-2020全球经济展望（宏观）.docx

Global Research 7EconomicsGlobalArend KapteynEconomist +44-20-7567 0531Seth CarpenterEconomist +1-212-713 4173Reinhard CluseEconomist +44-20-7568 6722Rafael De La FuenteEconomist +1-203-719 7127Gyorgy KovacsEconomist +44-20-7568 7563Pierre LafourcadeEconomist +1-203-719 8921James MalcolmEconomist +81-3-5208 6214Edward TeatherEconomist +65-6495 5965George TharenouEconomist +61-2-9324 3520Tony VolponEconomist +55-11-2767 6337Tao WangEconomist +852-2971 7525November 2018Global Economic PerspectivesGlobal Economic Outlook 2019-2020The end of "low for longer"The aggregate global output gap closed in Q2 this year, median unemployment is at the lowest level since 1980, wage pressures in many developed markets are approaching post-GFC highs (and in Japan at two-decade highs), labour shortages in several European countries with high unemployment seem as acute as in Germany (3% unemployment), and 72% of countries already have headline inflation running above the central bank's target-a share greater than during the 2003-2007 boom years. Economically, the global economy does not look like a patient in need of more monetary life support.But the evidence of being "late cycle" is at best mixedWe assessed 120 prior "late cycle" episodes for 40 countries, and found that on several dimensions the behaviour of the data over the last four quarters in the US, Eurozone and Japan is completely incongruous with any of the recessions that took place since 1980. Employment growth in Japan, house price and productivity growth in the US, core inflation in Europe, and various aspects of consumption and investment behaviour are not consistent with being near a business cycle peak. The central banks are thus understandably cautious. Despite core inflation in DM expected to be back above 2% by end 2019, on par with pre-crisis levels, central banks are expected to lift their policy rates to only half of 2008/2009 levels by end 2020 and still be below neutral.And our now-casting models suggest the global economy is skiddingWe launch our global now-cast in this report and show how the deterioration in global 'hard' and 'soft' data is the most severe since the Eurozone crisis. We view the softness as transitory and we project consumption and investment to re-accelerate in 2020. But the knock to confidence this year is notable: 'survey' implied growth globally has fallen from 4.9% to 3.5% YTD.Where do we deviate from 2019 consensus?We are significantly below consensus growth forecasts in 2019 for a large number of countries, including recession-hit Turkey (-160bp below consensus) and Argentina (- 200bp), several of the Asian economies (-80bp in Thailand and Malaysia), and also the Eurozone (-20bp). We are significantly above consensus on Japan (+70bp next year on growth and a similar magnitude on inflation) and Brazil (+70bp on growth). We remain more cautious on the US near term than most, because of tariff concerns, but are more constructive post-2020. The fiscal cliff is more like a gentle h川一we've raised potential growth to 2% in 2020 and believe the Fed, after a pause, resumes hiking in 2021.What can go right?The biggest upside risks to our forecast are (i) trade tensions de-escalating; (ii) trade diversion effects being stronger than we think (we show for several Asian economies how market share gains could outweigh value chain disruption); and (iii) the US growth acceleration being more structural.What can go wrong?The biggest downside risks are (i) central banks falling behind the curve if inflation surprises to the upside (on just a 50bp deviation from our inflation baseline we would see policy rates 70bp higher in DM); (ii) the US fiscal cliff being steeper than we project; (iii) US-China tariffs getting extended to all goods (we modelled the implications of USD/CNY potentially moving to 8.0); (iv) European peripheral debt stress intensifying on Italian debt concerns; (v) Brexit (we've estimated the global cost of a 'hard' Brexit at roughly half of the US/China trade tensions); and (iv) oil prices being significantly above/below our baseline, which has a large (transitory) effect on policy rates in EM.This report has been prepared by UBS Limited. ANALYST CERTIFICATION AND REQUIRED DISCLOSURES BEGIN ON PAGE 227.softens. This is somewhat puzzling because domestic demand in the G3 looks healthy and, in the case of the US, manufacturing production-which is typically comparatively import-intensive-has picked up significantly. Perhaps the overall slowing in US investment YTD (ex-inventories) is the culprit, as it seems unlikely tariffs played a role earlier in the year. But the puzzle extends to the Eurozone and Japan, where import demand is slackening despite robust investment patterns (while confidence indices seem to be turning, capex intention surveys are holding up better than broader business sentiment ones). Whatever the case may be, the deceleration in trade volumes and prices has carried through to global industrial production and pulled the global economy down from last year's lofty levels. Currently, EM Asia seems to be virtually holding up global import demand on its own. That is not sustainable, and the implementation of US/China tariffs should weigh yet more heavily on trade volumes.Trade aside, however, our baseline forecast for '19 and '20 is more constructive. We expect global economic growth to settle at roughly its long-run level (3.7%). We generally still see ample slack to contain any undue acceleration in core inflation andprovided we are right-markets seem to be pricing central banks' intentions pretty fairly for 2019.By the time we reach *20, however, things get a lot more uncertain. We spend considerable time below going through six alternative scenarios, including one where DM central banks fall behind the curve. The range of outcomes for growth, inflation and policy rates in those scenarios is vast, and much more interesting than the baseline.The difference between trade tensions escalating and de-escalating either puts us 40bp below the global growth baseline or 20bp above. Trade escalation would lower our DM policy rate baselines substantially but increase them for EM (see our Scenario analysis further down for detail). Trade de-escalation does the opposite.The largest deviation from our baseline would come from us being wrong on the steepness of the US fiscal cliff. We think of it more as a gentle hill and have penc川ed in a minor growth drag. If we are wrong on that, combined with being wrong on a number of other assumptions, we get dramatically different outcomes for global growth (lower by nearly 100bp), in part because of the disruption to financial markets. Most DM central banks would return to the lower bound, the Fed would resume QE, and despite some pressure on EMFX, even EM central banks would be cutting rates.An alternative scenario where central banks fall behind the curve on inflation produces somewhat similar growth dynamics but is ultimately much less damaging in that the shock is self-correcting. The journey is volatile, however, with DM central banks raising their policy rates 70bp above our baseline if we are only 50bp wrong on our inflation forecast.We also explore the possibility that the US accelerates away from the rest of the developed markets. Our prior was that the associated business and policy rate divergence would create financial market volatility. It does that to some extent but productivity gains reduce inflation pressure in the US at the margin, which mitigates the additional Fed hikes (though we see fed funds at 4% in that scenario). Even without any structural improvement in growth elsewhere, we view this as having a 'de-anchoring' effect on other central banks and lifting their policy rate paths.disappointedD. failed33. Her sister has become a lawyer, she wants to be.A. whichB. whoC.whatD. thatIt is often said that teachers have very easy life.A. the; aB. /; /C. /;aD. the; /34. This girl is Linda's cousin.A. pretty little SpanishB. Spanishlittle prettyC. Spanish pretty littleD. littlepretty Spanish35. -You know, Bob is a little slow understanding, so-一So I have to be patient him.A. in; toB. on; withC. in;withD. at; for36. I really can't understand her like that.A. you treatB. you to treatC. whyGlobal Economic Perspectives 7 November2018treatD. you treating37. I send you 100 dollars today, the rest in a year.A. followsB. followedC. tofollowD. being followed38. in the mountains for a week, the two students were finally saved by the local police.A. Having lostB. LostC. BeinglostD. Losing39. The manager,it clear to us that he didn't agree with us, left the meeting room.A. who has madeB. having made C.madeD. making答案：BCDAD DCCBB ACBDA DBDAC ADBDA DBBAC ADACA CDCBBGlobal Economic Perspectives 7 November2018A global now-castNotwithstanding the underlying economic improvement, it appears the global economy is hitting another major soft patch. Why worry about that in an Economic Outlook that should look years ahead, not just a few months? Well, the present is obviously the 'base' from which we forecast-a wrong base sets up a bad forecast. More substantively, if we don't understand what's driving the current downturn we have little hope of forecasting the future with any degree of accuracy.As regular readers will know, we use a factor-augmented, mixed-data-frequency Launching our global now-cast model to now-cast DM real GDP growth. We've now extended that to EM, adding a few bells and whistles with which we will shortly retrofit our DM estimates (see Box 1). We combine the two models to generate a global now-cast, covering roughly 200 indicators and 23 countries, and enabling us to explain 86%, 88% and 92% of the variation in EM, DM and global GDP growth, respectively .Figure 13: Global real GDP growth vs nowcastFigure 14: Root mean square errors (the lower the better)% QoQ saarGlobal real GDP - Nowcast model8.000020406081012141618Source: UBS, Haver, CEICThe fit improves cy u。"二 mng hard sndsc'ft data into the TodeWhat are the models telling us? That the global economy may be hitting its worst cyclical slump since 15H1, but the magnitude of the slowdown in 'hard' and 'soft' data (around 160bp relative to growth momentum in '17) is more on par with the '12 Eurozone crisis-induced slowdown (i.e. worse than '15). We are tracking global growth at 2.9% in the third quarter, against 3.5%, 3.7% and 4.5% in 18Q2, 18Q1 and full *17, respectively.4'Soft' data refer to consumer and business confidence surveys, bank lending surveys but also some stock market indicators. Hard data refer to measured levels of activity, e.g. industrial production, trade, employment, credit, retail sales and housing.Global Economic Perspectives 7 November2018The global economy has hit a (large) speed bumpThe global economy has hit a (large) speed bumpHowever, that's just the model. Despite the good historical fit, the model under- predicted global growth by 60bp in 18H1 (roughly one standard deviation) after consistently over-predicting it last year (see chart below left). The miss in Q2 we think was largely related to the US and Japan-two large-PPP-weight countries that accelerated (the model does not assign weights to countries; it extracts the collective co-movement in the data). Preliminary GDP releases for Q3, where available, are consistent with DM growth rolling over, though not as sharply as the DM model is suggesting. We find these now-casts most useful in thinking about the distribution of risks around our forecasts (given the magnitude of improvement/deterioration it is signalling), particularly when there are large shiftssuch as we are seeing now, but we would not get too hung up on the precise levels indicated for a particular quarter.Figure 15: Global growthFigure 16: DM growth% QM Global real GDP - Nowcast model saar% QM Global real GDP - Nowcast model saarGlobal growth2012141618Source: UBS, Haver, CEICSource: UBS, Haver, CEICFor DM, the now-cast is currently suggesting that growth has slowed to a bit under 1% annualized in Q3, down from 2% in Q2 and 1.6% in Q1. The chart above (middle) shows how the model was spot on in Q1 but then was too pessimistic in Q2. Much of the weakness in the model output seems again to relate to the Eurozone, where the now-cast fit is excellent (witness the low root mean square error in the table above), and has consistently been pointing to much weaker prints than expected by consensus (in line with the disappointing flash Q3Figure 17: EM growthFigure 17: EM growth%QoQ EM real GDP - Nowcast modelr.u tEM growth6.5 4- - - EM nowcastr.u tEM growth6.5 4- - - EM nowcast3.0 -(r-12141618Source: UBS, Haver, CEICEurozone now-cast has been softer than consensus all yearFigure 20: EM hard vs soft% QoQ EM real GDP - Nowcast modelSource: UBS, Haver, CEICLatam & EMEA are decelerating;Asia is moving sidewaysGDP release).Figure 18: Global hard vs softFigure 19: DM hard vs softFor EM, the now-cast is similarly pointing to a soft patch, from a little under 6% in 2017 to 5.3% in Q1, 4.7% in Q2 and now tracking at 4.4% in Q3. The slowdown appears concentrated in EMEA (substantial loss of momentum in Turkey following its currency weakness) and Latam (Argentina, and earlier tightening of financial conditions in Brazil which is now reversing). Growth in Asia, in the now-cast, is moving largely sideways. This holds even when we exclude India and China from the data-two countries where there is a particularly poor correlation between high frequency indicators and national account statistics.Figure 21: Latam growthFigure 22: EMEA growthFigure 23: EM ex India/China growthFigure 23: EM ex India/China growthSource: UBS, Haver, CEICSource: UBS, Haver, CEICGlobal Economic Perspectives 7 November2018Global "survey implied growth" has declined from 4.9% in Dec '17 to 3.5% in Q3-18Global "survey implied growth" has declined from 4.9% in Dec '17 to 3.5% in Q3-18One particularly useful metric is 'survey-implied' growth rates, as a proxy for how much confidence in the global economy has deteriorated Strictly speaking our *soft data' does not just include surveys but they dominate within the sample.Global Economic Perspectives 7 November2018. Policy uncertainty is running at historically high levels and it is starting to have an impact on confidence, both in EM and DM. For EM, growth rates implied by the soft data have come down from 6.2% in Q1 to 4.4% in Q3, and for DM, implied growth has slowed from 3.8% in Q4-17 to 2.4% in Q3. Remarkably, US growth implied by soft-data continues to edge up to close to 4% (see chart below right). Animal spirits matter-including for hard data-and they seem to be holding up much