Macro outlook: global growth or US yield curve, which will flatten first?

Can global growth stay on the current 4% plateau? Yes, under three conditions: G7 central banks accept loss of credibility, the US yield curve bull-flattens, and China preemptively launches another public spending spree to cushion from a new housing crash.

Key points:

• A solidly-growing global economy will face a sharp tightening of monetary conditions by G7 central banks in 2018.
• Market belief in the possiblity of a smooth normalization of monetary policies would send long-term bond yields much higher, to levels incompatible with current hefty valuations of asset prices. But elevated valuations seem necessary to sustain global final demand through wealth effects and collateralization of new debt flows.
• Therefore, the bond market holds the key to the continuation of the global boom: it is only in a scenario where yield curves flatten in defiance of central bank's normalization rhetoric that the current regime of global growth and overvalued asset prices can continue.
• It is thus a narrow pass that markets will have to cross in 2018 in order to avoid a new financial crisis. In addition, the risks to the global economy will be exacerbated by a bull market in the US dollar (Fed's balance-sheet contraction), a potential shortage of eurodollars (Basel III's LCR) and a slowdown in Chinese commodity demand (a consequence of this year's bear flattening of the Chinese yield curve). At the confluence of all these aggravating factors lie the unpopped housing bubbles of Canada, New Zealand and Australia. These economies could be the harbingers of the next Minsky moment.

Global growth remains close to its 4% plateau reached at the beginning of 2017. This regime appears robust as growth is both evenly distributed geographically, and, for now, non-inflationary, despite being significantly above potential (estimated at 3.5%). Notably, the boom in US shale prevents the oil price from imposing an immediate speed limit on global growth, as was the case in the previous cycle.

So, what could go wrong? Two things: attempts to normalize monetary policy in advanced economies, and the Chinese financial cycle.

Central bankers remain puzzled by the lack of inflationary pressures in G7 economies, but nevertheless are about to collectively impose a sharp monetary tightening to the global economy in 2018, be it for financial stability reasons in the US (the Fed fearing an equity and housing bubble), technical reasons in Japan (shortage of JGBs), political reasons in the euro area (33% cap on ECB holdings of any country's sovereign debt), or because of a return of the external constraint in the UK (widening current account deficit, excessive household leverage). Central banks rhetoric aims to convey the message that this inicipient 'normalization' of monetary policy (short rates, balance-sheet) and yield curves should not entail any major macro-financial disruption. We disagree.

The level of US asset valuations (equity, credit, housing) does not seem to us compatible with the bear steepening of yield curves entailed by this scenario of 'smooth' monetary normalization and contemplated by the market consensus. Coincidentally, the current demanding valuations of asset prices have been, since the beginning of this recovery, necessary to feed final demand for goods and services through the working of wealth effects (and their partial monetization via equity withdrawals - although this channel has been less important than in the previous bubble).

I will just present two charts that illustrate the overvaluation of US assets.

Regarding equity, we look at the equity risk premium. We first compute an Equity Expected Rate of Return, using a classic Dividend Discount model. We use the path of future dividend growth as reported by the consensus of equity analysts for the next 5 years. Beyond the 5 year horizon, we make the hypothesis that dividends grow at the average nominal GDP growth of the last 10 years (which is quite generous considering the future drop in potential growth). We then deduct the 10-year TIPS rate from this Expected Rate of Return to obtain an Equity Risk Premium (ERP). As a matter of fact, it stands out that analysts' dividend forecasts are extremely inert, i.e. they do not exhibit any cyclical swings, so that the ERP still embeds some large short-term and medium-term variations. As shown on the chart below, the multi-year cycles in the ERP are indeed highly correlated with the long cycles in productivity (proxied as a 5y moving average), with an advance of roughly 4 year.

In plain English, the equity market adjusts the equity risk premium to its expectations of future (roughly 4-year ahead) productivity growth. Sometimes the market overestimates future productivity gains, as it did during the IT bubble: as is visible on the chart, the ERP tightened to almost 4% in 2000 (keep in mind the ERP is shown with a 4-year lag in the chart), thus anticipating productivity gains accelerating to 4.5%, while productivity ex post peaked at 3.5%.

What does this chart tell us for US equity valuations today? Basically that the market prices in future productivity gains averaging close to 3% in the next 5 years. Productivity gains have hovered around 0.5% in the past half-decade, so this scenario factors in an acceleration in trend productivity similar to that of the 1990s with the IT revolution. Although technological progress could generate a cluster of new products and process innovations (AI, robotization, digitalization, uberization, ...), we doubt the US society can deliver such hefty efficiency gains given the collapse in education standards, health levels (opioid epidemic, obesity), weakness in productive capex and rising threats of more protectionism (NAFTA, KORUS under threat) and overall populism.

For now, it is hard to pin down a catalyst for a reversal of this bull market: the US earnings season is solid (5.9%yoy growth in Q3), and households seem to be on a buying spree. According to the University of Michigan survey, on average, households see a 65% probability that equities will rise next year, a record high for that series that started in 2002 (after the burst of the IT bubble). These retail inflows seem sufficient to neutralize the slowdown of share buybacks. But clearly, a spike in the 10y rate would destroy the argument of the 'cheap' relative value of equities versus bonds. And of course, the sensitivity of valuations to long-term bond yields is even more stringent in the case of real estate assets.

The same reasoning applies to the US housing market. The post-2008 relapse in rental yield seems at odds with the collapse in the trend in productivity gains. The house price is the present discounted value of the expected stream of future rentals, and rentals should follow the trend growth of inflation plus productivity. As illustrated in the chart below, rental yields should have increased, not decreased, as productivity gains further collapsed in the 2012-2017 period.

Therefore, the bond market holds the key to the continuation of this global growth regime:

1. either yield curves bear-steepen, crashing risky assets, and the global boom ends abruptly;
2. or yield curves bull-flatten, and the current regime of strong growth and inflated asset prices continues for a while -- but the further flattening of yield curves would be in open defiance of central banks' thesis of 'normalization'.

In fine, central bankers will either accept their total loss of credibility, or will crash the global economy.

We insisted last month on the key role played by the Chinese yield curve on the domestic real estate cycle. The dire predictions inferred from the flattening of the curve seem to be panning out, as could be seen in the latest batch of RE data in China. Notably, our preferred gauge of the real estate cycle, the growth of areas sold, further slowed down to 10%yoy, compared with +36% a year ago. As illustrated in the chart below, the path of the 10-year rate tells us we are in for a new crash similar to 2014.

Given the rising weight of services in Chinese GDP, real estate woes might be less likely to generate an overall downturn in the economy. And anyway, we are (no longer) trying to time a hard landing of the Chinese debt-fuelled growth regime. As long as the government retains control of enough national savings through public banks (i.e. there is not too much diversion of savings towards the uncontrolled shadow banking system) and keeps the capital account closed, there will always be the possibility to cushion any bust with large-scale fiscal spending -- at the cost, of course, of more "malinvestment" and further departure from a market economy, leading the Chinese economy back to a soviet-style centrally-planned system. The endgame will be a loss of trust in the public banks resulting in flight from all form of bank deposits and debt-backed assets, and/or social upheaval against corruption and inequalities. But it is not our purpose here to time this Black Swan event. What matters for our market analysis is to time the ebbs and flows of Chinese demand for raw materials, because these are the main drivers of commodity prices, and thus determine the fate of many commodity-oriented economies, in emerging markets (Russia, Brazil, Chile...) and the developed world (Canada, Australia, New Zealand). In other words, when Chinese bubbles go bust, a hard landing is more likely to take place in, say, Brazil than in China...

In short, because the PBoC aimed at reining in shadow lending and containing capital outflows, the Chinese yield curve has already bear-flattened, and thus is alreaday threatening the recovery in some commodity-heavy economies. As shown below, a flat yield curve is dragging the Chinese manufacturing PMI towards recession territory, and could soon end the rally in base metal prices.

As a result, faltering commodity demand from China will take its toll on large economies around the world, from Brazil to Australia.

The last important point we want to make regards the dollar. As we pointed out last month, we believe the Fed's balance-sheet contraction will fuel tensions on the availability of eurodollars, as is already visibile in the cross-currency swap market. The prospect of a tax holiday for corporate profit repatriation would also be a strong tailwind for the dollar. Countries with large funding needs in dollars and dependent on raw material exports to China will thus be particularly vulnerable in 2018. Some emerging market economies come to mind: Colombia, South Africa, Chile, Indonesia... but Australia, Canada and New Zealand also fall in this category, with the additional feature of hosting a real estate bubble in part financed by Chinese capital flight...

Canada, Australia and New Zealand exhibit the dangerous combination of (1) persistent current-account deficits, (2) large household debt burdens, (3) strong dependence on China's commodity demand, and (4) runaway house price inflation in the past decade (see the 3 charts below).

As an aside, Sweden also concentrates some of these vulnerabilities (high household debt burden, house price inflation), but what looks like a Swedish housing bubble does not seem to be dependent on foreign capital inflows: the country has a structurally positive current account balance. In addition, Sweden is less dependent on the import demand from the Chinese construction sector, and Stockholm real estate does not seem to be a magnet for Chinese capital flight, unlike Toronto, Vancouver, Sydney and Melbourne.

The degree of dependence of these housing bubbles on dollar-denominated funding can be assessed through the cross-currency swap market. Australian and New-Zealand banks seem to be the most exposed as the Aussie and Kiwi swaps exhibit a positive basis on the dollar leg (on the 5y-maturity in the chart below), while the basis is structurally (post-2008) negative for the other G10 currencies (EUR, JPY, GBP, SEK, CAD). This suggests that Aussie and Kiwi lenders issue debt in USD and then swap the proceeds for their domestic currency. On the contrary, Euro and Japanese banks raise funding in their domestic currencies and then swap the proceeds for dollars in order to fund their large portfolios of USD-denominated assets. The fact that the basis has turned negative for the CAD might be a mitigating factor for the Canadian situation.

The Canadian, New-Zealander and Australian housing bubbles were amongst the rare ones to survive the 2008 market crash, but could well be the first to blow up in the next crisis. The following quotation from the Australian Treasurer Scott Morrison in mid-2017 gives an idea of the degree of preparedness of the authorities:

"Access to capital is what enables this country - given that we are a net importer of capital and always have been - to have a strong banking and finance system the world can have confidence in and Australians can have confidence in."

If there is one commodity that is never in short supply on financial markets, it definitely is hubris.

Raphael Gallardo

@RaphGallardo

Written on October 25, 2017

This is a personal post, that is not sponsored by my employer. The opinions expressed here represent solely my own. The content of this post does not constitute investment research or advice. Data series referred to in this note or in the charts are available from public sources.