- Bund yields are now at levels implying that the eurozone has gone ‘Japanese’. We think this conclusion is premature.
- In Japan, wage deflation became entrenched eight years after asset prices peaked. In Europe that peak in asset prices occurred in 2007, prior to the global financial crisis, eight years brings us to 2015.
- In Japan, poor fiscal policy and a slow monetary policy response worsened matters. So far, the monetary policy response in the eurozone has been faster and more progress has been made on structural reform as well.
- Europe’s demographics are as poor as those of Japan, but immigration policy is far more open. There is no deflationary spiral driven by falling wages as was the case for years in Japan.
- Nevertheless, the eurozone is in ‘ageing mode’, i.e. in a low growth environment; it should be able to avoid prolonged deflation, but official inflation targets are likely to be missed for years to come.
- Even in such an environment, current yields appear too low and reflect other factors, not just concerns over ‘Japanisation’.
We define ‘Japanisation’ as a move to an economic environment characterised by a deflationary spiral fuelled by permanent wage cuts. This leads to a build-up of deflation expectations which structurally undermines demand and changes consumption and corporate spending behaviour en route.
Japanisation of the eurozone? Looking at current German Bund yields, one could ask if this is still just an academic question or a market verdict? A deeper look into the issue is essential if investors are to understand whether Japanisation is the main driver of market dynamics or just a spurious rationalisation of something else occurring in the European bond market.
There are many similarities between the situation in Europe today and Japan’s experience over the past 25 years. However, we believe there is still some way to go before one can conclude that Europe will undergo Japanisation. It took at least a decade to realise what was happening in Japan and we will only know whether this occurs in Europe in years to come.
To be able to compare the key economic and market dynamics in the eurozone and Japan, we will look at what happened in Japan after the peak of equity prices in December 1989. For Europe, we consider the peak of property prices in 2007 as the starting point. This coincided with the start of the global financial crisis (in 2008 & 2009) and worsened with the escalation of the euro crisis in 2010.
Source: Bloomberg, BNP Paribas IP
Comparing Japan and Europe
Asset prices – very different. Japanese assets (property & equities) were unsustainably expensive by December 1989. The bursting of that bubble was relatively quickly followed by the property bubble deflating as well. In 2007, eurozone asset prices were not as over-inflated as those in Japan and did not drop as steeply.
In Japan, valuations on the equity and property markets were far higher than that seen in Europe and both the property and equity markets popped within a couple of years of each other. In Europe, equity valuations peaked in 2001 (in the dot.com bubble), but did not return to those levels. The region is also more heterogeneous: whilst the property market peaked in 2007, property bubbles were seen in a few countries (e.g. Spain). Even now, Germany still has a relatively cheap property market.
Exhibit 2: Equity market CAPE (Cyclically adjusted price-to-earning ratio)
Source: MSCI, Datastream, BNP Paribas IP
Debt overhang – of a similar magnitude. In the lead-up to the peak of the bubble, the Japanese economy was growing quickly, fuelled by rising leverage. By 1990, non-financial debt as a share of GDP exceeded 350% (more than 150 percentage points above US, UK or eurozone levels). By 2010, eurozone non-financial debt as a share of GDP was close to 300% and was thus also very high.
Banking strains & recapitalisation – Europe in better shape. The bursting of the Japanese asset bubbles (equities and property) put huge pressure on the overstretched banking system. It was not recapitalised until 1998, causing the large debt overhang to lead to a ‘zombification’ of the banking system for many years. Bank lending declined and loan growth only started rising again in 2006.
The global financial crisis in 2008 led to large strains on the global banking system (including Europe’s), but the clean bill of health resulting from the ECB’s comprehensive asset quality review in late 2014 has left banks in a position to start lending once more.
Exhibit 3: Japanese and eurozone bank lending growth Y.o.Y. change (%)
Source: Bank of Japan, ECB, BNP Paribas IP
Having the right fiscal and monetary policy mix is vital to avoiding a long-term crisis:
Fiscal response in the face of crisis – similar. In the 1980s, the Japanese political system was dominated by one party, the LDP, which had the power to take tough fiscal decisions and put the brakes on growth. By late 1989, the LDP started losing its grip on power, but fiscal policy remained restrictive until 1991. Amid growing pressure and a more fractious political system, fiscal policy was then loosened quickly. However, in 1997, ‘Fiscal Restructuring Targets’ were enacted, including an increase in VAT. This was the straw that broke the camel’s back – the fragile recovery stopped and almost a decade of deflation ensued.
In 2009, the fiscal policy response in the eurozone initially relied more on automatic stabilisers than on large stimulus packages. During the global financial crisis, spiking fiscal deficits had been a timely and appropriate response. Similar to the fiscal reform that Japan tried to enact in 1997, eurozone governments shifted to a fiscal austerity agenda in 2010, in the face of the euro sovereign crisis. However, in 1998, Japan had quickly reversed its fiscal restraint, replacing it with a large stimulus package to address the economic collapse.
In 2014, in the face of rising political pressure across eurozone and a need to stimulate demand, some eurozone governments were given leeway to meet their EU budget targets in a move that bears some similarity to Japan’s suspension of the Fiscal Structural Reform Act in 1998. However, the relaxing of targets did not mark a large U-turn and eurozone governments have so far not launched large-scale stimulus (unlike Japan). Instead, they are responding flexibly to current market conditions.
Monetary policy responses – ECB quicker to react. The Bank of Japan (BoJ) tightened monetary policy in the face of a growing asset bubble, raising the official discount rate by 175bp to 6% by late 1990. Rates were cut from 1991 onwards, but the Japanese yen continued to strengthen until 1995, amplifying the tight monetary conditions. It took until 1999 for Japan to have zero interest rates (ZIRP). Quantitative Easing (QE) was only introduced in March 2001, two years after deflation started.
The ECB’s response to the great financial crisis and the eurozone crisis was initially relatively slow compared with that of other central banks dealing with crises at the time. The ECB even briefly raised policy rates in July 2011 (in the middle of the crisis) due to inflation concerns. In moving, this January, to unorthodox monetary policy measures including full-blown QE the ECB has launched a more expansionary policy response than that of the BoJ at that time.
Currency strength – better for Europe. In the past seven years, the euro has moderately weakened, while the yen strengthened significantly until 1995 (see below). This eroded Japanese competitiveness and put a strain on wages to help restore some competitiveness. As for the euro, recent monetary policy has led to a significant depreciation, helping to restore competitiveness. At the margin, this should help alleviate the need for companies to cut costs (via wage cuts) in years to come.
Exhibit 4: Real effective exchange rates yen/euro
Source: Bank for International Settlements, Bloomberg, BNP Paribas IP
Slack in the economy – similar. The output gap in the eurozone remains large and continues to exert dis-inflationary pressures. This output gap is larger today than it was for Japan in 1998. While currently the eurozone is returning to a stronger recovery track and the output gap is starting to close, there is a long way to go.
Exhibit 5: Economic output gaps in Japan and euro area
Source: OECD, BNP Paribas IP
Wage growth – on a similar path. In Japan, wages and the currency rose in the first six years after the peak of the bubble, eroding competitiveness swiftly and significantly. Japanese companies were forced to make sharp adjustments to regain competitiveness. However, hiring and firing rigidities, which still affect Japan today, left them with only one way to do so: through wage cuts. This turned into a vicious cycle of further cuts and lay the foundation for a permanent drop in prices, i.e. deflation. This is why wage growth takes pride of place when it comes to indicators of Japanisation.
Exhibit 6: Graph showing nominal labour compensation growth per employee (Japan and eurozone)
Source: OECD, BNP Paribas IP
The large output gap in the eurozone should continue to impact wage growth trends. Although it varies widely across the currency bloc, wage growth in the eurozone has stabilised at around 1%. At the onset of the financial crisis, wages fell in some countries in one-off adjustments to restore competiveness. If wages were to decline across the whole eurozone for a prolonged period, this would be a precondition for permanently declining prices.
Demographics – similar declines of working age population expected. The Japanese working age population started declining only in 1995. After peaking in 2010, the eurozone working age population is forecast to decline, but not by as much as Japan’s (see below). This will put a dampener on potential growth for both economies in the decades to come.
Exhibit 7: Working age population in Japan and euro area
Source: OECD, BNP Paribas IP
Immigration policies – different. Japanese anti-immigration policies have worsened the ageing and labour market immobility problem. In 2010, the foreign population of Japan accounted for just 1.7% of the total. Eurozone immigration law is more liberal and according to OECD data, the foreign-born population for those countries that report immigration statistics in the eurozone is closer to 6.5% (2011). That is considerably higher than in Japan.
Labour market regulation – high in both. Both product and labour market regulation were high in Japan in the 1990s. These barriers to entry and rigidities limited the capacity for creative destruction and hurt Japanese companies. In the eurozone, labour market hiring and firing flexibility varies across the region, but the largest economies are showing greater flexibility (France) and are improving (World Economic Forum). Whilst in Japan, this competitiveness has continued to decline over the past decade as well.
Reform willingness & progress – better in Europe. After more than 20 years, structural reform is progressing slowly in Japan despite the high approval ratings and the LDP’s political majority. The eurozone appears to be on a faster track. Some countries are already harvesting the first fruits of reforms (Ireland and Spain), which should help to lift trend growth in the region over the longer term.
What Japanisation means for asset classes
Over the past 10 years, Japanese government bond (JGB) 10-year yields have averaged 1.2%, having been at 6% in December 1989.
- Forward Japan inflation expectations as discounted in market pricing (5yr 5yr inflation) declined sharply in the 1990s and then more slowly in the early 2000s, but still remained above zero.
- Eurozone inflation expectations have also fallen over the past eight years. The fall has been less dramatic as inflation expectations started at a lower point.
Exhibit 8: Graph showing Japanese inflation expectations and CPI inflation
Exhibit 9: Eurozone inflation expectations and CPI inflation
Sources: Eurostat, Ministry of Internal Affairs and Communications, Bloomberg, BNP Paribas IP
- Property prices fell for over a decade in Japan as the deflationary spiral seeped into asset prices (see below).
- Eurozone property prices have fallen less.
Exhibit 10: House prices in Japan and euro area
Source: Japan Real Estate Institute, Eurostat, BNP Paribas IP
- Japanese equities underperformed world equities for two decades (see below).
- European equities have underperformed since 2007, but not by as much
Exhibit 11: Equity performance MSCI Japan and MSCI Europe vs World
Source: MSCI, Datastream, BNP Paribas IP
- Increased uncertainty is a side effect of Japanisation and is also experienced in times of higher market volatility. From 1969 to 1990, the realised volatility of Japanese equities was in line with that of US equities (17% vs. 16%). However, since 1990, Japanese realised volatility has been noticeably higher (20%) than US equity volatility (15%).
- Since 2007, Euro equity volatility has averaged 26%. Over the same period, US equity volatility has been lower at 22%. From 1999 to 2007, US realised volatility lagged that of the eurozone at 20% versus 24%.
Bund yields at these levels are different to JGB yields
Other factors have been pushing down yields in Europe. These could prove to be fleeting, allowing yields to rise in the future.
The ECB’s monetary policy response has been more pro-active and more drastic than the BoJ’s was at this point of the cycle (seven years after the asset bubble peak):
- A negative deposit rate was introduced in June 2014 and in September, it was cut further to -0.20%. Japan never moved to negative key rates.
- The ECB shifted to a more accommodative stance in 2011 in the face of the eurozone sovereign debt crisis (three years after asset prices peaked). The BoJ changed its stance much later.
- The move to full-scale QE in January 2015 underscores the ECB’s fight against the build-up of deflation expectations, which deteriorated in 2014.
- Currently, German yields are negative up to six-year maturities. Apart from investors accepting that they have to pay a premium for ‘riskfree’ assets, expectations of QE did contribute to yields falling to these levels. ECB asset buying should alter the demand/supply balance quite dramatically by the autumn of 2016. This may encourage some investors to position portfolios in anticipation of lower interest rates. Some analysts are starting to suggest that negative yields for 10-year maturities could be seen in the near term.
Low inflation expectations are not exclusively a eurozone phenomenon. Disinflationary forces are pushing bond yields down across the world. The fall in oil prices has helped to convert headline inflation into deflation. However, we see this as temporary and not necessarily as a precursor to a prolonged period of deflation. Despite this, longer-term inflation expectations have fallen in the eurozone as they did in Japan in the late 1990s.
Today, for a large pool of money, investment decisions are not driven solely by forward-looking fundamental factors, but factors and risk aversion considerations such as the forex reserves of central banks, accepting a premium for ‘safe haven’ assets and ALM-driven flows as well. These investment flows continue to supress yields.
Bunds have moved much quicker to Japanese-style yield levels and this appears overdone in light of the associated verdict of markets that Europe could be going the way of Japan. For us, this is a premature conclusion.
Despite many similarities, it is too early to assume Japanisation in the eurozone, but ‘ageing’ will occur. A low growth environment is likely for years to come, but this is not the same as Japanisation (a deflationary spiral). Due to demographics alone, eurozone growth is trending down. We expect it to be 1%-1.25% in the longer term. This makes business cycles more vulnerable to shocks and longer periods of stagnation. Growth falling close to or below zero could become more frequent, while periods of very strong cyclical growth look less likely. Large output gaps also increase the likelihood that inflationary pressures will stay subdued, but in the absence of falling wages, long-run deflation risks are more limited.
The bigger risk is that the ECB’s inflation target will not be reached for a prolonged period. However, this is not just a challenge for the ECB, but many other central banks as well (e.g. average inflation for the last 10 years in Switzerland is 0.5%).
Against this backdrop, we believe the recent historical low in Bunds is prematurely discounting market expectations that Europe (and Germany) will go the way of Japan, reflecting weak economic growth and other global factors.
A Japanese scenario should not be misinterpreted as the death of bond market volatility: 10-year Japanese yields were at 0.45% in December 2003 and back at 2% in 2006. In Japan, average real GDP growth for 2001-2013 was 0.8% p.a. and average deflation was -0.2%. We expect the current soft patch of growth for the eurozone to pass by the spring of 2015. A tailwind for ever lower Bund yields in recent years has been the highly accommodative policy of G3 central banks. This will gradually unwind in coming years – we expect that the US Federal Reserve will start hiking rates later this year.
We use nominal growth forecasts as a simple reference point for Bunds’ fair value: taking the latest consensus long-term forecasts for growth and inflation, and expecting that the ECB successfully manages inflation back to its 2% target by 2019. In this scenario, a substantial setback for Bunds is likely, with yields back towards 2% in the medium term and above 3% by 2024. The path of adjustment may be smoothed somewhat by the announcement of QE. Even taking this into account and using our own expectations for lower trend growth, and assuming inflation rates stay lower for longer (below the ECB’s target), current yields are still too low and are only justified if prolonged deflation ensues.
Using a CAPE valuation framework, developed market equities do not look cheap today. However, within developed market equities, European equities can offer value. European equities currently have a CAPE of 15, below their historical average of 18 and below that of most developed markets. Over the medium term, we favour European equities over US equities and we expect a 5.5% local equity return for European equities over the next 5-7 years.
In the near term we expect moderate economic growth in the eurozone (and the world) to result in corporate revenue growth accompanied by limited cost pressures in 2015. While pricing pressures should remain problematic, the pick-up in volume growth should offset this in the near term. Thus margins can expand from their cyclical lows today. This earnings growth should be a key catalyst for equity returns in 2015. We expect multiple expansion to be a smaller contributing factor as ECB accommodative action is likely to be offset somewhat by the Fed’s modestly tighter policy.
Exhibit 12: Graph showing Japanese, global and European equity market
Source: MSCI, Datastream, BNP Paribas IP
Euro government bonds – we retain a slightly lower duration exposure than an average bond benchmark, but still retain the capacity to reduce it further in the future. Whilst Eurozone government bond yields are extremely low, we expect them to stay lower for longer due to the ECBs quantitative easing measures and because of the disinflationary pressures that Europe is likely to continue facing for years to come.
European Equities & Real Estate – we remain overweight European equities compared to a traditional developed equity benchmark (MSCI World). We expect the growth environment in Europe to improve sufficiently to unlock the relative value in this market (compared to other developed equity markets).
We are currently raising our exposure to equities at the expense of real estate. We think the market pessimism on eurozone growth and deflation is overdone. We expect equity returns to be higher than real estate returns (which benefitted a lot more over the past year from the fall in long dated bond yields).
 GDP-weighted average of Germany, France, Italy, Spain and the Netherlands
 The countries that report immigration statistics are Austria, Belgium, France, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal and Spain.
 5-year 5-year forward inflation is the inflation implied in financial markets. Thus the inflation rate for 5 years starting in 5 years’ time (ie a measure of expectations of inflation in the distant future).
 This is our propriety model 5-7 year annualised gross return expectation for MSCI Europe in Euros