NN IP: Die globale Wirtschaftspolitik im Umbruch

Forget about the economy for now and look at politics and policy

As we have often argued, the real economy, the financial system and the political arena are very much interconnected, but the strength of the feedback loops between them is state dependent. Over the past year, we have seen the waxing and waning of various forms of political and policy risks but on the whole, markets have behaved in accordance with the reflation trade. Equity markets and safe Treasury yields have risen while credit spreads have declined. Meanwhile, despite the recent moderate appreciation, the dollar has depreciated on balance over the past year following an initial appreciation which took place immediately after the election of Trump. The dollar depreciation makes sense from a number of perspectives. First of all, the (positive) marginal change in (expected) growth momentum has generally been larger outside the US (Euroland, emerging markets) which makes sense because output gaps in these regions are still larger. What’s more, the market has been focusing on downside risks to the Fed’s intended gradual hiking path (which is very imperfectly communicated by the dot plot) while there have been bouts of (exaggerated) excitement about the ECB’s first tentative steps towards the exit. In short, the ECB policy outlook is much more susceptible to tantrum-like events because the Fed is already on a conventional hiking path. Finally, the dollar tends to depreciate in a risk-on environment because US safe and liquid assets perform the function of the world’s benchmark risk free asset. 

A risk-on environment will thus reduce the demand for global liquidity. The kind of political and policy risks seen over the past year (Trump protectionism, uncertainty about US fiscal stance, geo-political tensions, European elections, uncertainty about the composition of the Fed Board of Governors etc.) might well have been sufficient to trigger a risk-off environment in a different economic setting. Still, because of the resilient and broad-based economic growth momentum seen over the past year, the threshold beyond which political/policy uncertainty starts to exert a persistent negative effect on risk appetite has clearly been raised. In short, the economy seems to be taking care of itself as the momentum is pretty much self-sustaining, so investors had better spent relatively more time pondering the political and policy risks. 

Japan and Spain in focus

Having said that, a large enough negative political or policy shock will of course still cause market sentiment to sour substantially. In fact, we have seen and still see this on the “local level” even though not always. Last week we discussed Japanese political risks, but the probability that the Abenomics experiment will be cut short is not yet reflected in Japanese equities or the yen, perhaps courtesy of economic data which suggest the above-potential growth momentum is continuing. Incidentally, the new Party of Hope became much clearer this week about its intention to end Abenomics if it should rise to power. In particular, it wants the BoJ to head towards the exit. In our view this is an assured recipe for the “Mother of all Tantrums” which, if it were to happen, would probably mostly be expressed by a yen which goes through the roof. Meanwhile, Spanish stocks and sovereign spreads displayed some volatility on the back of the outcome of the Catalonian referendum. Until a few weeks ago, it very much seemed to be a local affair but now it has clearly caught the attention of the rest of the world and in particular of the EU. The situation remains pretty fluid, but it is good to bear in mind that in the end also this issue is at least partly driven by a conflict about the distribution of tax income (even though history is also a main driver). Catalonia wants to significantly reduce its large net positive tax distribution to the Spanish national government. Any solution will in the end probably imply a larger degree of autonomy especially on the financial level, which means that the central government will face a regional distribution problem and may be able to redistribute less to poorer regions. This is a recipe for political tensions which are in fact already rising for the central government. In particular, it now seems difficult to get a majority for the 2018 budget which needs to be presented to the EU by mid-October. The government could always extend the 2017 budget but this will likely not meet the deficit targets agreed with the EU. All this is probably the main driver behind the volatility in Spanish spreads. On top of this, there is a tail risk (albeit small) that Catalonia will head for an unnegotiated and unilateral secession, in which case it will not honour its share of Spanish national debt. This would cause the Spanish debt-to-GDP ratio to rise from around 100% towards Italian and Portuguese levels closer to 130%. 

A change in the global policy mix

Meanwhile, at the global level the policy mix is in the process of changing towards (somewhat) tighter monetary policy and a dose of fiscal easing. If you put this statement into a standard “plug and play” economic model the outcome is a rise in real interest rates. However, it should be kept in mind that the real world is a bit more complicated than this model. Real yields have been on a declining trend since the early 1980’s and one of the main issues in financial markets is whether and to what extent this will turn into an upward trend. The early part of this story is very much driven by monetary policy, which had to remain pretty tight relative to the economic cycle to tame high and variable inflation expectations. By the mid-1990s the latter had become pretty much anchored to the target and real yields more or less started to move sideways. However, in the early 2000s a renewed downward trend set in. With the benefit of hindsight this may have been partly triggered by the decline in DM potential growth which set in at the time. The theoretical rationale is that this reduces the macro return and thus the incentive to investment. Still, in practice, the link between potential growth and equilibrium real rates is tenuous at best. A far more likely explanation can be found in Ben Bernanke’s famous savings glut hypothesis, which states that real rates were kept low by an excessively high global net savings appetite (low investment appetite is part and parcel of this). At the regional level this was expressed in rising EM but also German and Japanese current account surpluses. Meanwhile, the Anglo-Saxon nations acted as spenders and recyclers of last resort, also very much by virtue of their globally important financial centres, and experienced rising current account deficits. China played a big role in all this and at the time its net savings rate was boosted by the absence of a social safety net and the deterioration of a family safety net because of the one child policy. 

The savings glut hypothesis can be refined to the “safe asset shortage” hypothesis which states that the large and growing pool of savings came alongside an increased preference for investing these in safe and liquid assets. The prime example here is EM/China reserve accumulation. Meanwhile, the sovereign issuers of safe assets did not keep up the pace as the focus was on keeping debt-to-GDP ratio’s contained. Even the Bush tax cuts of 2003 only changed the downward trend in US debt-to-GDP in a more or less sideways trend. As a result, investors sought refuge in other assets deemed safe at the time (EMU peripheral debt). What’s more, the financial sector produced more assets which were widely regarded as safe. The underlying asset class for these AAA stamped derivatives was exactly the one which benefited most from the savings glut, i.e.  the US housing market.  Hence, when the whole house of cards built on irrational exuberance fell apart, the supply of safe assets shrank substantially as well. Rising budget deficits could only compensate for this to some extent and a rising demand for safe assets on the back of risk aversion as well as unconventional monetary policy contributed further to the fall in real rates.

In view of all this, it makes sense to believe we have turned the corner in the real rate story, i.e. the 30 year downward trend will probably morph into a moderate upward trend. In fact, we have already seen the beginning of this but of course we cannot be sure that this will continue. What’s more, the trend will probably be gradual and there could well be a lot of volatility around this trend.  In particular we can make the following observations:

• The improvement in global growth momentum combined with more risk appetite in financial markets will in itself put upward pressure on real rates and should also at some point help in pushing up expected inflation rates (even though this depends on the Phillips curve coming back to life). The extent to which this happens crucially depends on the cyclical position of the economy and, related to that, on the monetary policy reaction. If there is a large pool of idle resources and potential savings, the upward pressure exerted by an improvement in growth will be muted, even abstracting from the monetary policy reaction. The reason is that a decrease in net savings appetite can substantially raise the pool of actual available funds through the increase in overall income (saving is a fraction of this income). In addition to this, monetary policy will act as a drag on the tendency for real rates to rise if the output gap is large and inflation expectations are low. Of course, if monetary policy is successful at some point inflation expectations will rise.

• Over the past year there has been a distinct albeit gradual shift in the Fed and the ECB’s reaction function. The post-crisis reaction function was very asymmetric: Display a vastly underwhelming response to improvements in the data but provide a strong put option against negative shocks in the real economy or financial markets. In a sense, the asymmetry is decreasing now which is certainly visible at the Fed which does react more strongly to solid real data and provides a much weaker put option. For the ECB one can say the response to better data is still very muted but the put option is less strong. In addition and partly related to this, it seems the ECB and the Fed have become more tolerant to the prospect of inflation undershoots in the foreseeable future. Another way of saying this is that they no longer seem to want to return inflation to target as soon as possible. Instead, both Yellen and Draghi, in their own ways, advocate patience by putting more emphasis on the direction in which inflation is travelling rather than its level. Also they pay more attention to indicators of future inflation pressure (e.g. labour market slack) and downplay their forecast horizons somewhat. The reason for this seems to be an acknowledgement that monetary policy faces political constraints (ECB) or concerns about unintended negative consequences of a very easy policy stance (labour market overshoot, financial stability). Finally, central bankers clearly feel that deflation risks have largely vanished. In our analysis we have certainly noted this shift in central bank behaviour, but we have also seen on many occasions that the market tends to exaggerate this. When that happens, a new tantrum is effectively born because in unconventional space the market can get much more ahead of itself than it did in the past. One very important reason why the risk of the market getting too hawkish will remain present is that the shift in central bank reaction functions could well trigger a more persistent inflation shortfall if it is indeed the case that inflation expectations have fallen below the target. Obviously, this risk is biggest in Europe but it is also present in the US. Some readers may ask why our Fed call (three hikes until the end of 2018) is still more hawkish than market pricing then. First of all, our Fed call is based on what we think the Fed will do and not on what it should do. It will take a while before continued below target inflation will convince the centre of gravity on the current FOMC to change course. Of course, there remains a lot of uncertainty about whether the inflation slowdown is structural or temporary and we could well be wrong in our conviction that structural factors play an important role here. Secondly, the centre of gravity on the FOMC will change considerably due to Trump appointments, which is why the market may only think one rate hike ahead and remain agnostic about what happens after that. Furthermore, the light between our call and Fed pricing may well be less because of a negative term premium in the US money market curve.

• One aspect of monetary policy which is worth mentioning separately is the trajectory of the G3 central bank balance sheet. Fed balance sheet roll-off will be put on auto-pilot while the ECB is expected to stabilise its balance sheet towards the end of 2018. Meanwhile, the BoJ balance sheet is endogenous to the yield target. All in all, this will start to exert upward pressure on term premiums. At first this will be pretty muted, but once the aggregate balance sheet starts to shrink in a year’s time the effect could grow. In addition to all this, China’s pile of FX reserves is likely to fall further.

• Fiscal policy was a big drag on real rates between 2010-13 when concerted fiscal expansion hit the global economy in  the face of what was probably a pretty large multiplier (if the multiplier is x than a 1% change in structural deficit will trigger x% change in GDP). Since then fiscal policy has been neutral to mildly expansionary. For instance, we saw some fiscal stimulus in China and parts of peripheral Europe (financed by lower interest rate expenses). Going forward, we could get a fiscal stimulus of around 0.5pp in Germany and possibly an EMU-wide stimulus in the form of an increase in public investment financed by the ESM. In the US we will probably see tax cuts which amount to an increase in the structural deficit in the region of 0.5-1pp. Finally, in Japan the target date for reaching a balanced primary budget is likely to be pushed out in the future.

Adding it all up, the “plug and play” economic model thus probably gives roughly the right answer but it is important to emphasise that especially the monetary part of the equation still argues for a much more muted yield response to better growth than typically seen in the past. What’s more, it is not always clear how various risks will affect bond yields. EMU political risks will lead to higher peripheral spreads but lower Bund yields. US political and policy risks could go either way. If these threaten to push the economy into a recession then yields could fall. Meanwhile, the exact response of yields to a fiscal expansion will depend greatly on the expected fiscal multiplier as well as any expected supply side and monetary policy reaction. Also, uncertainty about the future course of the Fed because of personnel changes may well in itself raise the term premium. Against that, if Trump appoints low rates/weak dollar proponents this could lower the expected path of the policy rate but over time it could also raise the inflation risk premium.

Willem Verhagen, Senior Economist