.08

3.56

19. CHF

0.98↑

0.63

0.11

-0.13

-0.57

20. CAD

1.27↓

0.61

0.26

-0.13

5.86

21. RUB

58.41↓

0.25

0.51

0.12

-3.29

22. JPY

111.24↑

0.21

0.34

-0.04

0.58

23. MXN

18.63↑

-0.03

0.29

0.06

-0.47

24. INR

64.58↑

-0.04

63.33

0.24

0.26

25. IDR

13527↑

-0.51

13525.80

-0.18

-1.48

26. CNY

6.58↑

-3.08

5.55

0.25

4.67

Ranks of 4 basic currency strategies

The data from the table above results in the following rankings of the 4 currency strategies. See the table below. The lower the rank number the lower the statistical return. For example, based on changes in purchasing power, the expected return of the Hong Kong dollar is lower than that of the Indonesian rupiah.

Rank

Changes in

purchasing

power

Term
spread

1-month

Δ 10Y yield

6-month

momentum

1

HKD

TRY

HUF

TRY

2

KRW

BRL

AUD

ZAR

3

ILS

HUF

NZD

RUB

4

USD

PLN

SEK

NZD

5

CZK

ZAR

IDR

IDR

6

CHF

ILS

DKK

CHF

7

DKK

EUR

SGD

MXN

8

EUR

SEK

CAD

HKD

9

SGD

CZK

NOK

USD

10

PLN

NOK

CHF

INR

11

GBP

DKK

EUR

JPY

12

NZD

NZD

GBP

BRL

13

HUF

HKD

USD

AUD

14

CAD

GBP

PLN

ILS

15

SEK

AUD

ILS

GBP

16

MXN

SGD

JPY

NOK

17

BRL

USD

HKD

SGD

18

AUD

KRW

MXN

KRW

19

JPY

CHF

KRW

HUF

20

NOK

CAD

RUB

PLN

21

ZAR

RUB

CZK

CNY

22

RUB

JPY

INR

SEK

23

TRY

MXN

CNY

DKK

24

CNY

INR

BRL

CAD

25

INR

IDR

ZAR

EUR

26

IDR

CNY

TRY

CZK

Ranks of the 4 combined currency strategies

Below are the ranks of each currency in the 4 combination strategies. Behind each currency you will find the average rank of the 2 basic currency strategies. Before computing the average I have normalized the 4 individual ranks to numbers between 0 and 1. Hence the averages are also between 0 and 1. Again the lower the average and rank number the lower the expected return of a long position.

Rank

Changes in

purchasing

power

+

Term spread

Changes in

purchasing

power +

1-month

Δ 10Y yield
spread

Momentum

+

Term spread

6-month

momentum

+

1-month

Δ 10Y yield

1

ILS

0.14

DKK

0.22

TRY

0.00

NZD

0.10

2

PLN

0.24

HUF

0.24

ZAR

0.10

IDR

0.16

3

CZK

0.24

NZD

0.26

BRL

0.24

AUD

0.26

4

HKD

0.24

SGD

0.28

NZD

0.28

CHF

0.28

5

EUR

0.26

CHF

0.28

ILS

0.36

HUF

0.36

6

HUF

0.28

USD

0.30

HKD

0.38

USD

0.40

7

DKK

0.32

ILS

0.32

HUF

0.40

RUB

0.42

8

BRL

0.34

HKD

0.32

PLN

0.44

SGD

0.44

9

KRW

0.36

EUR

0.34

RUB

0.44

NOK

0.46

10

USD

0.38

SEK

0.34

CHF

0.46

HKD

0.46

11

SEK

0.42

AUD

0.36

NOK

0.48

MXN

0.46

12

TRY

0.44

KRW

0.38

USD

0.48

SEK

0.48

13

NZD

0.44

CAD

0.40

AUD

0.52

TRY

0.50

14

SGD

0.46

GBP

0.42

GBP

0.54

ZAR

0.50

15

CHF

0.46

PLN

0.44

IDR

0.56

GBP

0.50

16

GBP

0.46

CZK

0.48

SEK

0.56

JPY

0.50

17

ZAR

0.48

NOK

0.54

MXN

0.56

ILS

0.54

18

NOK

0.56

IDR

0.58

EUR

0.60

DKK

0.54

19

AUD

0.62

MXN

0.64

SGD

0.62

INR

0.60

20

CAD

0.64

JPY

0.66

JPY

0.62

CAD

0.60

21

MXN

0.74

BRL

0.78

DKK

0.64

PLN

0.64

22

JPY

0.78

RUB

0.80

INR

0.64

EUR

0.68

23

RUB

0.82

ZAR

0.88

CZK

0.66

BRL

0.68

24

INR

0.94

INR

0.90

KRW

0.68

KRW

0.70

25

CNY

0.96

CNY

0.90

CAD

0.84

CNY

0.84

26

IDR

0.98

TRY

0.94

CNY

0.90

CZK

0.90

As you can see most currencies score bad on at least one of the 4 basic strategies. In other words: in efficient markets there is no such thing as a free lunch.

Also be aware that these are very simple models. Other, more sophisticated, models can result in totally different predictions.

Statistical shorts

I find the combination of a favorable term spread and undervalued based on 5-year changes in purchasing power the most attractive forex strategy. This is the first column in the table above. I think this strategy generates the highest returns, in the long run. I also prefer it because I think it involves the least trading. As with any other investment strategy it does not always work though. Based on this strategy good shorts could be the Israeli shekel, the euro and the Hong Kong dollar, and the 2 East European currencies PLN, CZK.

Deutsche Bank thinks the Israeli shekel is the second most overvalued currency in the world. In other words the shekel is a short. The Bank of Israel is selling as well, although only in small amounts. That it did not change its ultra-low interest rates will not further push up its currency either.

See previous articles for more information on the Hong Kong dollar. The currency is pegged to the USD. Therefore it depends on the monetary authorities to decide when to de-peg. I do not think they will do that anytime soon. Recently the Hong Kong Monetary Authority issued bills for the equivalent of 10.3 billion USD to soak up excess cash.

The situation with the Polish zloty is difficult to judge. The economy is doing fine with wages going up and inflation so far under control. See also here. What still remains is an unfavorable ranking based on term spread and purchasing power. The current value seems to be high and seems only sustainable if positive news continues.

I have not found anything really justifying the bear case. Traders in the zloty might be speculating the Polish central bank might do the same as the Czech central bank: raise interest rates more than the European central bank. If that does not happen the value of the zloty might decrease. In the mean time reported labor shortages in the region make this unlikely.

For the Czech kurona the story is similar to that of the Polish zloty. A good economy and labor shortage drives inflation up. Therefore the central bank raised interest rates again. See also here. Short-term there might not be many internal drivers for the currency to go down.

The euro is also a short but statistically not as pronounced as the other 3 currencies. The euro has gone up on good GDP news from Germany. Other euro economies such as The Netherlands, Spain and Ireland have been performing very well. The EU Commission forecasts Germany to continue growing. Therefore Germany is probably going to be another engine for European GDP growth. A failure of German coalition talks followed by a new German election could spoil the party. Also lack of progress in Brexit negotiations may weigh on the euro.

Statistical longs

Based on the combination of changes in purchasing power and the term spread the most undervalued currencies are the Chinese yuan, Russian ruble, the Japanese yen, the Mexican peso, and, to a lesser extend, the Canadian dollar. The Indian rupee and the Indonesian rupiah are also good longs but this article focuses on currencies that can be traded via Interactive Brokers.

In the same report the Deutsche Bank mentioned the Israeli shekel as the second most overvalued currency in the world the Chinese yuan was described as the most overvalued currency. For this analysis Deutsche Bank mostly relies on fundamental econometric models. These models are fed with past data. Then parameters are computed using regression. Finally these parameters are used to predict exchange rates using present data.

I am skeptical about this method since it relies on extrapolation. And extrapolation is something we think we are good at but we are not. Clearly their conclusion defies a wide range of statistics predicting high returns for the yuan. Not only the Chinese yuan is undervalued based on 5-year changes in purchasing power, but also based on the term spread, based on the monthly change in 10-year interest rate of government bonds, and based on 6-month momentum. Part of the momentum might be caused by wider use as a reserve and trade currency: see also here and here.

However what would have happened if China had not restricted capital controls? See this excellent analysis. I cannot repeat often enough the forex markets are the most manipulated markets in the world. In the mean time China prints more yuans.

Despite higher oil prices all 3 petro-currencies, the Russian ruble, the Canadian dollar and the Norwegian krone, went down since last month. I cannot explain this but here are some thoughts. Among others diversification might have kicked in and also higher interest rates might have suppressed the effect of the oil price on exchange rates. I think, however, the market is just slow to respond. If the oil price stays at this level we will see the effect in these 3 currencies within a couple of months. In the mean time, how about having expensive medical treatments in Russia?

For the ruble it did not help either that the Bank of Russia purchased foreign reserves and cut its key interest rate with 0.25%. In general I also think the current euphoria on the US stock markets is taking away money from less developed countries like China, Russia, India, Indonesia and Mexico. The outflow of speculative capital depresses their exchange rates.

The Japanese yen strengthened compared to last month. The bottom in the exchange rate was when Trump complained about trade with Japan not being free and reciprocal. The Bank of Japan kept its monetary policy with massive asset purchases unchanged.

After last month's lows the Mexican peso recovered somewhat. A recent minimum wage hike may increase inflation and decrease exports. Lack of progress in the NAFTA negotiations with the Trump government might depress the exchange rate. See also here and here for more details. Actually the Trump administration has been providing a service to the Mexicans by talking their currency down. That decreases Mexico's export prices, at least short-term. Other uncertainties are elections and the appointment of a new head of the central bank.

Like the Mexican peso also the Canadian dollar could fall if NAFTA negotiations do not succeed. I recommend reading the linked article in the last sentence also to learn more about the NAFTA negotiations. Another reason to short the loonie is reduced chances of interest rate hikes. See also here. However, this article points out rising commodity prices will push the CAD up.

Since all exchange rates are expressed in terms of the US dollar I finalize with my remarks on the USD. The USD still looks overvalued based on changes in purchasing power and momentum but not so much based on last month's changes in long term interest rate and based on the term spread. What we have seen last 3-4 months is a reversal of the irrational gains from the Trump effect. So far Trump has not taken many measures stimulating the US economy. We are still waiting for Trump's tax reform. Rising commodity prices and rising interest rates outside the US could further correct the value of the USD. Finally, the USD has been pushed up by inflows of speculative capital invested in US stocks. The share prices cannot continue to go up forever. When the broad US stock market goes down, foreign capital will be retracted depressing the value of the USD.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.