Exchange Rate Perspectives

Transcription

1 Deutsche Bank Markets Research Global Foreign Exchange FX Spot Date 27 February 214 Exchange Rate Perspectives A few good basis points Arup Pal Daniel Brehon Strategist Strategist (+91) (+1) arup.pal@db.com daniel.brehon@db.com Deutsche Bank AG/Hong Kong DISCLOSURES AND ANALYST CERTIFICATIONS ARE LOCATED IN APPENDIX 1. MICA(P) 4/4/213.

2 27 February 214 Table Of Contents Currency Forecasts...3 The Big Picture: A few good basis points... Monitors: G1 FX Valuation Monitor: Lines in the Sand G1 Capital Flows and Basic Balance Monitor Commodity Price and Currency Monitor U.S. Trade Balance Monitor Central Bank Reserves Currency Composition Monitor... 4 Page 2 Deutsche Bank AG/Hong Kong

3 27 February 214 Currency Forecasts Industrialized Countries Currency Spot 3M 6M 12M US $ Exchange Rates U.S. DB USD Index Euro EUR/USD (Fwd. Rates) Japan USD/JPY (Fwd. Rates) U.K. GBP/USD (Fwd. Rates) Canada USD/CAD (Fwd. Rates) Australia AUD/USD (Fwd. Rates) N.Z. NZD/USD (Fwd. Rates) Switzerland USD/CHF (Fwd. Rates) Euro Cross Rates Japan EUR/JPY (Fwd. Rates) U.K. EUR/GBP (Fwd. Rates) Switzerland EUR/CHF (Fwd. Rates) Norway EUR/NOK (Fwd. Rates) Sweden EUR/SEK (Fwd. Rates) Source: Datastream, Reuters, Bloomberg Finance LP, DB forecasts Latin America Currency Spot 3M 6M 12M Argentina USD/ARS (Fwd. Rates) Brazil USD/BRL (Fwd. Rates) Chile USD/CLP (Fwd. Rates) Colombia USD/COP 2,8 2,6 2,7 2,1 (Fwd. Rates) - 2,69 2,8 2,119 Mexico USD/MXN (Fwd. Rates) Asia Currency Spot 3M 6M 12M China USD/CNY (Fwd. Rates) Hong Kong USD/HKD (Fwd. Rates) India USD/INR (Fwd. Rates) Indonesia USD/IDR 11,66 11,8 11,7 11,8 (Fwd. Rates) - 11,88 12,9 12, Malaysia USD/MYR (Fwd. Rates) Philippines USD/PHP (Fwd. Rates) Singapore USD/SGD (Fwd. Rates) South Korea USD/KRW 1,69 1,7 1, 1,6 (Fwd. Rates) - 1,7 1,79 1,86 Taiwan USD/TWD (Fwd. Rates) Thailand USD/THB (Fwd. Rates) Source: Datastream, Reuters, Bloomberg Finance LP, DB forecasts Emerging Europe Currency Spot 3M 6M 12M Czech Rep. EUR/CZK (Fwd. Rates) USD/CZK (Fwd. Rates) Hungary EUR/HUF (Fwd. Rates) USD/HUF (Fwd. Rates) Poland EUR/PLN (Fwd. Rates) USD/PLN (Fwd. Rates) Russia USD/RUB (Fwd. Rates) Turkey USD/TRY (Fwd. Rates) South Africa USD/ZAR (Fwd. Rates) Source: Datastream, Reuters, Bloomberg Finance LP, DB forecasts Source: Datastream, Reuters, Bloomberg Finance LP, DB forecasts Deutsche Bank AG/Hong Kong Page 3

4 27 February 214 G1 FX Forecasts: End of Quarter Spot Q1 Q2 Q3 Q4 Q4 USD-crosses EUR/USD USD/JPY GBP/USD USD/CHF AUD/USD NZD/USD USD/CAD USD/SEK USD/NOK EUR-crosses EUR/JPY EUR/GBP EUR/CHF EUR/SEK EUR/NOK Source: Deutsche Bank Page 4 Deutsche Bank AG/Hong Kong

5 27 February 214 Currencies: A few good basis points An asset class that provides stable returns with low volatility in changing macro environments is a compelling choice for investment. Despite recent wobbles FX as an asset class has provided consistent returns over the past 3 years [Figure 1]. Equally important is the fairly stable volatility in those returns over the same period. The Deutsche Bank Currency Returns (dbcr) Index, an equally weighted FX benchmark comprised of carry, valuation and momentum strategies, has averaged 3.% annual returns and % volatility since 198 [Figure 2]. Indeed, excess returns were above 3% in the 8s, 9s and last decade. These stable returns along with low correlation to equities and bonds promote FX as a prime contender for a place in every manager s portfolio. Liquidity concerns that ordinarily deter investors from diversifying away from bonds and equities towards alternative asset classes are not a constraint for investment in FX. Daily turnover exceeding trillion USD and rising (see BIS 213 tri-annual FX survey) dispels any liquidity related reluctance. Benchmark dbcr returns have been positive in a number of market regimes over the years: dollar bull (early 9s, late 9s) and bear (late 8s, 2s) cycles, periods of active central bank intervention and freely floating currencies, eras of strong and weak global growth. Using over 3 years of data we measure the positive effects of including FX in a global portfolio and observe whether it generates a systematic beta from an unconstrained position and not from a currency exposure management position that comes with being long/short any global bond/equity. We also decompose the FX returns according to strategies and analyze their performance over time. We pick out the currencies that contributed the most for each case. Finally we try to compare how the conventional carry, valuation and momentum strategies fared over the long term against a simple moving average strategy or a relative valuation strategy. Figure 1: dbcr returns over last 3 years 2% 2% 1% 1% % % excess returns (over risk-free rate), including transaction costs % Source: Deutsche Bank Figure 2: dbcr annual volatility 12% 1% 8% 6% 4% 2% FX market volatility % Source: Deutsche Bank Average annual vol of % Our currency returns benchmark (dbcr) 1 By the very nature of exchange rates, it is not possible to perfectly mimic the rules used to create an equity or bond index. The fundamental issue is that unlike equities and bonds, where one can be outright long an equity or bond, in FX markets by definition by being long one currency one has be to short another currency. Otherwise there would be no exchange rate risk. This poses many problems when constructing an index or benchmark, not least of how one would know whether to simply buy and hold EUR/USD (i.e. long euro/short dollar) over time or USD/EUR (i.e. long dollar/short euro). Also, what would the market-capitalization of EUR/USD be at any point in time? Would it be the value of all current EUR/USD open positions (if that could be measured)? Or would it be the net EUR/USD position? And so on. Therefore, like with equity (and bond) indices, a set of broadly agreed upon (and replicable) rules need to be used to determine the market return. The four principles we chose when constructing the dbcr Index were the following: 1. Developed world currencies - to meet liquidity requirements 2. Positive net yield (or carry) similar to bond indices 3. Positive momentum similar to market cap weighting and minimum capitalization rules 4. Undervalued similar to incorporating a fundamental metric such as earnings or revenue often used in equity indices. 1 See Hafeez (27) Currencies: Portfolio Saviour? and Benchmarking Currencies: The DBCR Index. Deutsche Bank AG/Hong Kong Page

6 27 February 214 Carry, momentum and valuation are the three most widely used FX investment styles. They form the core investment approach of many FX-oriented funds and are supported by decades of academic work. For the positive net yield, or carry trade, we simply buy the three highest yielding currencies and sell the three lowest yielding currencies based on short-term yields. This strategy exploits the widely documented forward rate bias that exists in FX markets. For momentum, we use a ranking rule based on annual spot returns, where one buys the three currencies with the highest return against the USD over the last twelve months, and sells the three currencies with the lowest returns. Finally, for valuation we use Purchasing Power Parity, the most basic valuation metric for currencies, which as a concept, has been around for over one hundred years. The trading rule we employ is to systematically buy the three most under-valued currencies and sell the three most over-valued currencies. In this way, though one may arrive at lower returns, one gets around the issue of picking over- and under-valuation extremes such as +/-2%, which could run the risk of ex-post optimization. Of course, the choice of top three/bottom three could also be liable to over-fitting, though results are broadly the same whichever number of rankings is used. We created our benchmark by simply taking the average of the three strategies/rules. Importantly, the returns of our synthetic FX returns index are well correlated to the actual performance of FX funds using currency manager indices. Comparing with Equities and Bonds Comparing our FX benchmark (live since 27) to bonds and equities suggest that FX has performed at par with the other asset classes over the last 2 years although it has outperformed bonds in the long run [Figure 3]. As seen in Figure 4, the Sharpe ratio for FX has been higher than both bonds and equities in the long run. In other words, FX risk-adjusted returns have outperformed equities or bonds since 198. For a consistency check we calculate annual Sharpe ratios for each asset class and plot their distribution over time. FX turns out to be a consistent performer with risk-rewards having a fat tail on the positive side [Figure ]. Figure 3: FX returns comparable to bonds and equities 16% 14% 12% 1% 8% 6% 4% 2% % Annualised total returns of benchmark* including interest or dividends if appropriate 8% 6% 4.% 9% 7% 1.7% 1.8% 8% 7%.%.6% 198-Jan Jan Jan Jan % 4.3% 1% 7% 3.4% Worst FX Combined FX Bonds Equities Source: Deutsche Bank; MSCI World for Equity, Lehman Aggregate for Bonds and DBCR for FX, worst FX is FX momentum Figure 4: FX Sharpe ratio superior over the long run Sharpe ratio Jan Jan Jan Jan 214 Worst FX FX Bonds Equities Source: Deutsche Bank; MSCI World for Equity, Lehman Aggregate for Bonds and DBCR for FX, worst FX is FX momentum Figure : Histogram of Annual Sharpe ratios to -1-1 to to 1 1 to 2 2 to 3 3 to 4 Source: Deutsche Bank; MSCI World for Equity, Lehman Aggregate for Bonds and DBCR for FX Equities Bonds Combined FX Page 6 Deutsche Bank AG/Hong Kong

7 27 February 214 In addition the correlation of FX returns to bonds and equities has been low historically. As the bolded numbers in Figure 6 show, the individual baseline strategies that constitute the benchmark have not had large correlations to bonds and equities over the long term (although carry vs. equity correlation has been rather high since 28). One of the reasons for dbcr s outperformance is that the underlying three strategies have low correlation to each other thereby providing diversification in the overall benchmark. Compelling risk-reward combined with low correlation of returns to other asset classes suggest a role for FX in a well diversified cross-asset portfolio. How can we assess what the proper proportion should be? Breaking down the benchmark Figure 6: Correlation of FX strategies against Equities and Bonds Bond Equity DBCR Carry Mom. Valuat. Bond 1% 24% -2% -14% 1% -24% Equity 1% 1% 24% -7% % DBCR 1% 66% 42% 8% Carry 1% -9% 22% Momentum 1% -22% Valuation 1% Source: Deutsche Bank, DBCR is the benchmark giving equal weight to carry, momentum and value The FX benchmark gives equal weight to carry, momentum and valuation strategies. Figure 7 shows that each strategy performs (approximately) equal on an absolute return basis historically, although risk-adjusted returns are slightly lower for momentum. The diversification resulting from low strategy correlation produces a higher Sharpe for DBCR [combined FX in Figure 8] in comparison to each baseline strategy. Figure 7: Average annual return of the baseline strategies in the FX benchmark 14% 12% 1% 8% 6% 4% 2% % Average annual total returns 9% 7%.7% 3.7% 1% 8% 6.8%.1% 8% 6% 4.% 198-Jan Jan Jan Jan 214 9% 7%.% 1.7% 1.8% Valuation Carry Momentum Combined Source: Deutsche Bank; Combined gives equal weight to each individual baseline strategy Figure 8: Sharpe ratios of the baseline strategies Sharpe ratio Valuation Carry Momentum Combined Jan Jan Jan Jan 214 Source: Deutsche Bank; Combined gives equal weight to each individual baseline strategy We further decompose each baseline strategy into their components [currencies against the dollar] and study their performance. Except for NOK and CHF, all currencies have a positive contribution to the benchmark over the long run with NZD, AUD, GBP and JPY contributing most. Return contribution of each currency is presented at the end of the report. Figure 9: Currency specific performance of each baseline strategy Sharpe Ratios; Jan 214 Valuation Carry -.3 NZD AUD GBP JPY CAD SEK EUR NOK CHF Source: Deutsche Bank; all pairs against the US dollar, valuation is PPP based, carry is net positive yield ranked top 3 long bottom 3 short Deutsche Bank AG/Hong Kong Page 7

8 27 February 214 Robustness of our momentum strategy Figure 9 makes evident that both valuation and carry perform consistently across currencies. To check whether our trailing return based momentum strategy is in essence capturing true FX momentum we compare it against a simple moving average (SMA) based strategy for all currencies. The strategy signal is derived by subtracting the 2 day simple moving average from the day simple moving average. We go long (short) the currency if the signal is positive (negative). Including a transaction cost of.% for every buy/sell we find that the SMA strategy performs slightly better across currencies although the difference is almost completely explained by the high transaction costs for the trailing return based momentum strategy. The returns have an 8 percent correlation and are therefore capturing the same phenonmenon [Figure 11]. Figure 1: Benchmark momentum v/s Simple Moving Average strategy 6% % 4% 3% 2% 1% Trailing return based momentum Average annual returns; Jan 214 Figure 11: Simple Moving Average same as trailing return based momentum Index based on the returns of the SMA strategy; averaged across curerncies % -1% -2% -3% Simple moving average Strategy AUD EUR JPY GBP CHF CAD NZD NOK SEK 12 Index on the returns of trailing return based momentum 1 strategy; averaged across currencies Source: Deutsche Bank; all crosses are against the US dollar Source: Deutsche Bank; individual strategy returns are averaged across currencies to get the total return for each index How much FX exposure is appropriate? The question of what proportion of a fund is to be allocated to an asset class is non-trivial. In-sample optimizations run the risk of the future being different than the past. Even proper out-of-sample backtesting can produce different results as one varies the time window or the definition of risk/reward. In the end the decision parameters boil down to a manager s perception and risk appetite. Both of these factors are affected by historical performance and so we perform a simple mean-variance historical analysis to determine the optimal mix of FX, bonds and equities Using MSCI World for global equities, the Barclays Aggregate bond index and dbcr as the primary asset classes, we find that as we increase FX allocation, the portfolio as a whole starts provides a better Sharpe ratio. Besides the extra return, increasing FX allocation also reduces the magnitude and duration of a drawdown [Figure 12]. In light of these findings we suggest a FX exposure equal in magnitude to bonds and equities. A 3% exposure helps reduce overall portfolio volatility without diluting returns. Figure 12: Summary portfolio statistics Allocation to FX* % % 1% 2% 3% 4% Using FX Combined Entire sample 9.3% 9.3% 9.3% 9.3% 9.3% 9.3% Best -year period 23% 23% 22% 21% 19% 18% Worst -year period % % 1% 1% 2% 2% Sharpe ratio Max drawdown -29% -28% -27% -24% -21% -17% Duration of max underperformance** Using FX Momentum Entire sample 9% 9% 9% 9% 9% 9% Best -year period 23% 23% 22% 21% 2% 19% Worst -year period % % 1% 2% 2% 2% Sharpe ratio Max drawdown -29% -27% -2% -2% -1% -1% Duration of max underperformance** * Assuming equal split in bonds and equities ** Number of years before previous peak is reached Source: Deutsche Bank Annualised Returns Annualised Returns Page 8 Deutsche Bank AG/Hong Kong

9 27 February 214 Money left on the table and currency manager returns 2 FX markets are perceived to be a zero-sum game but past performance suggests profit-seekers (hedge funds, money market funds and other investors) in the market have been consistently earning excess returns over liquidity-seekers (such as official and corporate end-users) [Figure 13]. Performance varies across currencies and time but overall non-commercial IMM positions appear to have made money at the expense of commercial (hedger) positions in every year since 1993 [Figure 13]. This anomaly is explained not by refuting the assumptions of an efficient market and more by differentiating the objectives and beliefs of different FX market participants. Looking at IMM net speculator positions to gauge the direction of spot we find that even with a 3 day lag (because of the data release of Friday corresponding to Tuesday s positioning) we are able to hit the correct direction of trade more often than not. Despite the absence of strong academic support to dethrone random walks as the most accurate descriptor of FX movements, IMM noncommercial position returns are not random. [Figure 14] Volatility has played a big role in determining the level of profits every year (which we will look in detail later), but after accounting for the varying participation from speculators over time our findings suggest that the average amount of money made has been stable in the past but have slightly underperformed since 29. A little background on our calculation of profits will help elucidate the observations we are going to make. Every Friday we get speculator s net long position on standard CME futures corresponding to the previous Tuesday. We can calculate the net profit from these positions and that of the week before. The only unobservable in this calculation is the entry point of the new positions that were entered over the past week. Keeping things simple we take the base case to the when this entry point is the average weekly price of the currency. Figure 1 shows the annual USD returns for the base case while Figure 16 disaggregates this performance through time. Figure 13:IMM non-commercial positions have made money every year since 1993 Annual P&L; 4 in billions of USD Non-commercial / Profit-seekers Commercial / Liquidity-seekers Source: Deutsche Bank Figure 14: Non-commercial position hit ratios applying a 3-day lag on IMM positions Source: Deutsche Bank; IMM positioning data Figure 1: Base Case profits broken into currency Aggregate speculator profits*; in billions of USD JPY EUR AUD GBP CHF CAD NZD *Aggregates since 1993 for AUD,GBP,CAD,JPY,CHF,EUR(since 1999) and NZD(since 24); al currencies against the USD Source: Deutsche Bank; IMM positioning Figure 16:Speculator returns stable despite time-varying participation in FX markets Average per contract return * (in thousands USD) stable over last 2 years AUD GBP CAD JPY CHF * return calculated using the base case assumption of average weekly price to be the entry point of new positions Source: Deutsche Bank; IMM positioning 2 For the original analysis of IMM non-commercial position returns, see Kearns and Manners (24), The profitability of Speculators in Currency Futures Markets, Reserve Bank of Australia. Our follow-up note replicated the results: Hafeez (27), Currency Markets: Is Money Left On the Table? Deutsche Bank AG/Hong Kong Page 9

10 27 February 214 Effect of Volatility and Trade-timing on Currency returns Generally a higher level of volatility should provide more opportunities for extracting returns from a market and this has been the case for FX too. Considering the base case for calculating profits, we see that the correlation between yearly profits and yearly volatility has been positive for all currencies. [Figure 17] But high volatility makes choosing entry points for a trade a nontrivial exercise. Therefore we need to account for the fact that bad timing might impair the spectacular returns we were getting in the base case. We remain pessimistic with our estimation of profits and introduce two price points x* low and x* high to be our entry points for the new positions taken in a week. We use the former when new positions taken are short and the latter when long. A lower (higher) value of x* low (x* high) essentially tells that we are selling (buying) at weekly lows (highs).the buy low, sell high reduces profits up-to the point when x* low is equal to the week s low and x* high equal to the week s high. To get the values of x* low (x* high) we subtract (add) from (to) the average weekly price, the weekly price range [max min] divided by a parameter that we vary. A parameter value of infinity would correspond to the base case as the amount subtracted would become zero. A high volatility increases the weekly range and shifts the x* high and x* low away from the weekly average and therefore diminishes our profits. As an example a parameter value of 4 reduces aggregate profits by 1 million USD per year on average. [Figure 18] 3 Figure 17: Higher Volatility increases opportunities for higher returns 1. Correlation;.9 Yearly profit vs Yearly Volatility AUD GBP EUR CAD NZD CHF JPY Source: Deutsche Ban; IMM positioning Figure 18: Higher volatility impairs market timing and reduces profits P Profit Llost tddue tto hi volatility; h l in millions of USD Average Loss CHF GBP JPY AUD CAD Source: Deutsche Bank; IMM positioning Next we compare the returns we generate from publicly available IMM positioning data with conventional currency managers indices. We chose the Stark Currency Trader Index (SCTI) for our analysis. 4 With our framework we try to estimate under what market entry conditions and leverage our returns comply with those of SCTI. 3 The parameter 4 roughly corresponds to a weekly volatility of one standard deviation if prices are assumed to be normally distributed 4 We chose to disregard the Parker Index because the pre-crisis Sharpe ratio of 2. for the Parker Index proved impossible to replicate with basic trading strategies. We do note that Parker and Stark are highly correlated to each other and our IMM position replication strategy but Parker s unusually strong pre-crisis performance stands out as an anomaly. Page 1 Deutsche Bank AG/Hong Kong

11 27 February 214 We construct an IMM positioning based index whose week-over-week returns are equal to the aggregate weekly profits from trading AUD, CAD, JPY, GBP, CHF, NZD and EUR futures divided by the capital invested. Estimating the invested amount is no easy task as it depends on the average amount of leverage put up by currency managers. Managers may even choose to be under-invested at times thereby complicating the return on investment (RoI) calculation. We assume the invested capital to be a constant which we vary in order to match the volatility of the SCTI returns. We consider two time windows, 23-8 and 29-present for our analysis. This structural regime break allows us to study the changes in currency trading behavior pre- and post-crisis. The SCTI had a cumulative return of 16.4% from 23 to 28 and -1.8% from 29 to date. The annualized volatilities of the SCTI returns over the same periods were.7% and 6.4%. We calculate profits using the base case (parameter value of infinity used for calculating x* low and x* high) and divide by capital invested (a constant) to get weekly returns. To match the levels of volatility of our index with that of SCTI we settle with a 2 billion USD capital as the invested amount. Next we change the entry points of our trade by varying the parameter (used to calculate x* of and x* high) to match the cumulative return over the two windows of our index and SCTI. A surprising result from our simple modeling is that trade timing has to be significantly poorer in the 29-present window than in the window for the returns to be comparable. In other words, managers may have been buying high and selling low intra-week while accumulating positions, possibly due to greater headline risk and tighter risk management in the post-crisis period. A parameter value of 4.3 works for the earlier regime (23-28) which is reasonably worse-off because it says that on average managers went long (short) at the weekly average plus (minus) one standard deviation of weekly prices. But our parameter must fall to 2. in order to produce negative returns similar to that of SCTI over the present regime (29-date). Figure 19 & 2 shows how the two return streams look against each other with the above parameter values and capital invested. Figure 19: IMM positioning based index against Stark Currency Trader Index Stark Currency Trader Index IMM positioning-based index * * Capital Invested is 2 billion USD, market-timing parameter is 4.3 for and 2. for 29-present, 2% management fee. Source: Deutsche Bank; IMM positioning Figure 2: Post-crisis currency returns seem to indicate poor market-timing Stark Currency Trader Index IMM positioning-based index * * Capital Invested is 2 billion USD, market-timing parameter is 4.3 for and 2. for 29-present, 2% management fee. Source: Deutsche Bank; IMM positioning Our assumptions are quite broad and poor intra-week market timing is but one possible explanation for sub-par post-crisis manager returns. We look for more market timing evidence by comparing SCTI returns with currency returns broken into momentum, value and carry as mentioned before. Unsurprisingly we find that SCTI returns are better correlated to that of momentum than both valuation and carry over both periods. Valuation and carry portfolio compositions change with lower frequency and therefore their performances should be less vulnerable to poor market timing. By contrast, a momentum portfolio is intended to catch short-run trends and unfavorable entry points can result in losses even if directional accuracy is high [Sell low and buy high can lose money despite a underlying strong trend]. The present underperformance of SCTI closely matches the poor returns in momentum trades [Figure 21] since average volatility [CVIX] in the second period has been two percentage points above that of the first period leading to poor market timing. With current FX volatility below crisis levels, market timing is becoming less of a concern now. Stronger trends emerging from global market normalization should present a better trading environment for profit seekers in FX markets. Deutsche Bank AG/Hong Kong Page 11

12 27 February 214 Figure 21: Poor momentum returns from poor timing have eaten into currency manager s overall returns Momentum Index Valuation Index Stark Currency Trader Index momentum underperformance closely resemble Stark Currency Trader Index since Carry Index Source: Deutsche Bank,Bloomberg, IMM positioning, DBMOMUSF for momentum, DBPPPUSF for valuation and DBBHVBUSI for carry Conclusion: FX returns have low correlation to bonds and equity returns besides being favorable on risk-adjusted terms in the long run. FX may be viewed as an asset class that provides an additional source of positive beta irrespective of one s bond or equity exposure. We suggest allocation of a significant (3%) proportion to FX in a global portfolio to increase the overall performance through diversification. The out-performance of currency-dedicated funds over the long run further encourages allocation towards foreign exchange strategies. Our concern for the future is whether FX liquidity seekers will continue to dominate flows in the future as they have in the past, allowing profitable opportunities for profit seekers to persist. Looking at the latest BIS survey on FX turnover, we believe this trend is going to continue as non-dealing small banks together with nonfinancial customers account for 33% of the daily activity surpassing that of speculators 22% [Figure 22]. With global trade, international M&A and portfolio investment expanding every year, the share of liquidity-seekers among the FX market players will likely remain stable, leaving plenty of money on the table for FX investors. Figure 22: FX activity by counterparty Hedge Funds,PTFs, Institutional Investors Non-financial Customers and Non-dealing Banks 22% Others 7% 33% Source: Deutsche Bank; BIS April-214 Survey 39% Dealers Page 12 Deutsche Bank AG/Hong Kong

13 27 February 214 Strategy specific currency returns For baseline strategies, the total return is the sum of all the returns when a currency was included in the strategy. Tr. Ret is the trailing return based momentum strategy and the Simple MA is the simple moving average strategy. Figure 23: AUD against USD returns Figure 24: EUR against USD returns Figure 2: JPY against USD returns 12% 1% 8% 6% 4% 2% % Jan 214 Total returns 24 - Jan 214 baseline Tr Ret Simple baseline baseline 8% 6% 4% 2% % -2% -4% -6% Jan 214 Total returns 24 - Jan 214 Simple baseline Tr Ret baseline baseline 14% 12% 1% 8% 6% 4% 2% % -2% Jan 214 Total returns 24 - Jan 214 Tr. Ret Simple baseline baseline baseline carry mom MA Value mom MA carry mom value mom mom MA value carry mom Source: Deutsche Bank; DBHVBUSI for carry; DBMOMUSF for momentum and DBPPPUSF for value Source: Deutsche Bank; DBHVBUSI for carry; DBMOMUSF for momentum and DBPPPUSF for value Source: Deutsche Bank; DBHVBUSI for carry; DBMOMUSF for momentum and DBPPPUSF for value Figure 26: GBP against USD returns Figure 27: CHF against USD returns Figure 28: CAD against USD returns 6% 4% 2% % -2% -4% -6% -8% Jan 214 Total returns 24 - Jan 214 baseline baseline Simple baseline Tr. Ret 1% 8% 6% 4% 2% % -2% -4% -6% Jan 214 Total returns 24 - Jan 214 Tr. Ret Simple baseline baseline baseline % 4% 3% 2% 1% % -1% -2% -3% -4% % Jan 214 Total returns 24 - Jan 214 Simple baseline baseline Tr. Ret baseline value carry MA mom mom mom MA mom carry value MA value carry mom mom Source: Deutsche Bank; DBHVBUSI for carry; DBMOMUSF for momentum and DBPPPUSF for value Source: Deutsche Bank; DBHVBUSI for carry; DBMOMUSF for momentum and DBPPPUSF for value Source: Deutsche Bank; DBHVBUSI for carry; DBMOMUSF for momentum and DBPPPUSF for value Figure 29: NZD against USD returns Figure 3: NOK against USD returns Figure 31: SEK against USD returns 14% Jan 214 Total returns 6% Jan 214 Total returns 1% Jan 214 Total returns 12% 1% 8% 6% 4% 2% % 24 - Jan 214 Simple baseline Tr. Ret baseline baseline 4% 2% % -2% -4% -6% 24 - Jan 214 baseline Tr. Ret Simple baseline baseline 1% % % % -1% 24 - Jan 214 Simple baseline Tr. Ret baseline baseline MA carry mom value mom carry mom MA mom value MA value mom carry mom Source: Deutsche Bank; DBHVBUSI for carry; DBMOMUSF for momentum and DBPPPUSF for value Source: Deutsche Bank; DBHVBUSI for carry; DBMOMUSF for momentum and DBPPPUSF for value Source: Deutsche Bank; DBHVBUSI for carry; DBMOMUSF for momentum and DBPPPUSF for value Deutsche Bank AG/Hong Kong Page 13

14 27 February 214 G1 FX Valuation Monitor: Lines in the Sand* Figure 1: The euro is expensive and the dollar cheap Figure 2: The dollar is 8% cheap to fair value Figure 3: EUR/USD: The euro is expensive though remains within the 2% threshold Figure 4: USD/JPY: The yen is very cheap to fair value EUR/USD PPP EUR/USD 2% Band % Band USD/JPY PPP USD/JPY Figure : USD/GBP: GBP is expensive Figure 6: USD/CHF: as well as CHF % Band USD/CHF PPP USD/CHF % Band USD/GBP PPP USD/GBP *Our measure of relative PPP is calculated using long-term averages from Jan-8 to Dec-4 and deflating by monthly CPI differentials. We refer to current spot rates as "cheap" or "expensive" with explicit reference to this measure of fair valuation; these statements are not intended in any way to be "buy" or "sell" recommendations. Page 14 Deutsche Bank AG/Hong Kong

15 27 February 214 Figure 7: USD/CAD: CAD overvaluation is being unwound Figure 8: USD/AUD: AUD is very expensive, beyond 2% threshold % Band 1. USD/CAD.9 PPP USD/CAD USD/AUD.9 2% Band PPP USD/AUD Figure 9: USD/NZD:.and so is NZD Figure 1: EUR/JPY: The euro is very expensive against the yen EUR/JPY 2% Band PPP EUR/JPY USD/NZD 1. 2% Band PPP USD/NZD Figure 11: EUR/GBP: Sterling is cheap against the euro Figure 12: EUR/SEK: SEK is very cheap versus the euro EUR/GBP.4 2% Band PPP EUR/GBP EUR/SEK 2% Band 4 PPP EUR/SEK Deutsche Bank AG/Hong Kong Page 1

16 27 February 214 Figure 13: EUR/CHF: CHF is expensive against the euro Figure 14: EUR/CAD: CAD is cheap against euro EUR/CHF % Band PPP EUR/CHF EUR/CAD 2% Band.9 PPP EUR/CAD Figure 1: AUD/NZD: NZD is fair value against AUD. Figure 16: CAD/NZD:.and is expensive against CAD AUD/NZD.8 2% Band.6 PPP AUD/NZD CAD/NZD 2% band.6 PPP CAD/NZD Figure 17: JPY/NZD: NZD is expensive against the yen Figure 18: GBP/JPY: JPY is very cheap against GBP JPY/NZD 2% Band PPPJPY/NZD GBP/JPY 2% Band PPP GBP/JPY Page 16 Deutsche Bank AG/Hong Kong

17 27 February 214 FX Behavioral and Fundamental Equilibrium Exchange Rates (BEER and FEER)* Figure 1: USD-cross BEER and FEER valuations Figure 2: EUR/USD is expensive vs. BEER FV Figure 3: USD/JPY is now above fair value vs. BEER FV Figure 4: GBP/USD is very undervalued vs BEER FV Figure : USD BIS TWI is cheap vs. BEER FV *Sources: BIS, Bloomberg Finance LP, Deutsche Bank. Notes: For details on model, see Exchange Rate Perspectives, Jan-13. BEER model is relative PPP adjusted for terms-of-trade and productivity effects. Relative FEER model is based on current account surpluses/deficits relative to long-term (structural) surpluses/deficits. Over/undervaluation calculated off TWIs and converted to USD-crosses using matrix algebra. EM graphs available upon request. Deutsche Bank AG/Hong Kong Page 17

18 27 February 214 Figure 6: USD/CAD is cheap vs. BEER FV Figure 7: AUD/USD is quite expensive vs. BEER FV Figure 8: NZD/USD is very expensive vs. BEER FV Figure 9: USD/CHF is quite cheap vs. BEER FV Source: Deutsche Bank Figure 1: USD/NOK is a bit expensive vs. BEER FV Figure 11: USD/SEK is expensive vs. BEER FV Page 18 Deutsche Bank AG/Hong Kong

19 27 February 214 Figure 12: EUR/USD is cheap vs. FEER FV Figure 13: USD/JPY is cheap vs. FEER FV Figure 14: GBP/USD is expensive vs. FEER FV Figure 1: USD BIS TWI is very cheap vs. FEER FV Figure 16: USD/CAD is quite cheap vs. FEER FV Figure 17: AUD/USD is fair value vs. FEER FV Deutsche Bank AG/Hong Kong Page 19

20 27 February 214 Figure 18: NZD/USD is expensive vs. FEER FV Figure 19: USD/ CHF is bit expensive vs. FEER FV Figure 2: USD/NOK is fair value vs. FEER FV Figure 21: USD/SEK is slightly cheap vs. FEER FV Page 2 Deutsche Bank AG/Hong Kong

21 27 February 214 G1 Capital Flows and Basic Balance Monitor United States (USD bn) Figure 1: The basic balance is worsening since 213 Figure 2: as non-treasury portfolio flows have turned negative USD Bn. Net foreign US inflows(excl. US Treasuries) minus the trade balance (bn, 12m accumalated ) Basic Balance USD TWI (Rs) Ln Trade Balance Net non-treasury portfolio flows Basic Balance and US Treasury and US Treasury Figure 3: The private basis balance is in better shape than the overall balance Figure 4: Official inflows to agencies and corporate bonds have surged since 29 2 USD Bn Private Basis Balance Overall Basic Balance USD Bn. Net Official Purchases of Corporate Bonds Net Official Purchases of Agency Bonds (Rs) and US Treasury and US Treasury Figure : Private non Treasury flows are at record lows Figure 6: Relative to the private basic balance, the dollar is expensive 6 USD Bn. 4 Net Official Flows 3 Net Private non Treasury Flows USD Bn. Ln Private Basic Balance USTW$ (ls) and Haver Analytics and US Treasury Deutsche Bank AG/Hong Kong Page 21

22 27 February 214 Figure 7: FDI outflows continue Figure 8: Portfolio flows hover around record lows th Qtr. sum FDI Inflows FDI Net flows FDI Outflows mo. Sum US Net Capital Inflows Net Bonds Inflows Net Equity Inflows and US Treasury and US Treasury Figure 9: Official sector buying of US bonds is now almost equal to private buying Figure 1: Treasury purchase by private sector has fallen substantially mo. Sum Net Private Bond Flows Net Official Bond Flows mo. Sum 6 4 Private Treasury Purchases Private Agency Private Corp US Net Purchases of Foreign Bonds and US Treasury and US Treasury Figure 11: USD TWI and UST purchases are improving from their lows Figure 12: Net equity flows continue to fall 6 4 Treasuries (ls) USD TWI (Major, rs) Ln USD bn US Stocks Foreign Stock Net Stocks and US Treasury and US Treasury Page 22 Deutsche Bank AG/Hong Kong

23 27 February 214 Figure 13: Equity outflows from the US continue in spite of impressive US equity performance Figure 14: Net equity flows fallen significantly in last one year US Stocks Foreign Stocks -17 S&P (Rs) Ln Net Equity Flows (12M Sum) USD TWI Major (Rs) Ln Correlation=% -2 Correlation(6 mo lag)=8% Source: Deutsche Bank, US Treasury and Bloomberg Finance LP Source: Deutsche Bank, US Treasury and Bloomberg Finance LP Figure 1: Foreign interest in USTs versus US equities is almost same Figure 16: The agency & corp bond inflows is close to zero 6 46 Foreign Purchases of US Treasuries Foreign Purchases of US Equity Corp and Agency Bonds(ls) Corporate bonds(ls) US TWI (Major,rs) log and US Treasury Source: Deutsche Bank, US Treasury and Bloomberg Finance LP Deutsche Bank AG/Hong Kong Page 23

24 27 February 214 Canada (CAD bn) Figure 17: The basic balance is showing some sign of improvement Figure 18: though net FDI flows remain outbound 8 6 Basic Balance 4Q CAD/USD - Quarterly Average (rs) Dec-9 Dec-94 Dec-98 Dec-2 Dec-6 Dec Net FDI Outward Direct Investments Inward Direct Investments -4 Mar-9 Mar-94 Mar-98 Mar-2 Mar-6 Mar and Haver and Haver Figure 19: Portfolio inflows is falling ever since its peak in the late 29 Figure 2: as foreigners appetite for Canadian securities continue to fall 1 1 Portfolio Flows 12-month Basic Balance 4Q rolling CAD/USD - Quarterly Average (rs) -1 Jun-91 Jun-9 Jun-99 Jun-3 Jun-7 Jun Canadian Flows Abroad Foreign Inflows Net Flows -1 Jan-91 Jan-94 Jan-97 Jan- Jan-3 Jan-6 Jan-9 Jan and Haver and Haver Figure 21: not withstanding net inflows in the equity market Figure 22: the debt market is witnessing steep fall in foreign interest Canadian Flows Abroad Foreign Inflows Net Equity Flows Canadian Flows Abroad Foreign Inflows Net Debt Flows Jan-91 Jan-94 Jan-97 Jan- Jan-3 Jan-6 Jan-9 Jan Jan-91 Jan-94 Jan-97 Jan- Jan-3 Jan-6 Jan-9 Jan-12-7 and Haver and Haver Page 24 Deutsche Bank AG/Hong Kong

25 27 February 214 Japan (JPY trillion) Figure 23: The negative basic balance has been accelerating sharply Figure 24: due to huge net FDI outflows Basic Balance 12M sum -1 USD/JPY (right scale, inverted) Jan-91 Jan-94 Jan-97 Jan- Jan-3 Jan-6 Jan-9 Jan Net FDI Direct Investment in Japan by Foreigners -14 Direct Investment Abroad by Residents Jan-91 Jan-94 Jan-97 Jan- Jan-3 Jan-6 Jan-9 Jan , MOF, and Haver and MOF Figure 2: Net capital inflows have seen a recent surge Figure 26: though bond flows remain outbound mo. sum -2-2 Net Equity Flows -3 Net Bond & Notes Flows -3 Net Money Market Flows -4 Net Capital Flows mo. sum Japan Flows -2 Foreign Flows -3 Net Bonds & Notes Flows and MOF and MOF Figure 27: as equity inflows surge to record highs Figure 28: and money market flows remain positive mo. sum Japan Flows Foreign Flows Net Equity Flows mo. sum Japan Flows Foreign Flows Net Money Market Flows and MOF and MOF Deutsche Bank AG/Hong Kong Page 2

26 27 February 214 United Kingdom (GBP bn) Figure 29: The basic balance remains negative Figure 3: although net FDI flows have entered positive zone Basic Balance 4Q sum GBP/USD (right scale, quarterly average) Net FDI FDI Outflows 1 FDI Inflows Mar-91Mar-94Mar-97Mar-Mar-3Mar-6Mar-9Mar Mar-91Mar-94Mar-97Mar-Mar-3Mar-6Mar-9Mar and Haver and Haver Figure 31: Portfolio flows are recovering from their lows Figure 32: Net debt IIP turned positive, while equity remains negative Net Portfolio flows Portfolio Outflow Portfolio Inflow -3 Mar-91Mar-94Mar-97Mar-Mar-3Mar-6Mar-9Mar Net Equity IIP Net Debt IIP GBP/USD (rs) Mar-9 Mar-93 Mar-96 Mar-99 Mar-2 MarMar-8 Mar and Haver and BoE Figure 33: Net holdings of equities Figure 34: Net debt holdings Net UK Equity Holdings by Foreigners Net Equity IIP Net Foreign Equity Holdings by the UK Net UK Debt Holdings by Foreigners Net Debt IIP Net Foreign Debt Holdings by the UK Mar-9Mar-93Mar-96Mar-99Mar-2MarMar-8Mar Mar-9 Mar-93 Mar-96 Mar-99 Mar-2 Mar Mar-8 Mar and BoE and BoE Page 26 Deutsche Bank AG/Hong Kong

27 27 February 214 Euro area (EUR bn) Figure 3: The positive basic balance remains intact Figure 36: as current account surplus outweighs the FDI outflows Basic Balance 12-month USD/EUR (rs) Feb-1 Feb-3 Feb Feb-7 Feb-9 Feb-11 Feb Current account 12-month Net FDI 12 month Jan-1 Jan-3 Jan Jan-7 Jan-9 Jan-11 Jan and Eurostat and Eurostat Figure 37: EUR/USD strongly correlated (.88) with bilateral basic balance with the US Figure 38: Bilateral basic balance explains 84% of EUR/USD movements since inception of the euro USD Bilateral Basic Balance,lagged (6m) EURUSD(rhs) EURUSD(t)= -.36*Bilateral Basic Balance(t- 6) Sample 1999M6-27M4 Observation:97 Adjusted R-squared= Correlation= Jan-99 Jan-1 Jan-3 Jan Jan-7 Jan-9 Jan-11 Jan EURUSD 1. Fitted EURUSD.8 Jan-99 Jan-1 Jan-3 Jan Jan-7 Jan-9 Jan-11 Jan and Eurostat and Eurostat Figure 39: The bilateral basic balance is highly correlated with net non treasury flows Figure 4: US purchases of euro area bonds have continued to be replaced by sales 1 1 Bilateral Basic Balance Net Non Treasury Flows Bilateral Trade Balance USD 12M Net Agency 12M Net Corporate 12M net US stock 12M net foreign stocks 12m Net Foreign Bond Dec-98 Dec-2 Dec-6 Dec Dec-77 Dec-82 Dec-87 Dec-92 Dec-97 Dec-2 Dec-7 Dec-12 and Eurostat Source: Deutsche Bank and US Treasury Deutsche Bank AG/Hong Kong Page 27

28 27 February 214 Figure 41: Net portfolio flows remain positive Figure 42: Equity inflows are above bond inflows Portfolio Outflows Portfolio Inflows Net Portfolio Flows -8 Jan- Jan-2 Jan-4 Jan-6 Jan-8 Jan-1 Jan Net Bonds and Notes Flows Net MM Flows Net Equity Flows -1 Jan-3 Jan Jan-7 Jan-9 Jan-11 Jan Source: Deutsche Bank and European Central Bank Source: Deutsche Bank and European Central Bank Figure 43: Equity inflows are accelerating in sync with equity market upsurge Figure 44: Foreign interest on the bond side boomed in late 26 and has slowed now Net Equity Inflows Net Equity Outflows Euro STOXX (rs) -4 Jan-3 Jan Jan-7 Jan-9 Jan-11 Jan Net Bond Inflows Net Bond Outflows 1Y Bunds Yield (rs) -4 Jan-3 Jan Jan-7 Jan-9 Jan-11 Jan Source: Deutsche Bank, Bloomberg and European Central Bank Source: Deutsche Bank and European Central Bank Page 28 Deutsche Bank AG/Hong Kong

29 27 February 214 Australia (AUD bn) Figure 4: The basic balance remains negative Figure 46: in spite of net FDI inflows Basic Balance 4Q sum AUD/USD (rs) Net FDI FDI Outflows FDI Inflows Dec-9 Dec-94 Dec-98 Dec-2 Dec-6 Dec-1. Jan-91 Jan-94 Jan-97 Jan- Jan-3 Jan-6 Jan-9 Jan-12 and RBA and RBA Figure 47: Net Portfolio flows have started recovering after the fall in 21 Figure 48: Foreign investors have favored Australian debt (negative IIP a liability for AU) Net Portfolio flows Portfolio Outflow Portfolio Inflow Net Equity IIP Net Debt IIP AUD/USD (rs) Jan-91 Jan-94 Jan-97 Jan- Jan-3 Jan-6 Jan-9 Jan Mar-9 Mar-94 Mar-98 Mar-2 Mar-6 Mar-1.6. and RBA and RBA Figure 49: and to a lesser extent equities Figure : with relatively modest purchases by Australians of foreign debt Net Equity Liabilities for AU Net AU Equity Holdings by Foreigners Net Foreign Equity Holdings by Australia Mar-9 Mar-94 Mar-98 Mar-2 Mar-6 Mar Net Debt Liabilities for AU Net AU Debt Holdings by Foreigners Net Foreign Debt Holdings by Australia Mar-9 Mar-94 Mar-98 Mar-2 Mar-6 Mar and RBA and RBA Deutsche Bank AG/Hong Kong Page 29

30 27 February 214 New Zealand (NZD bn) Figure 1: The basic balance deteriorates further Figure 2: on low net FDI inflows 1-1 Basic Balance 4Q sum NZD/USD (rs) Net FDI FDI Outflows FDI Inflows Jan-1 Jan-3 Jan Jan-7 Jan-9 Jan-11 Jan Mar-1 Mar-3 Mar Mar-7 Mar-9 Mar-11 Mar-13-1 and Haver and Haver Figure 3: Net Portfolio flows turns negative Figure 4: Foreign appetite for government bonds Net Portfolio flows Portfolio Outflow Portfolio Inflow -2 Mar-96 Mar-99 Mar-2 Mar Mar-8 Mar Government Bonds Held by Foreigners Treasury Bills held by Foreigners 8 Share held by Foreigners of Total (right scale) Jul-91 Jul-94 Jul-97 Jul- Jul-3 Jul-6 Jul-9 Jul-12 and Haver and NZ FinMin Page 3 Deutsche Bank AG/Hong Kong

31 27 February 214 Commodity Price and Currency Monitor Figure 1: CRB Commodity Prices and components Figure 2: Energy prices Raw industrial Foodstuffs Metals Livestock and products,(rhs) Fats and Oil,(rhs) CRB Commodity Prices,(rhs) Oil (WTI, $/barrel) Natural Gas ($/mmtbu, rhs) Jan-2 Jan-4 Jan-6 Jan-8 Jan-1 Jan-12 Jan-14 1 Jan-2 Jan-4 Jan-6 Jan-8 Jan-1 Jan-12 Jan-14, Haver Figure 3: Precious metals Figure 4: Industrial metals Gold Price (US$/Troy oz) Platinum Price ($/Troy oz) Palladium Price ($/Troy oz) Silver Price ($/Troy oz),(rhs) Aluminium Price ($/Metric Tonne) Copper Price ($/Metric Tonne) Lead Price ($/Metric Tonne) Zinc Price ($/Metric Tonne) Nickel Price ($/Metric Tonne),(rhs) Tin Price ($/Metric Tonne),(rhs) Jan-2 Jan-4 Jan-6 Jan-8 Jan-1 Jan-12 Jan-14 3 Jan-2 Jan-4 Jan-6 Jan-8 Jan-1 Jan-12 Jan-14 Figure : Commodity Currencies and Prices Figure 6: The dollar cycle and global growth cycle AUD/USD CAD/USD NZD/USD CRB (Rs) yoy,%.1. Correlation over entire sample = -.7 Correlation from May 2 = -.1 Ln World IP USTW$, inverted,(rhs) -.1 Jan-81Jan-8Jan-89Jan-93Jan-97Jan-1JanJan-9Jan Deutsche Bank AG/Hong Kong Page 31

32 27 February 214 Figure 7: Nominal CRB and World IP Growth Figure 8: Nominal CRB and the Dollar Ln Nominal CRB Index World industrial Production(rhs) yoy,% Ln Nominal CRB Index USTW$,inverted,(rhs) Ln Jan-86 Jan-89 Jan-92 Jan-9 Jan-98 Jan-1 Jan-4 Jan-7 Jan-1 Jan Jan-86Jan-89 Jan-92 Jan-9 Jan-98 Jan-1 Jan-4 Jan-7 Jan-1 Jan-13 Figure 9: Long-run Relationship- Nominal CRB Figure 1: Long-run Relationship- Oil Long-run elasticities: TWI: -1.88, World IP:.81 Real Interest Rate: -.3 Ln Ln Elasticities: Major TWI: -2.6 World IP:.3 R-square: Nominal CRB Index Oil Price Fitted Oil Price 3.4 Fitted Nominal CRB Index 4.8 Jan-86 Jan-89 Jan-92 Jan-9 Jan-98 Jan-1 Jan-4 Jan-7 Jan-1 Jan May- May-3 May-6 May-9 May Figure 11: RBA Commodity Price Index (Nominal) and AUD/USD Figure 12: Long-run Relationship-AUD/USD AUD (lhs) RBA Commoditiy Price Index (rhs) AUD Long Run Relationship Long-run elasticities: Commodity Price: US GDP: Page 32 Deutsche Bank AG/Hong Kong

33 27 February 214 Figure 13: ANZ Commodity Price Index (Nominal) and Figure 14: Long-run Relationship-NZD/USD NZD/USD NZD (lhs) ANZ Commodity Prices Index (rhs) NZD Long-run elasticities: Commodity Price:.77 GDP: 1.8 Long Run Relationship Figure 1: BoC Commodity Price Index (Nominal) and Figure 16: Long-run Relationship-CAD/USD CAD/USD CAD (lhs) BoC Commodity Price Index CAD Long Run Relationship Long-run elasticities: Commodity Price:.13 GDP: Figure 17: BoC Non-Energy Commodity Price Index (Nominal) and CAD/USD Figure 18: BoC Energy Commodity Price Index (Nominal) and CAD/USD CAD (lhs) BoC Non-Energy Commodity Price Index (rhs) CAD (lhs) BoC Energy Commodity Price Index (rhs) Deutsche Bank AG/Hong Kong Page 33

34 27 February 214 Figure 19: RBA Commodity Price (Nominal) Figure 2: RBA Commodity Price (Real) RBA Commodity Price Index (Nominal) Average RBA Commodity Price Index (Real) Average Linear Trendline y = 3Ex R² = Figure 21: ANZ Commodity Price (Nominal) Figure 22: ANZ Commodity Price (Real) ANZ Commodity Price Index (Nominal) Average ANZ Commodity Price Index (Real) Average Linear Trendline y = -2E-6x R² = Figure 23: BoC Commodity Price (Nominal) Figure 24: BoC Commodity Price (Real) 1 8 BoC Commodity Price Index (Nominal) BoC Commodity Price Index (Real) Average Linear Trendline Average y = -2Ex R² = Page 34 Deutsche Bank AG/Hong Kong

35 27 February 214 Figure 2: BoC Non-Energy Commodity Price (Nominal) Figure 26: BoC Non- Energy Commodity Prices (Real) BoC Non-Energy Commodity Price Index Average BoC Non- Energy Commodity Price Index (Real) Average Linear Trendline y = Ex R² = Figure 27: BoC Energy Commodity Price (Nominal) Figure 28: BoC Energy Commodity Price (Real) 2 2 BoC Energy Commodity Price Index Average BoC Energy Commodity Price Index (Real) Average Linear Trendline y = 2Ex R² = Figure 29: Commodity Price Indices Figure 3: Ratio of Commodity Price Indices RBA Commodity Price Index (Nominal) ANZ Commodity Price Index (Nominal) BoC Commodity Price Index (Nominal) Jan 1986 = Ratio of Australia to NZ Commodity Price Indicies (Nominal) Ratio of Canada to NZ Commodity Price Indicies (Nominal) Deutsche Bank AG/Hong Kong Page 3

36 27 February 214 U.S. Trade Balance Monitor Figure 1: The US trade deficit continues to recover Figure 2: US and world industrial growth remains sedate -1-2 USD Bn yoy,% Annualized Trade Balance World IP ex US IP (YoY) US IP (YoY) -1-9 Annualized Trade Balance,3m Jan-92 Jan-96 Sum Jan- Jan-4 Jan-8 Jan-12 Annualized Trade Balance,12m Sum -9-1 Jan-92 Jan-96 Jan- Jan-4 Jan-8 Jan-12-1 Source: DataStream, Deutsche Bank. Source: DataStream, Deutsche Bank Figure 3: The narrowing in the deficit reflected a outpacing of import growth by export growth Figure 4: Recently export prices have receded sharply while export volumes have increased slightly yoy,% u Export Value Growth -2 Export Volume -3-1 Export Price(rhs) Import Value Growth Jan-92 Jan-96 Jan- Jan-4 Jan-8 Jan-12 Jan-92 Jan-96 Jan- Jan-4 Jan-8 Jan-12 Source: DataStream, Deutsche Bank Source: DataStream, Deutsche Bank Figure : Export prices tend to follow the dollar Figure 6: Export volume growth closely follows external demand yoy,% Export Price USTRBROA,inverted (rhs) u Ln yoy,% Export Volume World IP ex US IP(rhs) 1-12 Jan-94 Dec-96 Nov-99 Oct-2 Sep Aug-8 Jul Jan-94 Jan-98 Jan-2 Jan-6 Jan-1-1 Source: DataStream, Deutsche Bank Source: DataStream, Deutsche Bank Page 36 Deutsche Bank AG/Hong Kong

37 27 February 214 Figure 7: Export volumes have remained below trend since 21 Figure 8: Export volume deviations from trend is no longer correlated with moving average of dollar valuation Ln Real Broad TWI Export Volumes(rhs) Ln USDTWI,Deviations from Trend (8 Quarter MA),inverted Real Exports,Deviation from Trend (rhs) Ln Correlation = Mar-8 Mar-84 Mar-88 Mar-92 Mar-96 Mar- Mar-4 Mar-8 Mar Dec-81 Dec-8 Dec-89 Dec-93 Dec-97 Dec-1 Dec Dec-9 Dec Source: DataStream, Deutsche Bank Source: DataStream, Deutsche Bank Figure 9: A brief end to the dollar upsurge doesn t seem to bolster export volume growth Figure 1: Recovery in import price is boosting the volume during the past few months 2 1 yoy,% Export Volume USTRBROA,inverted(rhs) Ln yoy,% Jan-94 Jan-98 Jan-2 Jan-6 Jan Import Volume -1 Import Price (rhs) -2 Jan-92 Jan-96 Jan- Jan-4 Jan-8 Jan Source: DataStream, Deutsche Bank Source: DataStream, Deutsche Bank Figure 11: Import price inflation has followed the dollar Figure 12: Import volume growth has generally been highly correlated with US domestic demand growth 2 2 Import Price USTRBROA,inverted(rhs) yoy,% Jan-94 Jan-98 Jan-2 Jan-6 Jan yoy,% Import Volume US IP(rhs) Jan-94 Jan-98 Jan-2 Jan-6 Jan Source: DataStream, Deutsche Bank Source: DataStream, Deutsche Bank Deutsche Bank AG/Hong Kong Page 37

38 27 February 214 Figure 13: U.S. Exports and Imports of Goods and Services (Balance of Payments Basis) (last 13 months) Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Units Exports (US$ bn.) Imports (US$ bn.) Trade Balance (US$ bn.) Export & Import Growth Exports (y-o-y%) 3.9%.3% 4.% 2.% -1.% 1.6% 1.4% 3.1% 3.4% 4.1% 1.3%.7%.2% Imports (y-o-y%) 2.% -1.% -.8% 1.9%.% -1.8%.1% -.9%.7% 1.1% 1.7% 3.2% -1.1% Growth Differential 1.4% 6.8% 4.8%.6% 4.% 3.4% 1.3% 4.% 2.7% 3.% -.4% 2.6% 6.3% Source: Deutsche Bank and Reuters Figure 14: U.S. Export and Import Orders (ISM Survey) (last 13 months) Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Units Export Orders (index) Import Orders (index) Exp.-Imp. Orders Source: Deutsche Bank and Reuters Figure 1: Regional Breakdown of U.S. Trade Balance (US$ bn.) ( ) Canada Mexico Brazil Western Europe Germany U.K Japan China Hong Kong South Korea Singapore Taiwan U.S. Total Source: DataStream, Deutsche Bank Page 38 Deutsche Bank AG/Hong Kong

39 27 February 214 U.S. Exports-Imports by Commodity Figure 16: U.S. Trade Balance Excluding China & Petroleum (Monthly & Annual Balance) Figure 17: U.S. Trade Balance Advanced Technology Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Source: DataStream Source: DataStream Figure 18: U.S. Trade Balance Petroleum Products Figure 19: U.S. Trade Balance Consumer Goods Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Source: DataStream Source: DataStream Figure 2: U.S. Trade Balance Capital Goods Figure 21: U.S. Trade Balance Industrial Supplies Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Source: DataStream Source: DataStream Deutsche Bank AG/Hong Kong Page 39

40 27 February 214 Figure 22: U.S. Trade Balance Automotive Figure 23: U.S. Trade Balance Food & Beverages Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Source: DataStream Source: DataStream Page 4 Deutsche Bank AG/Hong Kong

41 27 February 214 U.S. Bilateral Trade Balances by Country & Region Figure 24: U.S. Trade Balance with China Figure 2: U.S. Trade Balance with Japan Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Source: DataStream Source: DataStream Figure 26: U.S. Trade Balance with the Pacific Rim (Asia excluding China and Japan) Figure 27: U.S. Trade Balance with OPEC Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Monthly (US$ bn.) (bars) -2 Annual (US$ bn.) (line) Source: DataStream Source: DataStream Figure 28: U.S. Trade Balance with Western Europe Figure 29: U.S. Trade Balance with Canada Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Monthly (US$ bn.) (bars) -1 Annual (US$ bn.) (line) Source: DataStream Source: DataStream Deutsche Bank AG/Hong Kong Page 41

42 27 February 214 Figure 3: U.S. Trade Balance with Mexico Figure 31: U.S. Trade Balance with Latin America Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Monthly (US$ bn.) (bars) Annual (US$ bn.) (line) Source: Deutsche Bank Source: Deutsche Bank Page 42 Deutsche Bank AG/Hong Kong

43 27 February 214 U.S Current-Account Balance Monitor Figure 32: U.S. Current-Account Balance (198-21) Figure 33: U.S. Savings and Investment (Private & Government Sector Savings-Investment) Annualized Current Account as % of GDP Private Sector Balance Gov't Sector Balance Mar-81 Mar-8 Mar-89 Mar-93 Mar-97 Mar-1 Mar Mar-9 Mar Mar-81 Mar-8 Mar-89 Mar-93 Mar-97 Mar-1 Mar Mar-9 Mar-13 Source: DataStream Source: DataStream Figure 34: U.S. Current-Account Balance (last 13 quarters) (US$ Billions) Q3 21Q4 21Q1 211Q2 211Q3 211Q4 211Q1 212Q2 212Q3 212Q4 212Q1 213Q2 213Q3 213 Balance on Goods Balance on Services Bal on Goods & Services Investment Income Unilateral Transfers Bal on Current Account (annualized, as % of GDP) -3.2% -2.7% -3.1% -3.1% -2.7% -2.9% -3.% -2.7% -2.6% -2.% -2.% -2.3% -2.2% Source: DataStream Figure 3: U.S. Current-Account Balance ( ) (US$ Billions) Balance on Goods Balance on Services Bal on Goods & Services Investment Income Unilateral Transfers Bal on Current Account (annualized, as % of GDP) -4.2% -3.9% -4.3% -4.7%.3%.9% -6.%.1% -4.7% -2.7% -3.% -3.1% -3.% Source: DataStream Deutsche Bank AG/Hong Kong Page 43

44 27 February 214 Figure 36: U.S. Savings-Investment & Net Foreign Investment ( ) (US$ Billions) Private Savings Private Investment Private-Sector Balance Gov't Savings Gov't Investment Gov't-Sector Balance Gross Savings Gross Investment Savings-Investment Statistical Discrepancy Net Foreign Investment Source: DataStream Page 44 Deutsche Bank AG/Hong Kong

45 27 February 214 Central Bank Reserves Currency Composition Monitor Figure 1: Official FX reserves have quadrupled reflecting primarily the growth of EM holdings Figure 2: Mature market (MM) reserves have grown only modestly reflecting valuation & interest Source: FRB, Census, BEA, DB Global Markets Research Source: FRB, Census, BEA, DB Global Markets Research Figure 3: Many countries report the currency composition of reserves to the IMF, which publishes them in aggregate form Figure 4: The advanced countries (MM) all report the composition of reserves to the IMF Source: COFER, IMF, DB FX Research Source: COFER, IMF, DB FX Research Figure : while about half of emerging markets report the currency composition of their reserves Figure 6: The currency composition of (114 reporting countries) total FX reserves: levels Source: COFER, IMF, DB FX Research Source:, COFER, IMF, DB FX Research Deutsche Bank AG/Hong Kong Page 4

46 27 February 214 Figure 7: The USD share in world reserves fell during 22-4; then stabilized Figure 8: Advanced country FX reserve holdings Source COFER, IMF, DB FX Research Source:, COFER, IMF, DB FX Research Figure 9: the dollar share in industrial country reserves has been relatively stable Figure 1: Ex-Japan (our estimate) industrial country dollar and euro holdings have both risen Source: COFER, IMF, DB FX Research Source:, COFER, IMF, DB FX Research Figure 11: The share of euros and dollars is not very different Figure 12: EM holdings of dollars had climbed steadily Source: COFER, IMF, DB FX Research Source: COFER, IMF, DB FX Research Page 46 Deutsche Bank AG/Hong Kong

47 27 February 214 Figure 13: In EM, the main driver of reserve growth has been intervention (in USD bn) Figure 14: In EM, the dollar share fell steadily during 22-4 then stabilized Source: COFER, IMF, DB FX Research Figure 1: First active diversification, then leaning against the wind Figure 16: China has steadily diversified away from USD since 24 (our estimates) Source: COFER, IMF, DB FX Research Source: US TIC data DB FX Research Deutsche Bank AG/Hong Kong Page 47