And just as everyone was starting to bet on the great USD renaissance, here comes Thomas Stolper to spoil the party, by not only refusing to close out his EURUSD trade reco after losing 800 pips in two weeks (and still being profitable), but by actually doubling down: "We have changed our forecasts to project more Dollar weakness." The reason is that the US apparently has a thing called a massive trade deficit that has to be normalized: "Since the last revisions to our forecasts, the Dollar decline has roughly tracked the expected path. Large structural imbalances in the US are highlighted by weakness in the tradable goods sector.The outlook for monetary policy differentials and BBoP trends remains USD-negative. Dollar weakness is common during periods with slowing GLI momentum." The bottom line: "We now see EUR/$ at 1.45, 1.50 and 1.55 in 3, 6 and 12 months, and $/JPY at 82, 82 and 86." Oddly enough, there is no mention of the real reason to position for a USD plunge. (Hint: Hewlett Packard). On the other hand, this may be the time to go balls to the wall long the USD, as it appears that Goldman is doing another USD fundraising campaign courtesy of its clients. Oh, and speaking of Goldman's clients, it's best to baffle them with bullshit. Here is Goldman's Jim O'Neill with a blurb from his Sunday note on why China is going down (among other things): "it seems to me that a bigger risk premia is still necessary for the Euro. I can’t see how it can remain at about 1.40." Yes. From Sunday. If your head didn't go boom yet, that's ok. It will soon enough. And way to cover your bases there Goldman...
- We have changed our forecasts to project more Dollar weakness.
- Since the last revisions to our forecasts, the Dollar decline has roughly tracked the expected path.
- Large structural imbalances in the US are highlighted by weakness in the tradable goods sector.
- The outlook for monetary policy differentials and BBoP trends remains USD-negative.
- Dollar weakness is common during periods with slowing GLI momentum.
- We now see EUR/$ at 1.45, 1.50 and 1.55 in 3, 6 and 12 months, and $/JPY at 82, 82 and 86.
Since we moved to a more explicit Dollar weakening path last autumn, FX markets have broadly followed the expected trajectory. In many cases, the Dollar has now weakened well beyond our near-term forecasts and the driving forces of continued gradual depreciation are intact. We review the key arguments behind our view and focus specifically on Dollar performance in the context of the global business cycle and the latest BBoP trends. Our major FX forecasts are revised to reflect continued further USD weakness.
Dollar Decline To Continue
In a nutshell, the main reason for broad-based Dollar weakness is the persistence of economic imbalances in the US. US economic output remains geared towards non-tradable sectors, while the large current account deficit and the structural decline in manufacturing employment suggests weakness in tradable goods. This situation has several Dollar-negative implications:
- The structural current account deficit causes constant external funding pressures. For the Dollar to stabilise or even to rally, investors need to be convinced of the case for additional long-term investments in the US.
- With unemployment still high, fiscal consolidation looming and continued weakness in the real estate sector, the growth outlook remains less compelling in the US than in many other regions or countries. This makes it even more difficult to fund the current account deficit with investment inflows.
- The cyclical factors discussed in the previous point suggest this it is also highly likely that Fed policy will remain more accommodative than in most other countries. Interest rate differentials will likely remain USD-negative.
- The case for Dollar depreciation will strengthen as fiscal policy becomes increasingly tight in the US. The likelihood of early monetary policy tightening would also decline with tighter fiscal policy, as highlighted by our US economists.
- Structural and EM-related upward pressures on crude prices add to the imbalances. All else equal, real disposable income in the US would decline and, hence, so would domestic demand, adding to cyclical headwinds. Moreover, the rising fuel bill would increase the nominal trade gap and therefore the external funding needs.
None of the above five factors seems to be changing towards a more Dollar-supportive direction. On the contrary, there are increasing depreciation risks linked to structural forces in commodity markets, as well as fiscal consolidation.
Manufacturing, Manufacturing, Manufacturing
Ultimately, it is difficult to envisage a Dollar-bullish scenario without a notably stronger US tradable goods sector. This does include to some extent services and also the primary sector, but the most important by far is manufacturing.
Employment in the US manufacturing sector has fallen substantially in the last 10 years (see chart) and this development was likely the result of two forces:
- First, the competition from interest rate sensitive domestic sectors during the credit boom in the US, in particular real estate related sectors.
- Second, aggressive offshoring and the relocation of factories to the rest of the world, in particular Asia, has led to the disappearance of whole manufacturing industries in the US.
There have recently been some signs of strength in the manufacturing sector, in particular strong ISM surveys and some persistent hiring in the manufacturing sector. But the rate of job growth in this sector remains very low compared with the losses over the last 10 years. From 2001 to the trough of the credit crisis, US employment in the manufacturing sector has fallen by about 5mn to 11mn, at a rate of about 52,000 per month on average. Since then, we have seen renewed hiring of about 13,000 per month on average. In other words, job growth in the US manufacturing sector currently runs at only 25% of the pace of job destruction seen over the last decade.
This weakness in the manufacturing sector is also still clearly reflected in the external balance. The real trade deficit currently runs at about $50bn per month (in 2005 Dollars). Broken down by sectors, auto related and consumer goods sectors account for about $40bn, again highlighting the weakness in tradable goods.
In terms of outlook, our US economists expect a gradual further widening of the real US trade deficit in terms of GDP, which will likely keep the downside pressures for the Dollar firmly in place.
Lastly, it is also important not to mix level and change effects. The Dollar downside pressures will likely subside only after the external deficits have narrowed substantially. In practical terms, this means a substantial amount of manufacturing capacity has to be shifted back to the US, and this is a very slow process that is typically measured in years rather than quarters.
We are confident this adjustment will ultimately happen, but in the meantime it may be necessary for the Dollar to drift lower until relocation to the US becomes a very clear case. The undervaluation of about 12% relative to our trade-weighted GSDEER model may therefore become more pronounced in the foreseeable future.
New Forecasts and Risks
Taken together, the points above suggest there is still considerable downside potential in the USD. We therefore are revising our forecasts to reflect this ongoing trend. In particular, we are now projecting EUR/$ at 1.45, 1.50 and 1.55 in 3, 6 and 12 months. We now see $/JPY at 82, 82 and 86, which, compared with our previous forecasts, also reflects more broad USD weakness—albeit within the recent range.
Our views, as outlined across our macro and market research, remain constructive on Europe. Markets have long expected some form of liability management for Greece, and so a lot of bad news is already priced in. That said, reform progress in systemically important Spain continues at a steady pace. With contagion effects from the Greek debt debate limited, we think the recent correction likely represents an opportunity to position for Dollar weakness versus the Euro.
A number of other forecasts have been revised accordingly, as highlighted in the table. For specific comments on individual currencies, please refer to the country pages in this publication.
As a general feature of this forecast revision, it is important to highlight once again the difference between the expected moves relative to the USD versus the much more muted appreciation on a trade-weighted basis. For example, the Euro is expected to rally about 9% versus the Dollar over the next 12 months, but it will only appreciate by 2.4% on a trade-weighted basis. To a varying degree, this pattern holds for most currencies.
In terms of ‘fair value’, we are calling for a move deeper into USD undervaluation territory. The new 12-month forecasts imply a Dollar undervaluation on a TWI basis of about 16%-17%, whereas the trade-weighted Euro overvaluation will likely remain below 10%.
The risks to our USD bearish view are threefold:
- First, a much faster structural rebalancing of the US economy than we currently expect would fundamentally change the picture, although the hurdle seems very high for this to happen in the near term.
- Second, a period of broad-based asset weakness would likely still translate into Dollar strength given the prevailing correlations.
- Finally, there are still concerns about the European sovereign situation. Although not directly a factor for the US economy, a substantial deterioration of the sovereign debt situation in Europe would support the Dollar, mainly because it would weigh on the Euro.
Anaemic Portfolio Investment in the US
As we stated above, one of the core reasons behind our Dollar view rests on a Fed that is more dovish than other central banks. Our official forecast for the first Fed hike is not until 2013, which is significantly below what the market is currently pricing. The second important factor behind our Dollar views is the growing current account deficit and the possible deterioration in funding inflows. Foreign flows into US assets other than US Treasuries have remained very weak, possibly affected by the negative returns following the ‘tech bubble’ until 2000 and the subsequent housing bubble in 2004/07. To provide a snapshot of this, Table 1 shows Treasury International Capital (TIC) data on long-term portfolio flows into and out of the US. It shows average monthly net foreign flows into US Treasuries (column A), into US agency debt (B), into US corporate debt (C), and into US equities (D).
During the pre-crisis period, for which we use 1H2007 as an example, monthly net inflows into US Treasuries averaged $19bn, inflows into agencies were just shy of $30bn, into corporate bonds they were an impressive $48bn, and into US equities they amounted to $24bn, making a total around $117bn on average per month. That picture has shifted dramatically. On average in 2010, net foreign flows into Treasuries rose to just shy of $60bn per month, while net foreign flows into US agency debt were on average only $9bn per month. Net foreign flows into US corporate debt were down to -$1.1bn on average in 2010, i.e., an outflow, while flows into US equities were just $9bn. Net foreign inflows across assets had therefore shrunk to $76bn on average per month in 2010, from $117bn in 1H2007, and the composition of inflows has shifted dramatically away from agencies, corporate debt and equities, towards US Treasuries. Looking at the most recently released data for March 2011, this shift remains firmly in place, with foreigners continuing to shy away from US assets other than Treasuries. Moreover, US investors have recently accelerated purchases of foreign assets again, with the latest March number hinting at a sizable outflow of more than $30bn from that source. Net portfolio inflows therefore remain very weak in general.
There are two additional issues worth mentioning. First, given the low interest rate environment, we think that hedge ratios for foreign inflows into US Treasuries are now relatively high. This means that these inflows are not as Dollar-positive as a similar inflow into US equities. Second, foreign official buying (largely by central banks in emerging markets) constitutes, on our estimates, the bulk of net foreign purchases of US Treasuries. The official breakout by the TIC data here suggests that—of the $60bn in foreign net purchases of Treasuries per month in 2010—about $44bn reflect foreign official buying. This number is likely a lower bound estimate, since foreign central banks may also be purchasing US Treasuries through intermediaries. We highlight the importance of official buying because these purchases are in some sense ‘passive’, i.e., they reflect the decision by some emerging markets to peg their currencies to the Dollar. As a result, they mechanically have to buy US Treasuries to neutralise appreciation pressure on their currencies. These inflows to the US are therefore not driven necessarily by the same motives as foreign flows into US equities (such as growth expectations and/or the profit motive). For both of these reasons, we see the shift to Treasury purchases by foreigners as a development highlighting the difficulty in funding the US trade deficit, and in line with our view for further Dollar weakness.
To highlight the important role that Treasuries play in funding the US trade deficit, we have compiled the table above to compare non-Treasuries foreign private (non-official) inflows (column K, net also of US residents’ flows into foreign assets, columns F and G) to the monthly pace of the trade deficit (column J). This comparison shows that while private inflows into non-Treasury instruments were used to fully fund the US trade deficit in 1H2007, this is no longer the case currently, as indicated by our ‘private’ basic balance measure (column L). Indeed, that private basic balance for 2011 year-to-date has been on a deteriorating trend, which is one fundamental reason why we see the current Dollar rally as temporary, reflecting risk appetite. On the fundamental flow picture, the outlook remains—in line with our Fed call—for more Dollar weakness.
Currency Performance During Slowing GLI Momentum
Global industrial activity continues to expand, but the pace of growth is declining. As we have discussed in recent research, our Global Leading Indicator (GLI) has exhibited signs of a slowdown over the last few months. In the last print, our metric of GLI momentum (3mth rolling % changes on the GLI) fell to 0.43% from 0.47%, with most of the components in the indicator deteriorating on a monthly basis. This has occurred even though on a yoy basis the GLI continues to accelerate at a solid pace of 4.5%.
What does this slowdown in GLI momentum mean for FX?
To answer this question, we first assess how different currencies behave during times of GLI momentum deceleration. We calculate monthly returns for the most liquid currencies against the Dollar since January 1999; a sample of about 150 observations overall. We then split our sample into months when the GLI is losing momentum and months when it is gaining momentum. Finally, for each currency, we average out the monthly returns for the months that the GLI momentum is declining relative to the month before. This subsample consists of 88 observations corresponding to 59% of the full sample. Comparing average returns across currencies, three key observations emerge:
- Overall, G3 currencies tend to strengthen during times of declining GLI momentum. The USD, EUR and the JPY—as well as the CHF—outperform most other currencies. Interestingly, the JPY, CHF and EUR outperform the USD on average during times of loss of GLI momentum.
- Pro-cyclical EM currencies tend to underperform as the GLI loses momentum. The BRL, CLP and MXN in Latin America, the TRY and ZAR in EMEA, and the IDR and INR in NJA tend to be among the worst performers.
- Interestingly, managed Asian currencies such as the MYR, TWD, CNY and SGD tend to remain largely unaffected by a slowdown in GLI momentum. Indeed, they have strengthened on average vs the USD during those periods. This may well be a spurious result given that the most important determinant of these exchange rates is policy decisions related to local inflation trends.
Overall, there appears to be a strong cyclical split between FX out- and underperformers during times of GLI momentum deceleration. However, it is important to bear in mind that these results may well be distorted by the fact that periods when GLI momentum decelerates include contractions in global output. We are not currently in such a contractionary stage, nor do we expect to be within our forecast horizon.
Therefore, we refine our sampling exercise even further to exclude observations during times of recession. We focus only on times when GLI is losing momentum but global output is still expanding, much like the phase we are in at the moment. This limits our sample to a still large set of 75 observations.
Our results are practically the same but for one key difference: on average, the USD tends to depreciate against most currencies as the GLI loses momentum but global growth is still positive.
What does this mean for our FX views? First, our broad bullish NJA FX bias, which we have expressed through long CNY, MYR and PHP recommendations, still appears to be justified despite the slowdown in GLI momentum. Historical data do not suggest that a deceleration in the pace of global activity growth has a significant effect on managed NJA currencies.
Second, one needs to be cautious with respect to pro-cyclical EM currencies and high yielders during periods of slowing GLI momentum. The Turkish Lira is among these currencies and exposed to a large current account deficit in addition. Slowing momentum therefore would support our tactical long EUR/TRY recommendation.
Finally, and most importantly, our findings reassure us that our overall bearish Dollar stance is not negatively affected by this slowdown in global activity momentum. On the contrary, as long as markets do not start to reflect concerns about an outright decline in global output, historical data of Dollar performance during similar phases of the global cycle support further our forecast for USD depreciation.