Investors assume a strong link between central bank balance sheets and currencies. This may oversimplify things. While balance sheets matter in some cases (USD/JPY), elsewhere the link seems spurious (EUR/USD) or non-existent (GBP/USD). The differences don’t always depend on whether Quantitative Easing is in place
Residents of the United States, please read this important information before proceeding
As interest rates approached zero in all major economies, their influence as a monetary tool became muted. Ever-inventive central banks then turned to new tools and started intervening in markets via channels of different sorts – most obviously via purchases of government bonds (“Quantitative Easing” in Japan, the US and UK) or large-scale lending to commercial banks (“Long Term Refinancing Operations” in the euro area). These unconventional measures were seen as the “next level” of monetary easing.
As with all monetary easing, the effects on respective currencies should be negative. Indeed, the received wisdom dictates that the larger the country’s (or monetary union’s) balance sheet, the larger the downside risk to the currency. However, we believe that this view largely oversimplifies the realities of the monetary world.
First, not every balance sheet expansion is the same. Comparing the loans and sterilised bond purchases made by the ECB with the unsterilised purchases of the Fed is equivalent to comparing apples and pears. They are simply not the same. In the first case, the aggregate money supply is unaffected. In the second, the central bank is effectively “printing money”. But even if the types of balance sheet expansions are similar (such as QEs in the US and UK), they still may not be the key drivers of exchange rates. Below we explore the relationship between growth in central banks’ balance sheets and three main FX crosses: EUR/USD, GBP/USD and USD/JPY.
The sensitivity (beta) of FX to central banks’ balance sheets is small, unstable over time and statistically insignificant
EUR/USD, ECB and Fed – comparing the incomparable
On first sight (Figure 1), there appears to be a relationship between the relative balance sheet expansions and EUR/USD. However, closer inspection reveals that this relationship has been very unstable over time (see for example the rolling 26-week beta, or sensitivity, in Figure 2). In our view, this instability stems from two factors. First, the different nature of balance sheet expansions between the ECB and the Fed. While the Fed expanded its balance sheet largely by buying bonds, mostly from banks, thereby contributing to growth in the money supply (printing money via QE1, QE2 and now QE3), the ECB has not. When the ECB has bought bonds its actions have mostly been offset (“sterilised”) by sales elsewhere, and much of its balance sheet expansion has been the straightforward making of loans. Of course, balance sheet expansion that is not associated with money printing is unlikely to be as negative for a currency.
The second difference between the two is one of timing and context. While the Fed’s balance sheet expansion (or market expectation of that) weakened the dollar in the past (as it stabilised financial markets at a time when the USD was a safe haven, prompting USD-funded trades), the ECB’s indications that it may expand its balance sheet (for example, via LTROs or OMTs) had in fact a positive effect on the euro because the ECB’s potential actions were seen as a backstop for the euro’s existential worries, reducing the risk of a collapse.
Figure 2: Beta of EUR/USD seems unstable
The second half of 2011 provides a case in point. The EUR/USD collapsed first as a result of market’s grave concerns about the viability of the region’s banking system. The fall in the euro area financial indicators (EUR/USD including) was followed by the ECB announcement of the LTRO program (cheap funding to euro area banks). As a result, the euro stabilised and rebounded from the 1.27 level it reached by mid-January 2012 to above 1.30 and stayed there until May 2012. Here, the EUR rebounded despite the fact that the ECB announced a material balance sheet expansion (the size of the two rounds of LTROs exceeded EUR 1 trillion). The same could be said about the Greek crisis in May/June last year, which was followed by the announcement of OMT program (whereby the ECB may buy short-term government debt of countries in need). This promise of further balance sheet expansion (via OMT) actually led to a rebound in EUR/USD from 1.23 to 1.37 in early February 2013 (rather than a collapse).
Hence, we remain cautious of using relative balance sheet movements on their own to try to predict EUR/USD moves. Not only have the ECB and the Fed so far engaged in very different types of balance sheet expansions, but the ECB’s balance sheet expansion has been lately employed as a response to deteriorating euro area financial conditions. It acted as a systemic support for EUR, rather than a cyclical damper.
Figure 3: GBP/USD and central banks balance sheets
Figure 4: Beta of GBP/USD is almost non-existent
GBP/USD and balance sheets – other things matter more
While there has been some loose and unstable relationship between EUR/USD and the relative sizes of the balance sheets of the ECB and the Fed, there seems to be no relationship at all for the pound and the BoE’s balance sheet (Figure 3). The sensitivity (beta) of FX to central banks’ balance sheets is small, unstable over time and statistically insignificant (Figure 4).
In our view, this weak relationship is simply because other things appear to matter more for GBP. These have driven the currency’s behaviour independently of the forces stemming from relative balance sheet expansions. In particular, the euro area crisis in late 2011 and mid-2012 kept GBP supported via safe-haven flows into the currency (due to: sterling’s high liquidity; the attraction of gilts, then rated at AAA; the absence of an alternative given the ceiling placed on the Swiss franc; and regional proximity).
Given the euro area backdrop, GBP “survived” the material, unsterilised BoE balance sheet expansion (via QE) from late 2011 to Q3 2012. Bizarrely, but in line with the above reasoning, the collapse of GBP in the first quarter of this year was in fact accompanied by a materially favourable move of relative balance sheets between BoE and the Fed in favour of sterling (Figure 3). Other factors than the size of central banks’ balance sheets were very visibly at work (namely, a rethink of sterling’s safe haven status).
USD/JPY – where balance sheet expansion does matter
The yen and the BoJ have dominated headlines in recent weeks as the BoJ announced large-scale monetary easing, aiming to double the Japanese monetary base over the next two years (in its quest to reach its 2% inflation target).
As our projections in Figure 5 show, such large-scale balance sheet expansion will put JPY on a different level to USD, EUR and GBP. The BoJ will increase its balance sheet from currently 34% of GDP to more than 60% in two years. Given the ambiguities noted above, the question is whether such large scale money-creating relative balance sheet growth matters for USD/JPY. In this case we think it does. Figure 6 shows that the relative balance sheet expansion between the BoJ and the Fed does have a relationship with USD/JPY. Moreover, we show in Figure 7 that the sensitivity (the “beta”) of USD/JPY to movements in relative balance sheets has been high over time (a correlation of 0.66 from 2010 until today). Moreover, this sensitivity has remained in positive territory (and did not change signs, unlike EUR/USD). Taken at face value, this suggests that for every 1% year-on-year increase in the BoJ’s balance sheet relative to the Fed’s, USD/JPY should appreciate by 0.66%.
Figure 6: USD/JPY and central banks balance sheets
Figure 7: Beta of USD/JPY is significant
Based on a conservative assumption that the Fed will taper all asset purchases in 2014 by an equal amount every month and the already announced pace of BoJ balance sheet expansion, the BoJ balance sheet should increase by around 30% relative to the Fed by the end of 2014. This would suggest a further 20% appreciation in USD/JPY (assuming that beta of 0.66). This may be excessive. Markets are forward looking and, since Q4 2012, expectations of an aggressive monetary response have been building. Following the BoJ announcement on 4th April, the aggressive monetary policy is now the market’s base case. But if a further 20% JPY weakness from current levels appear to be a bit aggressive, such a large scale balance sheet expansion is not supportive for JPY and should at least keep the currency soft. We thus expect JPY to continue struggling and expect USD/JPY to break above 100 over the months ahead.
We expect JPY to weaken further against USD as the ultra-dovish policy at the BoJ (and material QE) should continue weighing on JPY. But with some significant JPY weakening already behind us, investors should be selling JPY against USD during periods of JPY rebounds (such as the one seen on 15th April, when JPY temporarily rebounded from USD/JPY 98 to 96). For clients with existing long USD/JPY positions, we see further room for yen weakening. Our current forecast pencils-in USD/JPY at 103 in six months.
In terms of EUR/USD, we expect it to weaken to 1.25 and 1.23 over the 6- to 12-month time horizon. But we do not expect relative balance sheet expansion to be the main driver of the cross. Rather, the market expectations of the Fed tapering its asset purchases and normalising its monetary stance should support USD against EUR, where the probability of the ECB’s next step is skewed towards further easing. This, coupled with our view of USD turning pro-cyclical, should lead EUR/USD lower.
For GBP/USD, relative changes in central banks balance sheets did not have significant explanatory power in the past. However, with the euro area situation stabilising (and hence taking the safe-haven support away from GBP), it may become a more important driver of GBP/USD. Nonetheless, we expect the diverging fundamentals (both economic and monetary) to weigh on GBP against USD over the quarters ahead.