The BOJ’s SNB moment? JGBs, USD/JPY and potential negative convexity

As the BOJ steps up its rinban operations to defend its yield cap, it is demonstrating the unusual pro-cyclciality of it’s monetary framework: during times of improving growth expectations and rising yeilds the pace of BOJ liquidity creation is biased to increase, and vice versa. The failure of the era of Abenomics to end Japan’s multi-decade experience of deflation argues in favour of the BOJ defending it’s yield cap. As such, with a lag domestic investors could start to allocate more funds to JGBs and there is the potential for the JPY IRS 7fwd 3yr (the pivot point of the curve and our favoured way of expressing bullish views on JPY interest rates) to rally 25-30bp over the coming 3-6 months. However, we cannot recommend such a trade with any degree of conviction. Our bias is to seek trades with a positively convex return profile, but being received JPY interest rates essentially makes investors “short” the potential for a possible BOJ regime shift in policy formation. While we feel this is unlikely, Japan’s long history of policy mis-steps and the experience of the SNB and its decision to break its FX floor, both provide reason for caution. We have greater comfort in expressing a short-JPY FX view, particularly against the USD. While USD/JPY is clearly correlated to the JGB trade, there are some independent dynamics that could argue for the recent correction lower in USD/JPY running its course and reversing.

The BOJ’s yeild cap is confirming the pro-cyclicality of its monetary framework

The BOJ’s policy framework requires faster liquidity creation when it is less needed

One of the more interesting if not unusual aspects of the Bank of Japan’s (BOJ’s) switch to yield curve targeting as its monetary anchor is that it embeds a clear pro-cyclicality into the conduct of monetary policy. (As for generals, so for central bankers: USD/JPY and the BOJ’s pro-cyclcial policy framework.) In a climate of diminishing growth and inflation expectations, demand for JGB duration would be expected to increase and flatten the curve. If it flattened to the point where the gains accruing to JGB investors from sliding down the curve no longer compensated for the negative carry of holding bonds, the BOJ would be required to either cut policy rates further (which it is loath to do) or more likely taper it’s JGB purchases to try and re-steepen the curve. In contrast, during periods of improved growth expectations the BOJ may be required to accelerate JGB purchases in order to defend its yield cap on longer dated securities. (Chart 1.) Unsurprisingly, we are critical of the BOJ’s new yield curve methodology: it constrains the BOJ from implementing an optimal and efficient monetary policy; monetary policy is meant to be counter- rather than pro-cyclical, which in turn injects an inherent event-risk into markets given the potential for the BOJ to alter it’s current framework. (The BOJ’s new policy framework and jar Jar Binks.)

Chart 1. The steepening of the JGB curve is threatening the BOJ’s yield curve targeting framework


Source: Bloomberg

The BOJ will need to keep expanding its rinbans to renew confidence in it’s yield cap

The BOJ is currently demonstrating the pro-cyclicality of its policy stance. As global growth expectations have increased JGB yields have moved higher. 10yr yields rose above the BOJ’s 0% soft ceiling and spiked to 13.5bp on 3rd February. The response has been an aggressive acceleration of BOJ Rinban operations. The BOJ announced a surprise fixed rate rinban in the afternoon of 3rd February when it offered to purchase “unlimited” 10yr JGBs at 11bp. The BOJ’s rinban operations for February are expected to exceed the JPY8-12trn a month guidance from the BOJ as it defends a new cap on 10-yr yields of around 10-11bp, or in the BOJ’s words “close to 0bp”. The counter-cyclicality of BOJ monetary policy suggests that the retracement in USD/JPY since President Trump’s inauguration might be over and that the JPY could renew a weakening trend. Similarly, the steepening of the JGB curve may entice an asset allocation into bonds from domestic investors if there is sufficient confidence that the ~10bp ceiling in 10yr JGBs will be held. (Although given that the prior 0% soft threshold was allowed to break and that the Bank refuses to provide a clear line in the sand for its yield cap, it may take a considerable number of expanded rinban operations to provide that confidence to investors. (Chart 2.))

Chart 2. The yield curve targeting framework will require an acceleration of JGB purchases during periods when growth expectations are improving


Source: CEIC

Economically, there is little reason to fear the BOJ allowing it’s yeild cap to break

From an economic standpoint, there is little reason to expect the BOJ to blink and allow its yield curve-targeting threshold to be broken via a substantial rise in yields:

  • Price stability remains elusive. The Japanese economy is far from escaping its medium-term path of deflation. The Y/Y rise in the CPI was just 0.3% in December while core inflation was 0.1% and trending lower. (Chart 3.) The BOJ’s target of achieving 2% inflation “around” FY 2018 appears increasingly unrealistic, with the lingering adverse impact on economic growth of the April 2014 consumption tax hike acting to complicate further the path to price stability.

Chart 3. Japan has failed to reverse deflation during the Abenomics era


Source: CEIC

  • Wage growth is laggardly. The outlook for price stability continues to be dulled by weak wage growth. (Chart 4.) Average monthly earnings rose just 0.1% Y/Y in December and have consistently surprised on the downside and belied the ostensive signs that the labour market is tight. (The unemployment rate is just 3.1% and the jobs-to-applicants ratio of 1.43 is the lowest level since July 1991.) As we have highlighted before, there is a structural downdraft on Japanese wages that means that there is a reduced sensitivity to these measures of labour market tightness and indeed to demographic factors that will shrink the workforce. Large companies in Japan became more aggressive profit maximisers under the PM Koizumi reforms of the early 2000s and as such have been aggressively driving down costs in their network of SME suppliers. (Chart 5.) This contrasts the previous system whereby there was a symbiotic relationship between large firms and SMEs and each sector’s profit cycle was highly correlated. Now, SME profitability is lagging as large firms force them to reduce costs and this sector accounts for 65% of all employment and hence is the driving force behind aggregate wage gains. (Japan – a stir of echos: anticipating QQE3 and beyond.)

Chart 4. Japanese wage growth continues to disappoint…


Source: CEIC

Chart 5. …and a key reason is the breaking of the formerly symbiotic relationship between large firms and SMEs in the early 2000s


Source: CEIC

  • Growth remains weak. Japan’s economic performance over the period of Abenomics has been poor, in part due to the policy error of imposing a consumption tax hike in April 2014, which undermined a building economic momentum. (In response, GDP declined by 7.1% saar in Q2 2014). Japan’s real GDP grew 1.0% Y/Y in Q4 and since the eve of the consumption tax (Q1 2014) has merely grown a cumulative 1.1%. In the 15 quarters of Abenomics, Japan’s real GDP has grown by just 5.1%, despite an economy that at the time had still not recovered the sizable degree of output lost to the GFC. (Chart 6.)

Chart 6. The April 2014 consumption tax hike undermined the momentum of economic recovery


Source: CEIC

  • The practical limits of monetary easing have not been reached. For all of the talk of a shortage of JGBs, there remains a considerable pool of assets the BOJ can purchase before supply-constraints become binding. If the Japan is to break its deflationary psychology then banks, asset mangers and asset owners need to adopt less deflationary portfolios so heavily skewed to risk-free assets. As Chart 7 shows, the BOJ’s accelerated asset piurchases via QQE has seen bank JGB holdings decline from a pre-Abenomics level of 18.7% of their total asset base to 8.0% at present. This process has removed JPY82trn of JGBs from bank balance sheets and forced them to seek fresh assets that are more risk-bearing. Returning the JGB holdings of Japanese banks to around 4% of their total assets, which is a level more normal internationally and which also reflects levels seen after the bubble economy burst in the late-1980s/early-1990s, would entail further JGB sales of JPY42trn. While these numbers seem large, Japanese banks hold far fewer JGBs than Lifers and Pension funds. We believe that the BOJ can maintain an JPY80trn a year pace of JGB purchases for at least another two years before supply constraints become more binding and it has to consider switching to other assets. (Although of course as the BOJ encounters ever more reluctant sellers of bonds as Japanese risk-free asset holdings approach international norms, there will be an increasing price convexity of JGBs with respect to QQE.)


Chart 7. The BOJ’s QQE policy is forcing Japanese bank holdings of JGBs to converge to more international norms of around 4% of total assets


Source: CEIC

There is a complimentaity between the needs to the economy and the BOJ defending it’s yield cap

For all of its many flaws, the current monetary framework for the BOJ therefore has the capacity to inject considerable additional stimulus into the economy as the pace of QQE accelerates. This can lay the foundation for a renewed weakening of the JPY. There is a complementarity at present between the need of the economy for more stimulus and defending a 10bp yield cap.

However, there are lingering concerns that things might not prove to be so simple…


Reason for caution #1 – Japan’s history of policy mis-steps

Post-bubble economy, Japanese policy settings have rarely been sufficiently coordinated

Some central banks have a tradition of hawkishness (Buba). Others have a tradition of pragmatism (the Fed). Others have an unfortunately reputation for policy mis-steps and implementing inappropriate monetary policy decisions. The BOJ fits into the latter camp, and more recently the problem has been compounded by macro-policy errors with respect to fiscal policy. In short, Japan has rarely had a true coordination of policy across reflationary levers since the bursting of the bubble economy.

  • Post-bubble to Abenomics. During the 1990s and 2000s the Japanese economy was supported by an expansion of the government spending and rise in public sector indebtedness. While the BOJ tried to match the government’s fiscal largesse with monetary easing, its efforts always fell short of a sufficiently aggressive and pro-active monetary reflation. This moderated the multiplier effect of the expanded fiscal spending. The problem was that the BOJ was guided by a flawed measure of the output gap, one that understated potential economic growth and thus the level of slack in the economy. (For the geeks, the BOJ uses a Cobb-Douglas production function, which renders productivity growth as a residual. This creates a cyclicality whereby when Japanese growth slows, so does the measure of potential economic growth, which means that the BOJ tends to under-estimate the slack in the economy). This model has seen the BOJ ease too-little too-late and traditionally be prone to tightening too soon given that it frequently feared a rise of inflation that was purely in the mind, or more appropriately algorithm, of it’s flawed economic model. Over the past two decades, much gnashing of teeth and beating of desk with forehead from this analyst has accompanied BOJ policy decisions referencing a diminished “deflation gap” and “mounting inflation risks” based on reference to this flawed but nonetheless guiding econometric model.
  • Abenomics Part 1. Under Governor Kuroda, the BOJ appeared to have embarked on a step change. The flawed model of the output gap was set aside and monetary policy was guided by the belated need for monetary and fiscal reflation aimed at definitely ending Japan’s two decade history of deflation. The BOJ launched QQE and its balance sheet ballooned from 30.4% of GDP on the end of Abenomics (end-2014) to 85.1% In Q3 2015. While the BOJ was at last showing sufficient ambition, this time it was the government, which proved hesitant. The second arrow of Abenomics (fiscal easing) remained in the quiver, with the flawed decision to raise the consumption tax in April 2014 undermining the building economic recovery. (Chart 5.)
  • Abenomics Part 2. In 2015 and 2016, the BOJ was maintaining its aggressive monetary easing and targeting a JPY80trn annual expansion of the monetary base, but there was an opportunity to accelerate asset purchases further. The economy was struggling to recover the pace of growth seen prior to the consumption tax hike and inflation remained weak. Instead of accelerating monetary easing, last September the BOJ moved to its yield curve-targeting regime which – as noted above – creates a pro-cyclical policy framework which is simply not the correct way of continuing a multi-decade fight against deflation.


History does not provide a great deal of confidence that the BOJ will stick to its current monetary framework.

Reason for caution #2 – The experience of the SNB

The BOJ’s current situation also has echoes of the SNB’s policy dilemma. From 6th September 2011 to 15th February 2015 the SNB established a 1.20 floor to EUR/CHF to prevent further over-valuation of the Swiss Franc. (Chart 8.)

Chart 8. The SNB removed it’s EUR/CHF floor in February 2015 despite having greater ability than the BOJ to maintain it’s balance sheet expansion


Source: Bloomberg

The SNB’s monetary creation was not in opposition to the Swiss business cycle

Like the BOJ’s current need to defend it’s yield curve ceiling, there was a complementarity between the SNB’s policy framework and the immediate needs of the business cycle. Just as the desire to end deflation is consistent with the BOJ accelerating it’s rinban purchases, so did the lack of Swiss inflation support aggressive monetary expansion by the SNB.(Chart 9.) GDP growth was also subdued, and averaged 1.6% Y/Y during the period of the FX floor.

Chart 9. Swiss inflationary pressures were negligible during the period of the EUR/CHF floor…


Source: CEIC


The SNB has far greater practical ability to maintain it’s FX floor than the BOJ has in defending it’s yeild cap

From a practical stand-point, the SNB encountered far fewer technical impediments to maintaining a pronounced monetary easing than the BOJ currently encounters. While the BOJ is primarily purchasing JGBs which have a finite supply, SNB was selling CHFs which it was itself creating. This enabled SNB’s FX reserves to surge from CHF337bn since the EUR/CHF floor was introduced to the present level of CHF646bn, and FX reserves now exceed the size of the Swiss GDP. (Chart 10.) Despite rising domestic concerns about a medium-term inflation risk stemming from such a pronounced monetary easing, both growth and inflation were stubbornly non-resistant to the expansion in the money supply. And yet, despite fewer practical restraints than the BOJ faces over the conduct of it’s monetary easing, the SNB chose to end it’s EUR/CHF floor on 15th February 2015. This adds an additional note of caution into the BOJ’s defence of it’s yield ceiling.

Chart 10. …and proved resistant to the SNB’s dramatic balance sheet expansion


Source: CEIC


Can we trust the BOJ’s fortitude?

Economically, there are reasons to back the BOJ’s yeild cap

There are strong economic reasons to expect that the BOJ will maintain its yield curve targeting regime and attempt to make progress towards the goal of price stability and a self-sustaining phase of economic growth. While Japan’s economy is experiencing an upswing in tandem with a global pick-up in growth, it remains weak and far from the level that will help secure price stability. (Chart 11).

Chart 11. Japan’s business cycle is picking up, albeit unimpressively


Source: CEIC


Governor Kuroda’s final period in office could add to the BOJ’s commitment to floor the accelerator

There is a human level, which also argues for the BOJ maintaining its current monetary framework. BOJ Governor Kuroda’s term as BOJ governor is due to end in March 2018. At the current pace of monetary expansion, his 5 year term of office will have been associated with the BOJ’s balance sheet swelling to over 100% of GDP, a rise of 55pp. Despite this explosion of liquidity growth, Japan’s CPI risks ending his terms of office as it began – in deflation. We would anticipate Governor Kuroda attempting to hold his foot down on the accelerator with respect to monetary easing over the remainder of his tenure. Throw into the mix the near-term positive of an expected period of strength in USTs which reached our targeted buy-level a fortnight ago, and there is reason to expect the BOJ’s yield cap to hold and for this to gradually entice an asset allocation into bonds from domestic investors. (Trumponomics and the potential for a USD overshoot.)

JGBs and JPY IRS expected to rally over the comign 3-6 months: look for a 25-30bp declien in JPY IRS 7fwd 3yr

On a tactical basis, there is reason to be bullish JPY interest rate markets. The JPY IRS 7fwd 3yr (the pivot point of the entire curve and our favoured way of expressing bullish views on the JPY interest rate market) currently yields 54bp and has an annual positive slide of 8.6bp. (Chart 12.) While this interest rate will struggle to reach the record yield lows of 5bp seen in July 2015 (which reflects how Trumponomics has subsequently pushed the pivot point for the 10-year part of the US yield curve 100bp higher (Trumponomics and the potential for a USD overshoot) and how the BOJ’s new policy framework would see such a rally presage fears of a premature policy taper) a move to 25bp is likely. This is particularly the case given that the scale of the BOJ’s QQE is absorbing an ever larger pool of the “float” of JGBs and hence has the potential to have an increasingly convext impact on prices as the sellers of bonds to the BOJ grow increasingly reluctant to do so.

Chart 12. The risks of the BOJ blinking with respect to its yield cap make us cautious about recommending received JPY IRS positions…


Source: Bloomberg

BUT. Being long JGBs embeds negative convexity into investment portfolios

However, despite this view, we struggle to recommend long-JGB/ received JPY IRS positions with any degree of conviction. This reflects how our instinctive bias is to seek trades with a positively convex return profile. Receiving JPY interets rates is a position with a potential negative convexity. While we do not expect the BOJ to break its yield curve targeting regime, a long history of Japanese policy mis-steps and the SNB’s experience of ending its FX peg when it had far greater technical capacity to continue than the BOJ has, naturally urges caution. We do not expect the BOJ to break it’s policy framework over the rest of this year but were they to do so and allow yields to rise, the price action could be disruptive and illiquid as domestic investors moved to hedge/ pare-back their duration exposure: the JGB market has several experiences of sudden and sharp ~100bp sell-offs. Set beside this risk, 8.6bp of annual slide and a potential 25-30bp rally is not overly compelling.

We therefore struggle to find a concompelling reason to be strategically long JGBs/ received IRS

Of course, with policy rates at -10bp there is a limit to how far the curve could eventually sell-off and steepen as the carry and roll would become too enticing for domestic investors, but that does not preclude the potential for a severe initial over-shooting of yields. For choice we would therefore be long-JGBs/ received JPY IRS at the current level, but would view such positions as tactical and with relatively muted DV01 risk, given that a received position is potentially betting against an adverse policy shock. Similarly, a lesson of the SNB’s policy shift is that short-gamma positions can be particularly unappealing during periods when investors are effectively betting against the non-negligible risk of a regime shift in monetary policy.

Long-USD – despite the corrlation to JGBs – provides a more compelling investment proposition

We find that the FX markets provide more appealing risk reward trades at present. Of course, being short-JPY and long JGBs are highly correlated trades. Were the BOJ to tolerate a rise in yields, the JPY would be expected to appreciate sharply. However, there are reasons to be more comfortable in being long-USD/JPY than long JGBs at present:

  • While it will take domestic investors some time to gain confidence that the BOJ is committed to capping JGB yields, the need for expanded rinban operations will – via an accelerated pace of liquidity creation – provide more immediate fuel for a weaker JPY. A higher USD/JPY would be expected to lead rather than be coincident to a possible rally in JGBs.
  • The current pick-up in the Japanese business cycle will lend support to USD demand: the JPY and USD/JPY in particular has been highly cyclical during recent decades whereby a stronger domestic economy and greater confidence among investors encourages an increased allocation to overseas assets. (As the old adage goes, “if you like Japan, sell its currency”).
  • We continue to expect a gradually unfolding of the key elements of President Trump’s eocnomic policy platform over the coming months, most notably the various versions of the Border Tax (Versions 1.0 to 3.0) and fiscal reform, will lay the foundations for renewed USD strength. (Trumponomics and the potential for a USD overshoot.)

Chart 13. …and long USD/JPY positions may prove a safer play


Source: Bloomberg

USD/JPY to return to 120

We therefore view weak-JPY positions as the better way to position for the pro-cyclcality of the BOJ’s policy framework. We believe the correction lower in USD/JPY since Trump’s election victory became manifest, is passing and recommend scaling into long positions. We expect a return to 120 wthin 3-6 months.