The many reasons for a weak US dollar

See the latest Australian dollar analysis here:

Macro Afternoon

Via Bank of Tokyo-Mitsubishi in January:

USD: hard to see a clear trigger to reverse near-term sell off

The US dollar has continued to weaken during the Asian trading with the dollar index breaking below the 90.00-level for the first time since the end of 2014. US dollar weakness at the start of this year has been reinforced by heightened concerns over the ongoing shift to more protectionist US trade policies. Our analysts in Hong Kong put together the following report (click here) which assesses in more detail the recent decision by President Trump to introduce new tariffs on solar panels and washing machines. Market participants will be closely scrutinizing upcoming comments from President Trump at Davos, and in his Sate of Union address to assess if trade will be a more important policy focus this year ahead of the mid-term elections.

The US dollar sell off has been relentless since the middle of last month, and there appears no clear catalyst on the horizon to turn the tide in the near-term. It is notable that the US dollar has continued to weaken over the last month or so even as US yields have moved higher. It highlights that the US dollar has become less sensitive to Fed rate hike expectations and the performance of the US economy.

The Fed could acknowledge recent favourable developments at next week’s FOMC meeting although we doubt they will prove sufficient to trigger a shift to their current gradual policy normalisation plans at the current juncture. The US rate market has already shifted to price in a higher likelihood of the Fed delivering the further three hikes they are currently planning for this year. The US dollar is likely to prove more sensitive to the following FOMC meeting in March which will be the first under new Fed Chair Jerome Powell. The Senate confirmed Jerome Powell as the new Fed Chair overnight. The market will be looking for confirmation of current expectations for policy continuity under Chair Powell. In these circumstances, there is a building risk that US dollar weakness could even begin to overshoot fundamentals in the near-term. A repeat of price action from back in early 2004 can’t be completely ruled out. After breaking back above the 1.2000- level late in 2003, the euro continued to strengthen sharply against the US dollar towards the 1.3000 level in the opening months of 2004 before settling back into the 1.2000 to 1.2500 range, which is around where we currently estimate long-term fair value for the pair.


The relationship isn’t perfect, but as a general rule of thumb, the USD declines in years when global growth is above potential. In such years, strong global growth causes commodity prices to rise, which lifts commodity currencies. In addition, strong global growth normally triggers a flood of investment out of developed economies and into emerging economies. As emerging central banks intervene to slow the appreciation of their currencies from these two factors, they sell the USD against EUR and other alternative reserve currencies, thereby causing the USD to decline against both developed and emerging currencies. That dynamic helps explains the USD depreciation in 2017 as well as the 2004-2007 period. The 10Y average of the IMF’s World GDP growth rate is 3.4% and we think that is roughly potential growth. The IMF estimates that global growth will come in at about 3.6% for 2017 and accelerate to 3.7% in 2018. In addition, the US’s twin deficit fundamental (sum of the current account deficit and fiscal or federal budget deficit as shares of GDP) is another factor that is negative for the USD. Turns in the twin deficit normally precede turns in the USD by 1-2 years and then trends match thereafter. The US’s twin deficit fundamental has been deteriorating for the past two years and is likely to deteriorate further in 2018 and beyond due in part to the tax cuts. When looking at all these factors together with the fact that USD phases tend to last 5-7 years, we feel that we have to forecast a continuation of broad USD weakness—albeit at a slower pace. We project that the broad USD index will fall 1.5% in Q1 and then 1.0% per quarter in each of the remaining three quarters of 2018.


Parallels to Clinton in the early 1990s Softening a strong dollar policy is one thing, but actively talking down the dollar is quite another. There is some precedence here. A newly-elected President Clinton tried to prise open the Japanese auto-part market in 1993/94. His Treasury Secretary, Lloyd Bentsen, opined how a weaker dollar would help narrow the US trade deficit with Japan. That policy soon got out of hand, however. Not only the dollar but also US Treasuries and equities sold off and a ‘Sell America’ mentality was born. It took well over a year, much FX intervention, coordinated interest rate moves between the Fed and the BoJ and ultimately the birth of the strong dollar policy under a new Treasury Secretary (Robert Rubin) in April 1995 to stop the dollar rout. Such ‘Sell America’ signs have been sporadic in recent years.

Our ‘Sell America’ Market Sentiment Tracker – which identifies sustained periods of a co-ordinated downturns in US assets (ie, lower stocks, higher bond yields and lower US dollar) – shows that we remain well below the degree of pessimism expressed during the early 90s Clinton administration. Looking ahead, we note that a stealth environment of ‘Sell America’ can exist as the Fed tightening cycle enters its latter stages – drawing parallels to the 2005-2006 market backdrop. This would be consistent with our outlook for gradually rising long-dated US Treasury yields – and strategically bearish US dollar – in 2018. 26 January 2018 2 However, for us to really get back to the ‘Sell America’ sentiment seen during periods in the 1980s and 1990s, we would need to see one of two things: (1) the Trump administration’s ‘America First’ ideology leading to the implementation of more meaningful protectionist policies or (2) a structural shift in the stock-bond correlation (which for other reasons beyond this note looks unlikely in the current ‘lowflation’ world). The first factor is the real risk in the near-term – and the next couple of weeks could be a watershed moment for world trade and protectionism, with President Trump’s Davos (26 Jan) and State of the Union (30 Jan) speeches likely to set the tone for US trade policy over the coming year.


We are well into 2018 and our feedback from recently attending the TradeTech FX conference in Miami is that the market is still struggling to understand or embrace dollar weakness. How can it be that US yields are rising sharply, yet the dollar is so weak at the same time? The answer is simple: the dollar is not going down despite higher yields but because of them. Higher yields mean lower bond prices and US bonds are lower because investors don’t want to buy them. This is an entirely different regime to previous years.

Dollar weakness ultimately goes back to two major problems for the greenback this year. First, US asset valuations are extremely stretched. As we argued in our 2018 FX outlook a combined measure of P/E ratios for equities and term premia for bonds is at its highest levels since the 1960s. Simply put, US bond and equity prices cannot continue going up at the same time. This correlation breakdown is structurally bearish for the dollar because it inhibits sustained inflows into US bond and equity markets.

The second dollar problem is that irrespective of asset valuations the US twin deficit (the sum of the current account and fiscal balance) is set to deteriorate dramatically in coming years. Not only does the additional fiscal stimulus recently agreed by Congress push the fair value of bonds even lower via higher issuance and inflation risk premia effects, but the current account that also needs to be financed will widen via import multiplier effects. When an economy is stimulated at full employment the only way to absorb domestic demand is higher imports. Under conservative assumptions the US twin deficit is set to deteriorate by well over 3% of GDP over the next two years.

The mirror image to all of this is that the flow picture into both Europe and Japan has been improving dramatically anyway. We have previously written about the positive flow dynamics in Europe as the flow distortions caused by extremely unconventional ECB policy are starting to adjust. But the Japanese basic balance has also shot up to a 4% surplus in recent years helped by a big improvement in the services balance (Chinese tourists) and a collapse of Japanese inflows into the US: treasuries simply do not provide enough duration compensation any more. To conclude, embrace dollar weakness, it has more to run.


But the weakness of the USD in response to structural considerations means that the cyclical story for the USD may not be as boring as it seems. We have argued for some time that the role of FX within monetary policy has changed. It is no longer simply a consequence of interest rate hikes or cuts. Its level can determine the scale and pace of those moves, and it can often act as a substitute for interest rate shifts.

In 2014, this meant a strong USD acted as a test-bed for policy tightening, eventually prompting the Fed to tighten only once in 2015 and again in 2016. The USD weakness we have seen since the last FOMC meeting may act in the opposite direction, galvanising hawks to hike three or more times and reassuring doves that inflation is likely to track back to target. A weak USD could become a more visible part of the Fed’s rate debate, potentially prompting a revival of cyclical forces as a determinant of the USD’s level. In 2014, the prospect of higher rates drove the USD higher. In 2018, the prospect of a weaker USD may drive rates higher.

Ironically, if this happened, it could prompt a snap back stronger in the USD. Chart 5 shows HSBC’s model of where an exchange rate should be based on both the level and slope of the yield curves of both currencies. We fit the model using the data from 13-months ago to 1-month ago. We then use the model parameters found over that 1-year period to calculate the FX moves expected by the model in response to changes in the swap rates over the last month. In the charts we plot these model-implied FX moves in black. We show 5th -95th percentile error bars as the grey shaded area.

We can then compare the actual moves in the exchange rate (plotted in red) with this range. When the red line is outside of the grey shaded area it tells us that the moves seen in the FX market have not been reflected in the swap rates. The results show that the USD looks notably cheap relative to interest rates in all instances. The results for NZD, CAD and CHF are similar. It suggests if cyclical factors for the USD can regain traction, the USD will have much lost ground that it can make up. We accept the circularity here. USD weakness may prompt the Fed to become more confident about raising interest rates and revive the cycle as a driver to the USD resulting in a stronger USD.

I find the capital outflows to EMs, Japan and Eruope the most convincing argument. But ultimately agree with HSBC. Especially so given my view that Europe and China are going to slow first this year, creating a big headwind for commodites, leaving the USD as the primary growth driver and high yielder.

David Llewellyn-Smith
Latest posts by David Llewellyn-Smith (see all)

Comments are hidden for Membership Subscribers only.