Market Musings Global Strategy 22 February 2017 | TD Securities | Toronto

Benchmarking Fair Value in G10 FX  The rhetoric on exchange rates has been

one of the key drivers for the major currencies over the past few months, and we do not see signs of it ebbing soon



We benchmark fair value in G10 FX using a few different valuation models to measure how far exchange rates have deviated from long-term fair value and look at possible triggers for correction



The aggregate signal indicates that JPY, EUR and SEK are the most undervalued while USD, CHF and NOK are the most overvalued major currencies



CAD and GBP remain overvalued on external imbalances, suggesting more downside over the coming months as they struggle with rebalancing

One of the notable developments in G10 FX has been the pickup in rhetoric on international exchange rate policy since the start of the year. This has sparked a debate about the valuation of the major currencies. Even though jawboning is not new to market practitioners, it comes against a backdrop of a young administration trying to find its feet. Recall that in the recent weeks, the Trump administration has 1) indicated that the USD is too strong 2) argued that the implied Deutsche Mark is “grossly undervalued” and 3) suggested that China and Japan use money markets to devalue their currencies. We think figure 1 sums up the administration's FX views quite well. It shows the deviation of the REER from the 20yma and the trade deficit with the US (% of US’s total world trade). We suspect the key driving force behind the rhetoric on exchange rates is the administration's economic goal of eliminating the US trade deficit with the rest of the world. This reflects their view that trade imbalances are predatory, so US deficits are a sign of stolen growth from abroad. These global Figure 1:Trade Deficits and FX Valuations New Zealand Switzerland Australia

REER (deviation from 20yma)

9% 4% -1%

-6%

Canada Norway EZ France Germany

-11%

UK

-16% China

-21%

Japan

-26% Mexico -31% -50%

-40%

-30% -20% -10% 0% trade deficit with US (% of total)

Source: Bloomberg, TD Securities

10%

20%

imbalances, they might suppose, are in turn symptoms of manipulation in the FX market. We do not share this view but we break down a few of the most commonly followed lowfrequency valuation models to gauge valuation in G10 FX and to add a bit of context to this ongoing debate about valuations. Just price it For this exercise, we focus on three commonly followed models of exchange rate valuation: purchasing power parity (PPP), fundamental equilibrium model (FEER) and behavioural equilibrium model (BEER). The most common method of exchange rate valuation (at least in Econ 101 textbooks) is PPP or the law of one price. This theory argues that exchange rates adjust to reflect the price differential between countries, so inflation differentials drive nominal exchange rates. This may be true for the long-run but most empirical evidence indicates that the half-life of PPP sits between 3-5 years. The downside to this approach is that there are often persistent departures from PPP over the shortrun, rendering this model ineffective for most FX practitioners. Moreover, it also assumes that exchange rates are stationary. This means that exchange rates adjust to a fixed, steady average. Given the host of variables and components that drive returns, this is also flawed logic regarding the movement of exchange rates. One of the other drawbacks is that PPP takes a bilateral approach to fair value. The logic in PPP argues, for instance that a currency may be overvalued against some currencies and undervalued against others even though it is close to fair value in PPP terms. Still, it is a widelyfollowed tool for guidance of the direction of exchange rates over long cycles and market participants use it as a simple gauge of long-run fair value (and for understanding BAT). We look at the average of three calculations (from three sources) to benchmark PPP: the OECD, Bloomberg and the REER deviation from the 10yma (see table below): Low-frequency fair value: a few key models FEER

PPP

BEER

Average

USD

-0.1%

16.2%

5.9%

7.3%

NOK

17.8%

-4.7%

6.1%

6.4%

CHF

-3.8%

11.1%

6.3%

4.5%

NZD

-1.3%

10.6%

-0.9%

2.8%

AUD

-1.3%

3.7%

-1.2%

0.4%

CAD

7.1%

-9.4%

-4.6%

-2.3%

GBP

9.3%

-9.8%

-19.2%

-6.6%

SEK

0.0%

-12.2%

-12.2%

-8.2%

EUR

-5.0%

-11.3%

-11.0%

-9.1%

JPY

-8.3%

-16.4%

-17.9%

-14.2%

So urce: IM F, P IIE, M acro bo nd TD Securities

Market Musings 22 February 2017 | TD Securities | Toronto

The only thing to FEER The second model in our sample is a FEER. This approach measures the exchange rate regarding the internal and external balance. In other words, the fair value of a FEER model is the exchange rate that equates to a sustainable level of the current account. This steady state of balance does not mean current accounts should balance out to zero but instead balances over time to a set of structural and cyclical factors. Recall, the accounting standard of the current account measures the savings and investment balance of a particular country. That means countries that are net importers of capital can still run sustainable current accounts deficits, so deficit countries do not always have overvalued currencies. A FEER model differs with other low-frequency fair value models (LFFV), notably BEER, in that it is not an econometric model. Instead, it is a method of calculation, so there is some basis for subjectivity in the estimates. For instance, the practical and academic literature does not have universal standards for measuring price elasticities, theoretical assumptions or measures of the output gap. We believe this leads to a wider approach for analysis that balances out the existing framework from sources like PIIE and the IMF. Our method adopts a few different approaches to arrive at a sustainable current account and ultimately the exchange rates needed to balance it. First, we incorporate estimates from the PIIE model. This method benchmarks the exchange rate to a sustainable level of the external balance (current account). Their baseline assumption is that the current account does not differ from zero by more than 3% of GDP, so the exchange rate works to balance deviations from there. In effect, the exchange rate over time helps to manage the drift of the current account from its longer-run structural level. Deviation of LFFV models in G10 FX

20.0% 15.0% 10.0% 5.0%

0.0% -5.0% -10.0%

Source: TD Securities, PIIE, IMF, Macrobond

JPY

EUR

SEK

GBP

CAD

AUD

NZD

USD

-25.0%

CHF

-20.0%

NOK

FEER PPP BEER Average

-15.0%

The PIIE model also assumes the exchange rate helps to maintain this equilibrium through changes in exports, not imports. The impact parameter captures the required adjustment to maintain a sufficient external balance and measures the impact on the current account for a 1% change in the exchange rate. We incorporate this impact parameter into our estimates and average it against the IMF’s figures. Second, we also look at impact parameters from the IMF’s External Balance Assessment (EBA) report. For the IMF data, we blend two elasticity figures: a country-level number and single coefficient from the IMF’s panel regression. For the IMF’s framework, the cyclically-based current account reflects a set of desirable macroeconomic fundamentals and policies. This means that any departure from cyclically adjusted external balance should lead to an adjustment in the exchange rate. External deficits, in turn, reflect overvaluation, leading in time to a depreciation in the exchange rate and vice versa for surpluses. Naturally, a country’s share of trade drives the speed of adjustment, so countries that are more open tend to see quicker adjustments in their exchange rate. The share of trade is important to understand how the impact parameter shapes the relationship between external balances and the exchange rate. The elasticity of closed economies (smaller ratio of trade to GDP) indicates that exchange rates will need to do more of the heavy lifting in the rebalancing process. This process also differs with highly open economies that tend to see a more rapid adjustment in the exchange rate when imbalances gravitate too far equilibrium. Another point worth noting is that the shift to globally integrated supply chains over the past 20 years has probably led to a reduction in the elasticity of exchange rates and trade balances, indicating a structural break in some of the old rules. This breakdown has been one of the developments that has limited the economic rebalancing in Canada despite the 25% drop (peak to trough) in the NEER over the past few years. Our final estimates blend these approaches to arrive at an aggregate FEER valuation for the G10 currencies. We define the sustainable current account as the average of the IMF’s structural balance and a rolling 10ymma. Then we look at the most recent estimate of the current account and measure the deviation of it from the blended sustainable level. This measures the external gap. The estimated exchange rate then becomes the adjustment level needed to eliminate this gap. The level of valuation ultimately depends on the impact parameters that we source from the PIIE and IMF’s estimates. In the end, both the PIIE and IMF approaches have their drawbacks, so our approach relies on average deviations in this framework to benchmark valuation.

2

Market Musings 22 February 2017 | TD Securities | Toronto

This BEER’s for you The third model in the sample is a behavioral exchange rate model (BEER). We use an in-house econometric model that uses three macroeconomic variables to measure fair value: relative productivity, terms of trade and net foreign assets (% of GDP). The model specification is based in part on a stockflow model that uses cross country variables to produce a single coefficient for the entire sample. The benefit of this approach is the valuation of the exchange rate is based on macroeconomic variables rather than relative prices - in the case of PPP. This marries the practical and theoretical views about medium-term determinants of the exchange rate (more details of the BEER model and specification in the appendix). G10FX valuations most undervalued

USD

most overvalued

overvalued to undervalued

20.0% JPY

NOK

10.0% 0.0%

EUR

CHF

-10.0% -20.0%

FEER

SEK

NZD

PPP BEER Average

GBP

AUD

CAD

worth exploring the market regime to gauge in more detail the underlying theme that is driving markets. Ultimately, we think that a better understanding about valuation models should enhance convictions around medium-term views and trades. Thanks for all this but what’s the trade? Exchange rates tend to get lumped together in a few tradable themes: momentum, carry, and valuation. Our MRSI and SRSI indicators are tools that fall into the valuation category (we also plan to test them as fair value tools in upcoming pieces). For now, FX markets are lacking a major theme since the US election. Despite high political uncertainty, low market volatility has sprung life back in to carry and momentum strategies. Even so, we expect the rise in political uncertainty and convergence of global growth will likely shift the focus in FX markets back to valuation-based strategies. A simple carry-based rule in the G10 has seen a return of 9.8% (annual) since the election. This environment has helped the likes of AUD and NZD outperform while some of the European currencies have lagged in the broader G10 given their low rate profiles. In the same way, momentum traders have chased the returns in carry, leading to a respectable 4.6% (annual) gain in simple momentum rule following models over the same time period. Both momentum and carry do indeed perform well when volatility is low.

Source: TD Securities, PIIE, IMF, Macrobond

Pulling it all together The goal of this exercise was to provide a set of tools that offer different methods for valuing exchange rates. The models differ in timing, theory, and rigor but combined provide a more complete picture of exchange rate misalignments in the G10. Keep in mind that while PPP is the standard workhorse, it does have a considerably longer half-life than BEER or FEER. This means the half of the predicted misalignment is plugged much faster with BEER and FEER models. Based on a vector error correction model, the half-life of our BEER model is around 18 months. The average FEER model has a half-life around two years, leaving PPP the least useful valuation model for practical forecasting purposes. We also stress that no valuation model works all the time. Exchange rates deviate from longer-term fair value for considerable periods of time. This reflects the competing nature of drivers with risk appetite, positioning, and politics – just to a name a few, pushing against fair value on a daily basis. We believe that the key to valuation models is benchmarking thresholds and combining signals to get a complete picture. If the models agree on the direction of valuation, then that is likely to increase the probability of a correction. Alternatively, when models disagree, it is probably

103.00

Carry and momentum back in favor in G10FX

102.50 102.00 101.50 101.00 100.50 100.00

99.50 99.00

G10 Momentum

98.50

G10 Carry

98.00 11/9/2016

12/9/2016

1/9/2017

2/9/2017

Source: TD Securities, Bloomberg

The flipside is valuation-based strategies tend to offer a hedge against trending markets. Indeed, the build-up in macroeconomics imbalances dovetails with the deviations from fair value over time. We believe that the foundation of carry and momentum strategies rests on the assumption that exchange rates will deviate from ’equilibrium’ for a considerable period of time. This feedback loop reinforces the logic of trend following rules until it breaks. For instance, the build-up of current account deficits reflect in part the movement of capital flows to higher-yielding currencies that could be overvalued based on inflation differentials. The perceived stability of the macroeconomic environment fuels further capital flows until a trigger causes a rush to the exits.

3

Market Musings 22 February 2017 | TD Securities | Toronto

Despite the perceived stability in markets at the moment, the Trump administration's focus on exchange rate valuation and international trade policy should tip the balance of risk back towards valuation strategies instead of carry or momentum. Notably, the administration has already flagged that the USD is too expensive against a handful of major currencies. 120

Valuation correlates positively with volatility

110

25 20

100

15 90

10

80 70 60 Jan-04

5

G10 Valuation (LHS) CVIX Jan-07

Jan-10

Jan-13

Jan-16

0

Source: TD Securities, Bloomberg

These include the current account surplus countries of Japan, China, and Germany, which the government has criticized for either currency manipulation or unfair trade practices. The exchange rate is arguably the most important tool to correct these imbalances, so one critical implication of the Trump economic agenda is the correction of the so-called imbalances in the FX market. Over time this could see a significant correction of the valuation gaps highlighted above, especially if this coincides with the convergence of central bank policy and global growth we expect over the next year. For now, this will put greater emphasis on a few key upcoming events, including the March G20 meetings and the release of the US Treasury’s semi-annual report on currencies keeping the risks focused on international policy. The spider web chart on the previous page (and the table on the first) provide a snapshot of the misalignments in the G10 from our analysis. A few observations stick out. First, the jury is out on the valuation of CAD. Both PPP and BEER argue that CAD is undervalued by around 7% while our FEER estimates suggest it is, in fact, 7% overvalued. This divergence in valuation is also seen in GBP. We think the fact that the models disagree on these currencies is important given both countries are dealing with external shocks that policy makers seek to manage through exchange rate adjustments. In other words, the external shocks (Brexit and oil) leave exchange rates in Britain and Canada vulnerable to do more of the heavy lifting given large external deficits.

For CAD, the BoC has discussed the rotation from energy to manufacturing for years and has recently concluded that a mix of structural and cyclical factors has limited the expected rebalancing. Our FEER analysis arrives at a similar conclusion, indicating that the exchange rate is still overvalued and elasticises have diminished over time. This leaves CAD vulnerable to further losses as internal and external rebalancing persists. Our FEER model places USDCAD fair value at 1.40. What’s more, the collateral damage to Canadian trade from a shift in US trade policy is another risk factor while the economy continues to work through the energy shock. This shift from energy to manufacturing has been slow to materialize and should also offer some caution to policymakers in the UK who also aspire to rebalance the economy away from a core sector (finance). Second, the models point to three undervalued currencies: JPY, EUR, and SEK. Each share a common thread of easy monetary policy and a current account surplus. The JPY sticks out like a sore thumb in the group with an average undervaluation of 14%. This suggests the LFFV of USDJPY is around100, leaving it vulnerable to external triggers. This level of undervaluation will also keep JPY in the political crosshairs of the Trump administration. Politically, this could limit the BoJ’s policy options over the next year, increasing the scope for convergence with the Fed or JGB tantrum. The ECB is in a similar environment, and the models agree that the EUR is undervalued as well. Our recent EURJPY trade idea is tactical but it also fits in with results of this exercise. In the G10, our bias is for more USD strength this year on one final jolt of policy divergence. For the most part, this reflects the divergence in output gaps and domestic prices. But we do not expect the greenback to make new highs for the cycle and is likely to struggle to test the highs already marked over the past few years. Indeed, our longer-run forecasts show it underperforming the consensus in the G10 next year, especially against the undervalued European currencies. Against this backdrop, a few valuation-based trades over the coming year would be consistent with EUR and JPY strength and underweight in NOK and CHF. This also coincides with a cyclical peak in the USD. As noted above, CAD and GBP could still struggle this year as they continue to work through external shocks but their outlook improves next year. This reflects the scope for progress on the ongoing rebalancing efforts. All told, a stronger greenback in H1 is likely to give way to a pullback later this year and early into next as global imbalances and valuations start to correct.

Mark McCormick +1 416 982 7784 4

Market Musings 22 February 2017 | TD Securities | Toronto

Appendix: The BEER model specification follows the form below: REER = α+β*ToT+β*productivity+β*NFA+Ε The building blocks for the BEER framework are three macroeconomic variables. Economic theory and practical literature highlight these variables as key determinants of medium-term exchange rate valuation. Below we provide a brief overview of the variables:



Productivity: an increase in a country’s relative productivity performance should lead to exchange rate appreciation through multiple channels. First, a rise in productivity growth raises national wealth and overall demand. The result is that the pickup in productivity growth leads to see a bump in real interest rates, leading to exchange rate strength. Secondly, according to the Balassa-Samuelson effect, higher productivity of the tradable sector places upward pressure on wages in the non-tradable sector leading to higher relative prices. Our productivity proxy is the IMF’s dataset on GDP per capita.



Terms of trade: a rise in the terms of trade increases a country’s profits and incomes, thus providing a boost to wealth. This lift in profitability stimulates capital investment, generating capital inflows and the return on investment, while the rise in income increases consumer spending. Higher profits also attract international capital flows, such as foreign direct investment and portfolio flows, which both exert upward pressure on a country’s exchange rate. The immediate effect of a change in the terms of trade on exchange rates depends in part on the openness of the economy and the size of trade in the country. Capital mobility is another important factor that captures the impact of trade shocks. We use the Citi commodity terms of trade indices for our estimates.



Net foreign assets (% of GDP): theory suggests that creditor countries tend to see their currency strengthen as international demand for the currency rises. Debtor countries, on the other hand, need a weaker currency to generate trade surpluses necessary to rebalance the economy and help to service the external deficits. The build-up of current account deficits (or surpluses) reveals a shift in a country’s exposure to net foreign liabilities (or assets). This process shapes the stock of the country’s balance sheet. For deficit nations, the accumulation of foreign liabilities (increasing current account deficits) over time is consistent with the depreciation of the exchange rate and a higher risk premium.

Regarding the methodology, the dataset panels pools the cross-section of the major G10 exchange rates and uses a panel cointegration econometric model to determine fair value. We use a balanced panel of quarterly data with ten cross sections and 66 observations. Before we run the data we use a series of robustness checks to test whether the data are non-stationary. Our robustness checks find evidence that the variables do not mean revert and share a common trend. Moreover, the use of a panel framework allows the estimation of a set of countries jointly in the same regression equation using a fixed constant. This process helps to control for country level effects and is preferred to the random effects one, especially since effective exchange rates lack comparison across countries. We note that the panel specification may introduce some bias in the estimates since our assumption indicates that the fixed effects are identical across all countries. Still, the main benefit of the panel is to increase the size of the sample and improve the robustness of the coefficients for quarterly data.

5

Market Musings 22 February 2017 | TD Securities | Toronto

GLOBAL STRATEGY TEAM Richard Kelly

Head of Global Strategy

44 20 7786 8448

Global Macro Strategy Annette Beacher

Chief Asia-Pacific Macro Strategist

65 6500 8047

Jacqui Douglas

Chief European Macro Strategist

Michael Hanson

Chief U.S. Macro Strategist

James Rossiter

Senior Global Strategist

Brittany Baumann

Macro Strategist

1 416 982 3297

Robert Both

Macro Strategist

1 416 983 0859

44 20 7786 8439 1 212 827 7186 44 20 7786 8422

Global Rates Strategy Priya Misra

Head of Global Rates Strategy

1 212 827 7156

Andrew Kelvin

Senior Rates Strategist

1 416 983 7184

Gennadiy Goldberg

Rates Strategist

1 212 827 7180

Prash Newnaha

Rates Strategist

65 6500 8047

Cheng Chen

Quantitative Strategist

1 212 827 7183

FX Strategy Ned Rumpeltin

European Head of FX Strategy

44 20 7786 8420

Mark McCormick

North American Head of FX Strategy

1 416 982 7784

Mazen Issa

Senior FX Strategist

1 212 827 7182

Emerging Markets Strategy

Global Strategy USA Canada

Cristian Maggio

Head of Emerging Markets Strategy

44 20 7786 8436

Paul Fage

Senior Emerging Markets Strategist

44 20 7786 8424

Sacha Tihanyi

Senior Emerging Markets Strategist

1 212 827 7403

Margaret Ismond

Emerging Markets Strategist

44 20 7786 8431

Commodities Strategy Bart Melek

Head of Commodity Strategy

1 416 983 9288

Ryan McKay

Commodity Strategist

1 416 982 5816

Australia New Zealand UK Europe Emerging Markets

Research Home Page: https://www.tdsresearch.com/currency-rates Bloomberg Page: TDGR

FX & Commodities 6

Market Musings Global Strategy 22 February 2017 | TD Securities | Toronto DISCLAIMER This material is for general informational purposes only and is not investment advice nor does it constitute an offer, recommendation or solicitation to buy or sell a particular financial instrument. It does not have regard to the specific investment objectives, financial situation, risk profile or the particular needs of any specific person who may receive this material. No representation is made that the information contained herein is accurate in all material respects, complete or up to date, or that it has been independently verified by TD Securities. Recipients of this analysis or report are to contact the representative in their local jurisdiction with regards to any matters or questions arising from, or in connection with, the analysis or report. Historic information regarding performance is not indicative of future results and investors should understand that statements regarding future prospects may not be realized. All investments entail risk, including potential loss of principal invested. Performance analysis is based on certain assumptions, the results of which may vary significantly depending on the modeling inputs assumed. This material, including all opinions, estimates and other information, constitute TD Securities’ judgment as of the date hereof and is subject to change without notice. The price, value of and income from any of the securities mentioned in this material can fall as well as rise. Any market valuations contained herein are indicative values as of the time and date indicated. Such market valuations are believed to be reliable, but TD Securities does not warrant their completeness or accuracy. Different prices and/or valuations may be available elsewhere and TD Securities suggests that valuations from other sources be obtained for comparison purposes. Any price or valuation constitutes TD Securities’ judgment and is subject to change without notice. Actual quotations could differ subject to market conditions and other factors. TD Securities disclaims any and all liability relating to the information herein, including without limitation any express or implied representations or warranties for, statements contained in, and omissions from, the information. TD Securities is not liable for any errors or omissions in such information or for any loss or damage suffered, directly or indirectly, from the use of this information. TD Securities may have effected or may effect transactions for its own account in the securities described herein. No proposed customer or counterparty relationship is intended or implied between TD Securities and a recipient of this document. TD Securities makes no representation as to any tax, accounting, legal or regulatory issues. Investors should seek their own legal, financial and tax advice regarding the appropriateness of investing in any securities or pursuing any strategies discussed herein. Investors should also carefully consider any risks involved. Any transaction entered into is in reliance only upon the investor’s judgment as to financial, suitability and risk criteria. TD Securities does not hold itself out to be an advisor in these circumstances, nor do any of its representatives have the authority to do so. The information contained herein is not intended for distribution to, or use by, any person in any jurisdiction where such distribution or use would be contrary to applicable law or regulation or which would subject TD Securities to additional licensing or registration requirements. It may not be copied, reproduced, posted, transmitted or redistributed in any without the prior written consent of TD Securities. Australia: If you receive this document and you are domiciled in Australia, please note that this report is intended to be issued for general information purposes only and distributed through the Toronto Dominion Australia Limited (“TDAL”). TDAL does not hold itself out to be providing financial advice in these circumstances. TD Securities is a trademark and represents certain investment dealing and advisory activities of Toronto-Dominion Bank and its subsidiaries, including TDAL. The Toronto-Dominion Bank is not an authorized deposit-taking or financial services institution in Australia. TDAL is a holder of an Australian Financial Services License (404698) and is regulated by the Australian Securities and Investments Commission. New Zealand: The Toronto-Dominion Bank is not a “registered bank” in New Zealand under the Reserve Bank Act 1989. Canada: Canadian clients wishing to effect transactions in any security discussed herein should do so through a qualified salesperson of TD Securities or TD Securities Inc. TD Securities Inc. is a member of the Canadian Investor Protection Fund. China, India and South Korea: Insofar as the document is received by any persons in the People's Republic of China (“PRC”) (which, for such purposes, does not include Hong Kong, Macau or Taiwan), India and South Korea, it is intended only to be issued to persons who have the relevant qualifications to engage in the investment activity mentioned in this document. The recipient is responsible for obtaining all relevant government regulatory approvals/licenses themselves, and represents and warrants to TD Bank that the recipient's investments in those securities do not violate any law or regulation, including, but not limited to, any relevant foreign exchange regulations and/or overseas investment regulations. The Toronto-Dominion Bank has a representative office in Shanghai, Mumbai and Seoul which should be contacted for any general enquiry related to The TorontoDominion Bank or its business. However, neither any of the Toronto-Dominion Bank offshore branches/subsidiaries nor its representative offices are permitted to conduct business within the borders of the PRC, India and South Korea. In locations in Asia where the Bank does not hold licenses to conduct business in financial services, it is not our intention to, and the information contained in this document should not be construed as, conducting any regulated financial activity, including dealing in, or the provision of advice in relation to, any regulated instrument or product. This publication is for general information only, without addressing any particular needs of any individual or entity, and should not be relied upon without obtaining specific advice in the context of specific circumstances. Hong Kong: This document, which is intended to be issued in Hong Kong only to Professional Investors within the meaning of the Securities and Futures Ordinance (the "SFO") and the Securities and Futures (Professional Investor) Rules made under the SFO, has been distributed through The Toronto-Dominion Bank, Hong Kong Branch. The TorontoDominion Bank, Hong Kong Branch is regulated by the Hong Kong Monetary Authority. Japan: For Japanese residents, please note that TD Securities is not licensed in Japan and this is being provided to qualified financial institutions (“QFI”) only under a relevant exemption to the Financial Instruments and Exchange Law. Singapore: This report is distributed in Singapore by The Toronto-Dominion Bank, Singapore Branch, and recipients in Singapore of this report are to contact The TorontoDominion Bank, Singapore Branch in respect of any matters arising from, or in connection with, this report. The Toronto-Dominion Bank, Singapore Branch is regulated by the Monetary Authority of Singapore. Where this report is issued or promulgated in Singapore, it is only intended for distribution to a person who is an accredited investor, expert investor or institutional investor (as defined in the applicable Singapore laws and regulations). United Kingdom and Europe: This document is prepared, issued or approved for issuance in the UK and Europe by TD Securities Limited in respect of investment business as agent and introducer for TD Bank. The Toronto-Dominion Bank is authorised by the Prudential Regulation Authority and subject to regulation by the Financial Conduct Authority and limited regulation by the Prudential Regulation Authority. TD Securities Limited is authorised and regulated by the Financial Conduct Authority. Insofar as the document is issued in or to the United Kingdom or Europe, it is intended only to be issued to persons who (i) are persons falling within Article 19(5) ("Investment professional") of the Financial Services and Markets Act 2000 (Financial Promotion) Order 2005 (as amended, the "Financial Promotion Order"), (ii) are persons falling within Article 49(2)(a) to (d) ("High net worth companies, unincorporated associations, etc.") of the Financial Promotion Order, or (iii) are persons to whom an invitation or inducement to engage in investment activity (within the meaning of section 21 of the Financial Services and Markets Act 2000) in connection with the issue or sale of any securities may otherwise lawfully be communicated or caused to be communicated. European clients wishing to effect transactions in any security discussed herein should do so through a qualified salesperson of TD Securities Limited. Insofar as the information in this report is issued in the U.K. and Europe, it has been issued with the prior approval of TD Securities Limited. United States: U.S. clients wishing to effect transactions in any security discussed herein must do so through a registered representative of TD Securities (USA) LLC. (a) the U.S. firm accepts responsibility for its contents, (b) persons receiving the report should effect transactions in securities discussed in the report through the U.S. firm, and (c) transactions in such securities by recipients of the report are actually effected only through the U.S. firm. If you would like to unsubscribe from our email distribution lists at any time, please contact your TD Securities Sales Contact or email us at [email protected]. You can access our Privacy Policy here (http://www.tdsecurities.com/tds/content/AU_Privacy_EandAP_en_CA). TD Securities is a trademark of the Toronto Dominion Bank and represents TD Securities Inc., TD Securities (USA) LLC and TD Securities Limited and certain investment and corporate banking activities of TD Bank and its subsidiaries. © Copyright 2014 The Toronto-Dominion Bank. All rights reserved.

7