The LSR Macro Picture - Lombard Street Research

The LSR Macro Picture

Global macroeconomic themes in focus

April 10, 2014

The rate escape The Fed has clearly mapped out its exit strategy over the next few years. But the chances of getting this exactly right look slim. If the Fed keeps policy as loose as its forecasts suggest, it could create new asset-price bubbles. Exiting faster might counter these risks, but neither the FOMC nor markets seem prepared for this. Chart 1: The next policy error? Fed funds adj for QE

Taylor rule

Fed funds rate

12

FOMC forecast

short-term policy rate, per cent

10 8 6 4 2 0 -2 -4 85

87

89

91

93

95

97

99

01

03

05

07

09

11

13

15

17

Escape plan The Fed faces a difficult task over the next few years as it starts to tighten policy. In an effort to avoid past errors, it has set out a clear strategy and timetable. If the economy evolves as it expects, QE will end in late 2014 and rate increases will start in 2015 Q2. It would then aim to hike by around 350bps over the following 18 months.

Bloatation device The Fed is reluctant to sell the assets it holds. This simplifies exit but leaves it with a bloated balance sheet that will take years of natural wastage to shrink. Holding these assets makes it harder for the Fed to judge how much stimulus it is providing to the economy and what level of interest rates it must ultimately reach.

Exit wounds According to the Fed’s guidance, monetary policy will remain extremely loose over the next few years. This is more likely to create asset-price bubbles than inflation – especially with excess global savings still finding their way into the US. If the Fed wants to fight these bubbles, it could drain liquidity and raise rates more quickly.

The rate escape Central banks have often struggled when trying to ‘exit’ from loose monetary policies and this time will be trickier than most. In an attempt to guide markets and prevent a repeat of past errors – particularly its 1994 experience, when an opaque Fed left markets behind the curve and caused a substantial selloff in bonds – the FOMC has carefully mapped out its strategy for the next few years. If its economic forecasts are correct, it will end QE in late 2014 and begin to raise interest rates in 2015Q2. It must then take interest rates back to their neutral level (3½-4%) by the time unemployment falls to the NAIRU. The FOMC thinks this will happen in late 2016 or 2017. This would imply around 350bps of interest-rate hikes over 18 months – remarkably similar to Greenspan’s ‘measured’ pace of tightening in 2004-06. But while this is the plan, it is subject to substantial risks and uncertainties. The Fed has said, while it is normalizing policy, it will not sell the assets it bought during QE. This means its balance sheet will only shrink if it allows these assets to mature (i.e. it stops reinvesting them). The FOMC wants to focus on rate increases because (i) it understands better how these influence the economy; (ii) it can reverse these rate increases if needed; and (iii) higher rates would provide a powerful signal that policy is returning to ‘normal’. But this means the Fed’s balance sheet will remain bloated for years to come, creating new challenges for monetary policy. In particular, it will make it harder for the Fed to assess its overall policy stance, meaning it is more difficult to judge how quickly it should raise rates and when exactly to stop. Compared with pre-crisis benchmarks, the Fed’s forecasts imply loose policy for a prolonged period. This would provide the perfect environment for new asset-price bubbles. Excess global savings could further inflate these bubbles by keeping US bond yields down in a repeat of the 2005 ‘conundrum’. Meanwhile, we are less worried about rising inflation because there is still plenty of slack in the economy, the dollar is likely to be strong and the global backdrop remains deflationary. Taking all of this together, there is a real danger that the Yellen Fed repeats Greenspan’s 2004-06 mistake. To avoid this, it would have to become more ‘macro-prudential’ by raising interest rates more aggressively and/or directly draining liquidity. This would hurt market sentiment in the short term, but could ensure a more sustainable recovery. Right now, neither the FOMC nor markets seem prepared for such a policy shift.

1. Escape plan Contrasting errors In 1994, an opaque Fed tightened monetary policy earlier and faster than markets had expected, causing an aggressive sell-off in fixed-income markets. This happened partly because a tepid economic recovery quickly turned into a boom and partly

LSR Macro Picture | April 10, 2014

2

because the Fed was bad at signalling its intentions to investors. Back then, central banks generally thought transparency was undesirable – in fact, 1994 was the first time the FOMC had even announced an interest-rate rise1. As a result, bond yields shot higher (Chart 2) and for a while this threatened to undermine the recovery. Chart 2: No 1994-style selloff 1994-95 (LHS)

2013-14 (RHS)

9

5.5

10-year US bond yield 8

4.5

7

3.5

6

2.5

Days from yield trough 5

1.5 0

30

60

90

120

150

180

210

240

270

300

Source: Bloomberg, LSR estimates

Chart 3: Market underestimated 2004-06 policy tightening 2-year bond yield

Fed funds target

6

per cent 5 4 3 2 1 0 04

05

06 Source: Bloomberg, LSR estimates

In 2004-06, the market environment was totally different. Alan Greenspan was so worried about repeating the ‘disorderly’ 1994 experience that he raised interest rates only tentatively. Throughout this gradual tightening cycle, markets consistently under-estimated the number of rate hikes in the pipeline. For example, the 2-year yield simply tracked the Fed funds rate higher, all the while suggesting just a few

1

For more on the comparison with 1994, see Yield Swerve, 27 June 2013.


3

more rate hikes. Loose monetary conditions and strong inflows of foreign capital from the emerging economies combined to drive excessive risk-taking in financial markets and a housing boom, both of which spectacularly collapsed after 2007.

This time is different? Some FOMC members, such as James Bullard, acknowledge these past errors and have promised the next tightening cycle ‘will be different’. Ben Bernanke believed greater transparency was the best way to achieved this, a view his successor Janet Yellen clearly shares. They hope, by setting out exactly how they intend to tighten policy and carefully explaining their ‘reaction function’ (how they will respond to new information), they can prevent the kind of market mispricing that caused problems in the past. Investors will be able to work out when and how much the Fed will tighten. Table 1: Fed taper schedule Announced monthly asset purchases, $bn

Meeting

Treasurys

MBS

Total

December

40

35

75

January

35

30

65

March

30

25

55

April

25

20

45

June

20

15

35

July

15

10

25

September

10

5

15

October

5?

0

5?

December

0

0

0

Source: LSR estimates. Note: each reduction in QE happens one month after the announcement

Fed ex-it plans With this in mind, Yellen and her colleagues are providing clear and unprecedented guidance about how they expect policy to change over the next few years – their ‘exit strategy’. We are now in the first stage of this, tapering QE. If the economy performs in line with their forecasts, we can expect this strategy to evolve as follows 2:  Stage 1: QE tapering. Bernanke suggested the Fed will reduce its asset purchases by $10bn ($5bn off Treasurys and $5bn off MBS) at each policy meeting this year. As Table 1 shows, this would mean ending QE in late 2014. It’s hard to be precise about the date because on this schedule the Fed will be left buying $5bn of Treasurys after its October meeting. With no meeting in November, it could either cut this last $5bn in December or end QE in September with a $15bn taper. Timing: now until late 2014

2

As outlined in the FOMC’s June 2011 minutes, updated in June 2013.


4

 Stage 2: Allow balance-sheet shrinkage. Currently, the Fed prevents its asset holdings from shrinking by reinvesting these securities as they mature. When the FOMC originally set out its ‘exit strategy principles’ in 2011, it said it would start to tighten policy by halting this reinvestment process. This would allow the Fed’s balance sheet to gradually shrink. Timing: early 2015?  Stage 3: Signal the first rate increase. At ‘the same time or sometime thereafter’, the FOMC says it will use its policy guidance to signal the start of the rate hiking cycle. It could also initiate temporary reserve operations that would ensure it can technically raise interest rates when needed. (We discuss this in more detail below.) Timing: early 2015?  Stage 4: Start hiking the Fed funds rate and the deposit rate. Janet Yellen recently signalled that she will start to raise interest rates around six months after QE ends. This suggests rate hikes will begin in 2015Q2. Raising the interest rate charged on banks’ reserves will allow the Fed to keep short-term rates trading close to their desired Fed funds rate. Timing: 2015 Q2  Stage 5: Stop hiking when rates reach neutral. This is not an explicit part of the Fed’s guidance, but we can infer it from the FOMC’s long-term forecasts. Ideally, policy tightening should end when the economy’s output gap is closed (unemployment is back to the NAIRU). The FOMC thinks this will happen in late 2016. At this point, to prevent inflation rising, the Fed might need to have interest rates back at their ‘neutral’ level – the rate that neither stimulates nor constrains demand. Given most FOMC members think the neutral rate is around 3½-4%, this would imply around 350bps of rate hikes between 2015Q2 and late 2016/early 2017. But according the latest Fed minutes, the pace of tightening might be even more gradual than this3. Timing: ends late 2016-2017  Stage 6: Asset sales. The FOMC currently has no plans to sell the securities it acquired while conducting QE Timing: uncertain (never?)

The Fed is keen to stress that these plans and timeline are conditional on how the economy evolves and are ‘not on a pre-set course’. As Charts 4 and 5 show, the Fed is expecting a decent pick up in US growth later this year, which will lead to a further signification improvement in the labour market. Similar to our own forecasts, most

3

In fact, the latest FOMC statement says: ‘The Committee currently anticipates that, even after employment and inflation are near mandate-consistent levels, economic conditions may, for some time, warrant keeping the target federal funds rate below levels the Committee views as normal in the longer run’.


5

FOMC members expect sustained 3% growth from the summer. If the actual data turns out very different from this projection, the exit strategy will change. Chart 4: Fed's unemployment forecasts Unemployment rate

End QE threshold (lapsed)

Rate hike threshold (lapsed)

Fed's NAIRU estimate

11

FOMC forecast

per cent of labour force

10 9 8 7 6 5 4 05

06

07

08

09

10

11

12

13

14

15

16

Source: LSR interpolation of Fed end-year forecasts

Chart 5: FOMC growth forecasts Lowest 4

Central tendency

Highest

percentage change on year earlier

3 2 1 0 -1 -2 -3 -4 -5 06

07

08

09

10

11

12

13

14

15

16

LT

Source: LSR interpolation of FOMC year-end forecasts, BEA

Taper jammed While future interest-rate hikes are data dependent, the timetable for tapering looks fairly fixed. The FOMC has already ignored weak data and now seems determined to end QE even if the economy remains subdued. There are various reasons for this. First, having created significant volatility last summer, policymakers do not want to jeopardise the current calm in bond markets by injecting new uncertainty. Second, postponing even one instalment of the taper could push the whole schedule back significantly. Third, the FOMC is worried about the longer term costs of QE,


6

particularly its potential impact on financial stability. By contrast, it is more relaxed about the ‘unintended consequences’ of low interest rates, so has altered the composition of the stimulus its provides.

2. Bloatation device While the Fed’s exit strategy should now be clear, investors must remember this is just a forecast and forecasts can change. Indeed, there are substantial uncertainties and risk to this outlook, as Chart 6 makes clear. Even if most FOMC members have signed up to this exit strategy, there are substantial disagreements on the FOMC about what the ‘appropriate’ levels of interest rates will be during the tightening cycle. This is because monetary policy faces a number of challenges that will make exit more difficult than it was in the past. Chart 6: FOMC disagreements about 2016 6

appropriate fed funds rate, year end (dots mark individual FOMC members' views) 5 4 3 2 1 0

2014

2015

2016

long term Source: Federal Reserve

Challenge 1: Monetary policy with a large balance sheet The exit strategy no longer includes asset sales. This means the Fed’s balance sheet will only shrink to the extent that it allows this to happen passively, by not reinvesting Treasurys and MBS as they mature. While the original exit strategy (as set out in the June 2011 FOMC minutes) did include asset sales, the committee recently adjusted this and said it would instead focus on interest rates. There are a number of advantages to tightening through interest rates rather than asset sales. First, the FOMC has a better understanding of how interest rates affect the economy, whereas the impact of Quantitative Tightening (QT) is unknown. It doesn’t feel it can reliably calibrate a programme of asset sales. Second, if something goes


7

wrong and the FOMC needs to reverse its policy tightening, it would be easier to cut interest rates rather than start buying assets again. Raising interest rates would also provide a signal to the private sector that monetary policy and the economy are returning to normal, which could have a positive impact on confidence. But, of course, the situation over the next few years will not be normal. The flipside of not selling these assets is that the Fed’s balance sheet will remain bloated for many years to come (Chart 7). Chart 7: The Fed's balance sheet Treasurys 2.5

MBS

$ trillions

2.0 1.5 1.0 0.5

QE3

taper

run off

0.0 08

09

10

11

12

13

14

15

16

17

Source: Federal Reserve, primary dealers survey, LSR estimates

And while exit should be technically easier using interest rates rather than asset sales, this larger balance sheet will make it more difficult to judge the appropriate level of interest rates. This is because these asset holdings are also providing stimulus to the economy, but in a way that is difficult to measure. This will be a problem not only when the Fed is trying to raise interest rates at the ‘right’ pace, but also when it is deciding when to stop. The economy’s ‘neutral’ interest rate might be higher with a larger balance sheet because the Fed would have to raise interest rates more to offset the impact of the assets it is holding. To get a rough idea of this, we can update some estimates provided by William Dudley, the president of the New York Fed, in May 2012. He suggested the Fed’s first two QE programmes had reduced long-term interest rates by around 150-200bps. Taking the average and running these estimates forward to include QE3 suggests the effective Federal funds rate is probably around 290bps below its headline level (Chart 8). But the critical point is that these estimates are imprecise and the FOMC will not be able to rely on them as a guide to policy.


8

Chart 8: Asset holdings provide additional stimulus Fed funds adj for QE

Fed funds rate

12


10 8 6 4 2 0 -2 -4 85

87

89

91

93

95

97

99

01

03

05

07

09

11

13

Source: Federal Reserve, LSR estimates based on William Dudley speech

Challenge 2: Monetary overhang QE has greatly expanded the Fed’s assets and the counterpart of this has been a huge increase in central bank reserves. The Fed bought Treasurys and MBS by creating new reserves in the banking system. Chart 9 shows these reserves have increased rapidly, causing a sharp expansion in ‘narrow’ money. Chart 9: QE caused huge expansion in the monetary base Reserves held with Fed

Monetary base

4.0

$ trillions

3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 06

07

08

09

10

11

12

13

14

Source: Federal Reserve, LSR estimates

Normally, of course, such a rapid increase in liquidity might fuel inflation fears. But those fears were unfounded because tight credit conditions and sustained deleveraging prevented the increase in narrow money from feeding an expansion in broad money. Instead, the ratio of M3 to M1 collapsed (Chart 10). Still, the Fed will


9

need to watch these developments closely. If growth picks up and bank lending accelerates, this monetary overhang could spill over into much stronger broad money. Chart 10: Collapse in money velocity Narrow money (M1)

Broad money (M3)

ratio of M3 to M1

16

10

$bn

14

9

12

8

10

7

8 6

6

5

4

4

2 0

3 80 81 82 83 85 86 87 88 90 91 92 93 95 96 97 98 00 01 02 03 05 06 07 08 10 11 12 13 Source: Federal Reserve, LSR estimates

In principle, the FOMC has the tools it needs to mop up this liquidity. It could: (i)

Hike the deposit rate: Congress has granted the Federal Reserve the power to pay interest on all reserves. As it raises its policy rate it can immobilize the current ‘monetary overhang’ by also increasing the rate it pays on these reserves. Private banks will not want to undercut the Federal funds rate if they are charged more on their deposits;

(ii)

Directly drain liquidity: The Fed could use a number of tools, such as term deposits and reverse repos, to reduce the supply of reserves.

(iii)

Sell some of the assets it holds: Less likely given the problems we’ve already discussed, the FOMC could reduce banking-sector reserves by selling Treasurys and MBS. Alternatively, the Treasury to could ‘overfund’ its fiscal deficit by issuing more bonds than it needs.

Some of these tools have been used before (e.g. during the money supply targeting of the 1970s) but not on the scale that might be needed over the next few years. Even if they are successful, they could cause significant volatility and uncertainty in financial markets make the Fed’s task more difficult.

Challenge 3: Fiscal costs of exit By holding the assets it acquired during QE to maturity and paying a higher interest rate on reserves, the Fed is likely to experience a sharp deterioration in its net. This means it annual remittances to the Treasury will decline while, simultaneously, it will


10

be forced to make larger interest payments to banks. From a monetary policy perspective, this doesn’t matter, but it will clearly expose the institution to political pressures. Official estimates suggest remittances could decline from $90bn a year in 2013, to zero during the exit period.

3. Exit wounds Since these challenges are likely to make the Fed’s ‘exit strategy’ more difficult than in past cycles, investors should be sceptical of FOMC claims that it can avoid previous mistakes just by being more transparent. In fact, Janet Yellen’s ‘optimal control policy’ suggests a clear bias, one that could end up looking like Greenspan’s policies after 2004. In brief, Yellen has argued that the FOMC will achieve a better balance between unemployment and inflation (its dual mandate) if it focuses on reducing the unemployment rate, even if this allows inflation to temporarily overshoot its target. Chart 11: Staying behind the curve? Fed funds adj for QE

Taylor rule

Fed funds rate

12

FOMC forecast


10 8 6 4 2 0 -2 -4 85

87

89

91

93

95

97

99

01

03

05

07

09

11

13

15

17

Source: Federal Reserve, LSR estimates

This approach is also implicit in the Fed’s policy guidance. Chart 11 compares its exit strategy to a simple Taylor rule. Even ignoring the Fed’s bloated balance sheet, the FOMC is planning to keep interest rates well below ‘normal’ levels. In fact, when the FOMC starts to tighten policy in 2015, it will be doing so at a pace alarmingly similar to Greenspan’s last tightening cycle. The Fed’s forecasts imply around 350bps of rate increases over 18 months, or 20bps per month. Greenspan’s much-derided ‘monotone’ tightening involved raising rates by 25bps per month.

End game: bubbles vs. inflation If monetary policy stays as loose as these forecasts imply, there is clearly a chance this leads to inflation and/or asset-price bubbles. We think bubbles are the bigger threat and, with neither equities nor housing cheap by historical standards, it would take


11

long for over-valuation to become a problem. Meanwhile, inflation seems less of a worry. There is still plenty of slack in the US economy, the dollar could rise, and China’s weakness will intensify deflationary pressures in the rest of the world.

Conundrum II Alan Greenspan always blamed low bond yields for the housing bubble in the mid2000s, arguing that they failed to respond to higher Fed rates. Excess global saving, particularly from the emerging economies, flowed into Treasurys keep yield low. Given the global saving rate is even higher now than it was back then (Chart 12), it is possible we could see a new conundrum as Janet Yellen tightens policy. This is particularly likely in a scenario where China’s economy is weak and the Chinese authorities continue to liberalize their capital account, allowing greater outflows. Chart 12: The conundrum mark II World saving rate 26

US real bond yields* 10

per cent

9

25

8

24

7

23

6 5

22

4

21

3

20

2 1

* deflated by 5-year average of core PCE deflator

19

0 83

85

87

89

91

93

95

97

99

01

03

05

07

09

11

13

Source: IMF, LSR estimates

In this context, the Fed’s excessive attempts to provide transparency could prove counterproductive, especially if they give investors a false sense of security.

A macro-prudential Fed? If asset-price bubbles are likely, we must consider the possibility that the Fed eventually changes its exit strategy in an attempt to protect financial stability. After all, when these bubbles burst they will have a significant impact on the economy and make it harder for the FOMC to meet its objectives. This was a point Jeremy Stein recently made in an FOMC speech. He urged his colleagues to pay more attention to market risk premia, specifically in the bond and credit markets. When these fall to very low levels, as is typical during boom periods, they are prone to sudden, sharp reversals that can have a substantial impact on the wider economy. Stein doesn’t claim the Fed should always lean against such risks, but argues that the benefits to doing so will increase as unemployment falls back towards more ‘normal’


12

levels. This could mean, by the time the FOMC is prepared to start raising interest rates, it might hike more quickly than its guidance currently suggests. Chart 13: Term premium in US Treasurys 2.0

percentage points, derived from 10-year yield

1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 03

04

05

07

08

10

11

13

Source: Federal Reserve, based on Kim and Wright (2005) model

Market impact From a short-term market perspective, this would be a bad news because it would imply tighter monetary policy without stronger economic growth. This is quite different from a scenario in which stronger data encourages the FOMC to raise rates more quickly. Tightening on the back of stronger growth would hurt bonds but not equities. In contrast, ‘macro-prudential’ tightening could hurt both stocks and bonds. While such policies might ultimately deliver a more balanced and lasting recovery, this won’t be the first thought on many investors’ minds. ‘Fortunately’ there is little evidence the FOMC is becoming macroprudential. While such concerns probably influenced the committee’s desire to taper QE despite relatively sluggish growth in early 2014, most FOMC members seem reluctant to use interest rates to target financial risks. Jeremy Stein has recently resigned from the FOMC and most of his colleagues seem to believe that ‘regulatory tools’ – not monetary policy – would be better for targeting financial risks. Yet, five years after the crisis, it is not yet clear what these tools might be. And while, in principle, the Fed could use asset sales to damp rapid gains in asset prices (e.g. selling MBS to slow the housing market), this remains an unlikely prospect given the technical difficulties.

Bottom line The FOMC believes greater transparency can help it to normalize monetary policy without repeating the mistakes it made in the past. It has provided clear guidance about how it will proceed over the next few years. Yet post QE, the Fed’s bloated balance sheet should mean ‘exit’ will be even trickier than in the past. The chances of getting this exactly right seem slim and right now Ms Yellen and her colleagues have


13

a clear bias, promising extremely loose policy over the next few years. As occurred with Alan Greenspan’s last tight cycle, this could provide the perfect environment for new bubbles.

Dario Perkins [email protected]

Due to the Easter holidays, the next macro picture will be published on 8 May. For a short summary of our views on the global economy click here.


14

Disclaimer This report has been issued by Lombard Street Research Financial Services Limited. It should not be considered as an offer or solicitation of an offer to sell, buy, subscribe to or underwrite any securities or any derivative instrument or any other rights pertaining thereto (“financial instruments”) or as constituting advice as to the merits of selling, buying, subscribing for, underwriting or otherwise investing in any financial instruments. This report is intended to be viewed by clients of Lombard Street Research Financial Services Limited only. The contents of this report, either in whole or in part, shall not be reproduced, stored in a data retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without written permission of Lombard Street Research Financial Services Limited. The information and opinions expressed in this report have been compiled from publicly available sources believed to be reliable, but are not intended to be treated as advice or relied upon as fact. Neither Lombard Street Research Financial Services Limited, nor any of its directors, employees or agents accepts liability for and, to the maximum extent permitted by applicable law, shall not be responsible for any loss or damage arising from the use of this report including as a result of decisions made or actions taken in reliance upon or in connection with the information contained in this report. Lombard Street Research Financial Services Limited does not warrant or represent that this report is accurate, complete or reliable and does not provide any assurance whatsoever in relation to the information contained in this report. Any opinions, forecasts or estimates herein constitute a judgement as at the date of this report based on the information available. There can be no assurance that future results or events will be consistent with any such opinions, forecasts or estimates. Past performance should not be taken as an indication or guarantee of future performance, and no representation or warranty, express or implied is made regarding future performance. This information is subject to change without notice, its accuracy is not guaranteed, it may be incomplete or condensed and it may not contain all material information concerning the company and its subsidiaries. The value of any securities or financial instruments or types of securities or financial instruments mentioned in this report can fall as well as rise. Foreign currency denominated securities and financial instruments are subject to fluctuations in exchange rates that may have a positive or adverse effect on the value, price or income of such securities or financial instruments. Certain transactions, including those involving futures, options and other derivative instruments, can give rise to substantial risk and are not suitable for all investors. This report does not have regard to the specific instrument objectives, financial situation and the particular needs of a client. Clients should seek financial advice regarding the appropriateness of investing in any of the types of financial instrument or investment strategies discussed in this report. Lombard Street Research Financial Services Limited may have issued other reports that are inconsistent with, and reach different conclusions from, the information presented in this report. Lombard Street Research Financial Services Limited is Authorised and Regulated by the UK Financial Services Authority. FSA Firm Reference Number: 502674. Registered Office: 9 Cloak Lane, London EC4R 2RU. Registered in England No. 6862824


15

Since 1989 Lombard Street Research has provided the global investment community with independent, thoughtful and pertinent macroeconomic analysis, investment advice and asset allocation recommendations. Our clients value our genuinely global and fully integrated research, which is essential in the modern investment environment. Global-macro thinking drives all of our investment themes. For more information about our services please contact your local Lombard Street Research office.

Pellipar House, 9 Cloak Lane, London EC4R 2RU www.lombardstreetresearch.com

London

Tel: +44 (0) 20 7246 7870

[email protected]

New York

Tel: +1 212 367 7644

[email protected]

Hong Kong

Tel: +852 2521 0746

[email protected]