廉价的资金更好的收益还有多少上涨空间.docx

Global ResearchAPAC Equity StrategyCheap money, better earnings. How much upside?We lift our MSCI Asia ex Japan target to 670 but struggle to see much upsideWe were bullish at the end of March but the extent of the rally has surprised us. Cheap money and recovering earnings are instinctively a positive backdrop. However, despite a desire to be upbeat, we're struggling to use historic precedent to make a bull case now after a 40% rally from March lows. Our valuation models, sentiment and earnings analysis point to some downside by year end.Earnings not as bad as we feared - we now expect -10% EPS growth this year In early April, we cut our earnings forecasts for this year to -25%, with a high degree of uncertainty. Our models, macro and micro data, signals from the early reporters in the 2Q earnings season and a weaker US dollar now suggest that earnings are not likely to be as bad as we feared. We now project -10% EPS growth this year (consensus at -2%) and 36% growth for next year. The rise in our 2021 EPS estimate (in line with consensus) drives the increase in our index target from 625 to 670.Better earnings appear priced-in. Our index target Is 4% below current levelsEquities are trading at 15.6x forward earnings, 1 standard deviation above mean and just shy of their levels in the early stages of recovery in 2009. The difference then was a near 40% peak-to-trough earnings decline, and a bigger rebound to come. That's not the case today. We have sympathy with the view that low rates will support higher valuations, partly on an asset allocation switch for yield. However, there is little evidence of this - high yield stocks with stable dividends have underperformed since early April suggesting limited demand (so far) for equity income in a zero rate world.Calibrating the (non earnings) upside scenarios - another 13% to come?This could be an abnormal environment with both a cyclical improvement and yield curve control limiting any back-up discount rates. As a scenario analysis, if 1. cyclicals rebounded in line with their multiples in past recoveries (we are overweight Korea and Japan to play this theme), 2. higher yield stocks re-rated in line with their yield gaps to government bonds, and 3. growth stocks re-rated in line with lower discount rates, combined this could push the market up 13.3% from current levels leaving the market on an eye-watering 12m forward P/E of 17.8x today. That's not our base case, but a credible upside-risk scenario. We think equities are fully priced for now, and investors are better paid to play relative returns with our core call being to overweight Korea and Japan where we still think valuations and recovery interact well together.6 August 2020Equity StrategyAsia PacificNiall MacLeodStrategist +852-2971 6186Matthew GilmanStrategist +852-2971 8173Jiamin ShenAssociate Strategist +852-3712 3126This report has been prepared by UBS Securities Asia Limited. ANALYST CERTIFICATION AND REQUIRED DISCLOSURES BEGIN ON PAGE 21. UBS does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision.Figure 21: Theoretical Upside to get the Yield Gaps back in sync with long run averagesSource: Refinitiv, UBSYield Gap to Local Currency, Regional Government BondsYield Gap to US 10 year TreasuriesAverage DY-BY gap since July 2009-0.530.44Current Gap-0.101.80Current DY2.412.41Current BY2.510.61Implied DY if DY-BY gap returns to 'average'1.981.05Equity upside to return the yield gap to average21.8%129.2%While we have sympathy with the view that in a yield starved world, investors may be required to take on the higher risk of equities to meet income requirements, there is scant evidence of this having worked significantly under the QE environment post 2009 - as Figures 20 and 21 show, the yield gap shifted higher as bond yields dropped on QE.And for the purposes of our current view, there's no evidence that yield stocks have been leading this rally, which they ought to have been doing if the market was being driven by the hunt for income. Yield stocks have underperformed year to date and since the bottom of the market.Figure 22: But dividend yield stocks are not re-rating, or at least outperforming. This market Is not at this point being driven by high yield gaps.Source: Refinitiv, UBSA narrow market driven by momentumThe evidence doesn't point to the market being driven by conventional yield seeking asset allocation flows. Instead the gains in the market are actually surprisingly narrow, particularly compared to the last major cyclical recovery in 2009.Source: Refinitiv, UBSNot only is stock leadership narrow, but market leadership is also narrow, with China dominating returns. Interestingly however within China, stock returns are quite broad.Figure 24: China relative performance vs. North and South AsiaSource: Refinitiv, UBSFigure 25: Distribution of YTD stock performance in MSCI China%。欠§ %够。寸dpi %。¥。85- %ocooCNdn %oeoL5 %0T0dn%0v0 UMOQ %OCXJOL UMOQ %ocooe UMOa %070coUMOa %。寸A UMOQSource: Refinitiv, UBSOne market or multiple markets? Upside risks.If we are wrong, and equities keep surging, what are the potential sources of those returns other than better earnings? We think the upside comes from various 'mini' markets or buckets or styles performing simultaneously -The cyclical recovery comes through as we expect, and traditional cyclical stocks re-rate in line with the move in PMIs.1. Yield curve control restricts a bond yield rise. Investors start to price in a sustained low rates, and "carry trade“ equities start to re-rate to new absolute highs. This is unusual of course, given that cyclical recoveries have normally been associated with steeper yield curves.2. Growth stocks continue to re-rate, on the expectation of sustained low rates.1. Cyclicals rally into the recoveryOur analysis of the (ex-internet) cyclicals, shows an unsurprisingly strong correlation of valuations to the levels of PMIs. While PMIs don't capture the degree of economic activity (just the breadth of whether things are getting better or worse), our historical back-tests show PMIs give better and consistent signals about cyclical stocks than other indicators like industrial production.Using the historical relationship and plugging in some 'what if' scenarios based on the recovery in 2009 points to another 5% upside for cyclicals to 'fair value' by year end.There are several caveats to this - one, as we said, PMIs don,t measure degree of recovery, just whether this month's data is up or down. Two, while our economists are expecting a strong sequential recovery in activity, of a much deeper contraction than in 2008/9, their forecasts also have economies taking far longer than 11 years ago to recover back to their prior peak. But this is a 'what if' analysis, and isn't designed to be a base case.Given similar PMI contractions this time versus 2008 by magnitude, if we assume the economic recovery would largely follow the same path. Historical trend implies the PMI to return to 58.5 for China and 64.1 for US by the end of this year.Figure 26: US manufacturing PMI new orders 2008 vs todayFigure 27: China manuf. PMI new orders 2008 vs todayPutting the assumptions together, a normalized PMI for the US and China would imply 5% re-rating for cyclical (ex-internet) stocks in Asia ex Japan by the year end (see the regression behind this in the grey box on the following page).Figure 28: Calibrating the upside based on a similar cycle to 20091.671.754.8%Current cyclicals P/BVModeled P/BV by year endImplied cyclicals upsideSource: UBSCalibrating the potential upside for cyclicalsCyclical sectors are by definition positively correlated with the macro economic cycle. One of the most common factors we look at to gauge economic activities is the manufacturing PMI new orders data, which tends to provide a timely guidance ol the business cycle.In order to calibrate the cyclical (ex Internet) upside, we correlate the price to book valuation and the PMI data from the US and China. Based on that relationship, we can calculate the potential of cyclical re-rating given a recovery in PMI presumably as the economy normalizes by the year end.Figure 29: Cyclicals (ex Internet) price to book vs ISM manufacturing new orders indices for the US and Chinam u -wd ouunoBJnuBzNooaSflo 68 >vm u -wd ouunoBJnuBzNooaSflo 68 >v0 5 0 5 0 57 6 6 5 5 4sj<ddo403530R2 = 0.3547Source: Datastream, UBSWe then build a single variable regression to quantify the correlation between the two. In summary, the combined series ol PMI from US & China explains 35% of the trend in cyclical (ex Internet) P/BV. Every 1 point increase in the combined PM series would imply a 2.6% cyclical re-rating.Figure 31: Regression statisticsFigure 30: Actual vs modeled AxJ cyclicals (ex Internet) P/BV2.42.221.81.61.41.210.80507091113151719Regression StatisticsMultiple R|60%R Square35%Adj R Square35%Std. Error17%Observations186Coeff.Std. Err.t StatP-valueIntercept0.1340.1390.9630.337US&CN ISM0.0260.00310.0563.09E-19Source: Datastream, UBSSource: Datastream, UBS2. An asset allocation switch to higher yielding stocksHistorically, as recoveries have come through, yield curves have steepened. However in this cycle our economists now expect the Fed to follow the BoJ and introduce some form of yield curve control (YCC).We showed earlier the overall market dividend yield gap measures. What if low bond yields cause a shift into other income producing assets, compressing the dividend yield on equities?We run a simple simulation of looking at the 'high yield' dividend income universe in Asia ex Japan (defined as stocks with a yield over 3.5% at start of this year), and compared the yield to both US and local government bond yields over time. It*s more comparable to look at the gap between yields after 2010, when the second round of QE from the Fed caused a more dramatic sustained drop in bond yields, than the period before this.Figure 32: Equity High Yield universe dividend yield gap to local currency Asia government bond yieldsFigure 33: Equity High Yield universe dividend yield gap to US government bond yieldThese show clearly that there*s a large yield gap in equities* favour right now. Realistically however, this partly reflects a likely cut to dividends. The dividend cut in 2009 for this group was 25%.Figure 34: Theoretical upside from high dividend yield stocks /f the yield gaps returned to average levelsSource: Refinitiv, UBSBased on Local Currency 1 Oyr Government Bond yieldsBased on US 1 OyrTreasury yieldsAverage Yield Gap to0.8%1.8%Current bond yield2.4%0.7%Current DY assuming 25% cut in DPS4.0%4.0%Dividend yield gap assuming 25% cut in DPS3.1%2.1%Required Dividend Yield to return Yield Gap to average3.2%2.5%Price Appreciation to return Yield gap to average level22.7%56.9%Price Appreciation if US IO Year yield is 1%-39.3%However, even factoring a 25% cut in dividends, if investors decided to seek income in equities, there*s clearly scope for a re-rating of this segment of the market. For sensitivity analysis sake, even factoring in a 1% 10 year yield would still require a 39% re-rating of equities to take the dividend yield- bond yield cap (cum a 25% cut in dividends) back to long run average. A mid-point of the two, would suggest equities could rally around 30%, leaving the high yield universe on a dividend yield of 3.1%.3. Growth stocks continue to re-rateRather than use conventional growth/value indices, to analyse growth stocks, over the last few years we've constructed our own subjective list of the core 'growth stocks* in Asia ex Japan, based on consistently high ROE, high P/E and EPS growth rates (see the appendix for our current updated list of these stocks, with weighted market caps over US$5bn). The list is dominated by a handful of mega caps.Figure 35: G40 historical valuation - forward P/ESource: Refinitiv. UBSIn order to quantify the theoretical impact of lower bond yields on valuations, we have constructed a simplified DCF model. We are well aware that companies in practice might follow different valuation framework and the assumptions are quite rough, but we believe this still could give us some directional guidance as an efficient technique to calibrate the upside risks.We first take normalized market cap, consensus earnings estimate and local government bond yields before COVID-19 to back out the implied long-term growth assumptions from year three into perpetuity. We assume these long-term estimates for growth haven't changed with COVID-19.Using this implied growth rate, current EPS forecasts for years 1-3, a constant risk premium of 5%, and the new bond yields for each stock (current local government bond yields, but we've used US 10 years for the US/HK listed stocks), we can then create a new *fair value* market cap.Figure 36: Upside breakdown of the G40 basket (rebased to 100)Source: Refinitiv, UBSRe-rating potential% of market capUpside contribution X market cap4.8%36.3%1.8%30.0%18.5%5.6%23.6%25.1%5.9%79.9%13.3%Figure 37 shows the upside ranges for how much each of these scenarios could provide, from current levels of the index.Figure 37: Cydicals (ex internet)High yielders going back to yield gap averages(ex- cydicals, to avoid double counting)G40 re-rating on low discount ratesCombinedSource: UBSIn total, if all three scenarios played out simultaneously, covering 80% of market cap there would be potentially 13.3% upside for the index, putting the market on 17.8x forward earnings - levels not seen outside of bubbles.AppendixG40 basketNameSectorMarket cap ($mn)12m fwd P/E5 years average P/EAlibaba Group Holding Adr 1:8CNConsumer Discretionary697,64027.227.3Tencent HoldingsCNCommunication Services666,38034.431.2JD Com Adr 1:2CNConsumer Discretionary85,80637.747.7Housing Development Finance CorporationINFinancials40,23229.032.4Baidu 'A' Adr 10:1CNCommunication Services32,86815.623.7TAL Education Group CI.A Adr 3:1CNConsumer Discretionary32,34874.757.9NaverKRCommunication Services43,29152.227.8CelltrionKRHealth Care33,81070.953.8Hindustan UnileverI