2017年展望：微妙的平衡（英文版）

摘要：流动性渐紧，大牛市很难。中国的实际利率已经下降到了历史新低。历史上，如此低的实际利率往往预示着加息或存准的上调。史无前例地低的实际利率水平是泡沫以及人民币快速贬值背后真正的罪魁祸首。中国经济很可能延续其“L形”轨迹。其实，2012年以来一直如此。

市场共识认为房地产调控将把资金从房地产“倒逼”进入股市。然而历史经验证明，每一次的房地产调控之后，市场表现都令人失望，一如07年9月，10年1月和11年4月房地产调控后所见。市场共识未能认识到房地产成交其实有着货币乘数效应，并加快货币流通速度。因此，任何遏制房地产的政策都将导致流动性紧缩，拖累股市。

如选择利率稳定以熨平去杠杆期间的市场波动，那么人民币汇率将不得不承受经济调整的重担。如果资本管制被用于缓冲在岸人民币CNY贬值的压力，则离岸/在岸人民币汇率将分化。这种汇率价格的背离如果超过一定程度，也将迫使央行进行市场干预。但是，离岸利率的提高最终还是会触发其他资产类别价格的波动，例如股票市场。这是一个非常微妙的平衡。

与广泛看涨的市场共识相悖，我们的股债收益率EYBY模型估算2017年（今后十二个月）上证综指可能的合理交易范围为3300 +/- 500。恰恰在十二个月前，我们的这个模型估算2016年的合理交易区间为2900 +/- 400 。这个结果显示2017年或比2016年要好，但波动幅度更大 。此外，在我们模型估算的可能的情景里，有2/3的结果低于目前的指数约3300点的水平。尽管美元走强，南向资金流动也应该有助于抵消香港要面对的资金流出的压力。

以下是今天英文报告《Outlook 2017: High-Wire Act》的全文。

------------------------------

Summary - It is hard to fathom a raging bull market amid tightening liquidity. China’s real interest rate has fallen to new lows that used to portend interest rate or RRR hike. It has been the real culprit behind the bubble in bond, property and commodities, as well as fast depreciating CNY. While inflation expectation is rising, growth may not eventuate, as investment will likely fall with property curbs. As such, the Chinese economy is stuck between mild cyclical reflation and outright stagflation. It will continue to traverse an L-shaped trajectory, as it has been since 2012.

For now, more restrictive property curbs are chosen instead to deal with surging property prices. Consensus believes that such moves will push funds from property to equities. But in the past, after each property curb, disappointing market performance, or even plunge ensued – as seen in September 2007, January 2010 and April 2011. Consensus has failed to recognize that property transactions are indeed monetary multipliers, and can accelerate money velocity. As such, the curbs on property transactions will cut liquidity, and become a drag to market performance.

If interest rates were to be kept stable to sooth the volatility during deleveraging, then CNY will have to bear the brunt of economic adjustment. If capital control is instigated to slowdown onshore CNY depreciation, then on/off-shore exchange rates will divert, obliging market interventions such as cutting offshore RMB supply and raising offshore RMB borrowing costs. But that higher interest rate offshore will eventually roil the other asset prices, such as equities. Something somewhere will have to bend sometime – it is a high-wire act.

While bonds will likely underperform equities, how far equities can rise will depend on how fast equity valuation expands relative to bond yield’s rise. In general, in a tightening liquidity environment, valuation multiple will compress. Contrary to bullish consensus, our EYBY model estimates the likely trading range in 2017 for the Shanghai Composite to be 3300 +/- 500, suggesting perhaps a better year than 2016 (the same model estimated 2900 +/- 400 for 2016 exactly twelve months ago) - but with wider return dispersion to reflect rising volatility. Further, 2/3 of the estimates are lower than the current index level of ~3300. As such, we remain guarded. With the expanded connect scheme and a depreciating CNY, southbound flows should help offset outflows from HK due to a strengthening Dollar to an extent.

2017 is destined to be a year of epic changes and volatility. We should focus on convexity trades with option-like payoff and think in probabilistic scenarios, instead of being overly engrossed by the perpetual futile debate of bull vs. bear. We see a strengthening Dollar, rising inflation and long bond yield, as well as a weakening CNY. In the first half, there should be opportunities in financials, materials, energy, industrials, tech and consumer discretionary.

------------------------------

“To light a candle is to cast a shadow.” – Ursula K. Le Guin

A Rate Hike?

Funding costs volatility suppressed, leading to the “void of yield” and highly-leveraged trades: Primary funding cost, measured by the 7-day repo rate that the PBoC prices its liquidity provision, has been kept steady since the burst of China’s stock market bubble in June 2015. The supposed objective was to sooth volatility during systematic deleverage in equities. Funds have subsequently rotated out of stocks into bonds. With stable funding costs, traders leveraged their positions to buy bonds in order to earn a bigger carry. Over time, more and more funds flooded the bond market. As yield plunged, traders had to lengthen their portfolio duration and borrow extremely short to squeeze out an interest spread – till it disappears. At one stage, over 90% of market borrowing was overnight.

Since the beginning of 2016, the gap between China’s 10-year treasury yield and the 7-day repo rate has narrowed consistently for the longest stretch in history. Traders lamented “the void of yield”, as they added more leverage (please see our report “Outlook 2016: the Chinese Curse” on December 9, 2015). In the end, this trade is like picking pennies in front of a bulldozer. Recently, the 10-year yield has double-bottomed, and started to rise – especially after the US election, which appeared to have wakened the inflation expectation globally (Focus Chart 1).

Focus Chart 1: The spread between 10-year and 7-day repo at its narrowest for the longest stretch in history. 

Historic-low real interest rate feeds asset bubbles: China’s real interest rate, after adjusted for inflation in goods, services and property prices, has fallen to its new low in history (Focus Chart 3). Such low rate has encouraged reckless risk taking, leading to housing, bonds and commodity bubble. Recently, the PBoC has been gradually increasing the proportion of longer-dated repos (Focus Chart 2). Such maneuvers are to discourage bond traders from excessively shortening the duration of their borrowing – a precarious trade. Yet as property and commodity prices are surging relentlessly, real rates are falling faster than these open-market tactics could catch up. A rate hike, or further curbs on property, or even both, are necessary.

Focus Chart 2: Repo duration has been extended, but real interest rate is falling faster. 

Already, we have seen a bout of new curbs issued on property purchases. These new measures to deflate the property bubble include increasing the percentage of down payment, restrictions on second-home purchase, tightening lending standards to those with outstanding mortgages and the requirement for buyers to have local “hukou” residency – as usual. So far, this latest round of property curbs has not yet tamed the burgeoning bubble, and property prices continue surging (Focus Chart 3).

Focus Chart 3: Real interest rate has plunged to new low; property bubble binding monetary hands. 

Property curbs tighten liquidity and will hurt equities – contrary to popular belief: Consensus believes that the curb on property will “force” funds rotating out of the property sector into stocks. We beg to differ. History suggests that property curbs tend to be followed by declining market return, or even outright market downfall, as can be seen after September 2007, January 2010, April 2010, January 2011 and February 2013 after property curbs were initiated (Focus Chart 4). 

The reason for a falling market afterwards is palpable: property transactions tend to be a monetary multiplier, with the newly-created credits associated with home loans. These transactions also accelerate monetary velocity with derivative purchases of furniture and appliances after closing. As such, property curbs will indeed have a liquidity tightening effect – contrary to popular belief. Such decline in liquidity offers a good explanation to the ensuing disappointing market return historically. 

Focus Chart 4: Property curbs tend to initiate a slowdown in sales, housing starts, investment, prices and stock market.

Long Bond Yields Will Rise, Concurrent with Inflation; Bond Bubble Sets to Burst

Regardless, the market is already hiking rates with rising long yields, hurting leveraged bond positions: In our recent special report “A Price Revolution” on November 14, 2016, we summarized our presentation on global asset allocation at the Westlake Hedge Fund Summit, which was held one day before the historic US election. We believe that labor wages growth falling consistently below labor productivity growth for more than three decades, or the exploitation of labor’s surplus value is the fundamental cause of persistently low inflation. And hence the secular bull market in bonds. Recently, labor wages growth has started to outpace labor productivity growth, heralding the return of inflation and the end of the bond bubble (Focus Chart 5 and 6). Since our Westlake presentation on November 7, long yields globally have surged, led by the US 10-year treasury.

Focus Chart 5: US 10-year a history of surplus value exploitation; productivity gain in adequately compensated

Focus Chart 6: “Inflation is always and everywhere a monetary phenomenon”. 

Historically, long yields measured by China’s 10-year treasury, PPI, CPI, commodities, property and money supply have been rising in tandem (Focus Chart 7). Increasing money supply should lead to higher inflation, and thus rising commodity prices and yields. Yet since the start of 2016, inflation, commodity and property have all surged concurrent with China’s narrow money growth, except the 10-year yield. This is the longest stretch of time that the 10-year yield has fallen in its history (Focus Chart 7).

Focus Chart 7: Commodities, property, PPI and money supply surging – 10y has just started. 

That said, the 10-year has failed to fall below its historical lows twice, and has indeed started to rise above its long-term moving average – a positive signal to short bonds from a trading perspective. Historically, whenever the 10-year rose above its long-term moving average, it tended to form an uptrend that would last more than a couple of months (Focus Chart 8). 

Focus Chart 8: China’s 10y has surged through its long-term average, after the longest fall in history. 

As such, even if the PBoC refuses to raise interest rate when confronted by asset bubbles and mounting inflation pressure, market interest rates have been increasing to compensate for what is obviously missing, and should start to affect asset prices. Even though the interest rate spread should begin to widen, the effect of rising long yields will be more pronounced on bond prices, resulting in capital loses for many. It is not yet clear whether these loses could lead to margin calls and forced liquidation of leveraged bond positions built on the losing premise of steady funding costs. Regardless, as tension in the bond market builds, volatility is set to rise and will eventually spill.

Sacrificing the CNY?

If interest rate were to be kept steady, the CNY will have to depreciate to bear the brunt adjustment: Recently, the CNY has been depreciating rapidly towards 7, well pass the range of 6.8-6.83 - where the last phase of CNY appreciation started in 2010. With a half-open capital account, rapidly plunging real rates with rising inflation pressure is a sure recipe for currency depreciation. Should it intensify, it would not be a surprise to see tighter control on cross-border capital flows to slow down capital outflows induced by rapid CNY depreciation. 

Even if the CNY could depreciate in a gradually isolated environment, the offshore CNH market still represents a challenge. The on/off shore RMB markets are influenced by similar but not entirely the same factors. If the pace of CNY depreciation is managed, then the CNH will be under the pressure to reflect the underlying economic fundamentals. And the on/off shore exchange rates will start to diverge further. Of course, the PBoC could intervene with restricted offshore CNH liquidity and hiking short-term RMB borrowing rate in the offshore market. But such acts will then cause the on/off shore interest rates to divert, instead of the exchange rates. It will roil the offshore asset prices denominated in RMB whenever the pace of depreciation between on/off shore exchange rates diverts. Something, somewhere at sometime must bend.

Falling FX reserve portends further depreciation pressure and capital outflow: Globally, the countries that historically have been hoarding the most US dollars are seeing their FX reserve declining fast, except Japan. Saudi Arab’s reserve is hurt by falling oil price. And the petrodollar, a form of liquidity that used to support the US treasury market and provides offshore US dollar supply, is rapidly drying up. Meanwhile, China’s reserve has declined from four trillion US dollars to close to three trillion, together with a depreciating CNY (Focus Chart 9).

Focus Chart 9: Global central banks’ international reserves falling; the CNY depreciating fast. 

The recent CNY depreciation has not caused market panics, contrary to last August and early this year. Consensus believes that the CNY is reflecting the weakening economic fundamentals, and thus other asset prices will not have to adjust - similar to the effect of a weakening JPY on the Japanese economy and market. This view is not entirely correct. The difference is that Japan holds large investment positions overseas, whereas Chinese asset allocation is still largely RMB denominated. A depreciating JPY will make Japan’s overseas investment more valuable, but not necessarily the CNY to China.

FX reserve accumulation has been the most important channel of money creation in China (Please see our report “The Most Crowded Trade” on September 12, 2016). CNY depreciation will not affect stocks only if the PBoC is seen not intervening with precious FX reserve. A discharge of FX reserve will mean tightening macro liquidity, if without other forms of liquidity replenishment by the PBoC, such as the SLF and MLF. The exchange-rate reform is to let the market determine where the CNY should trade, and it appears this objective is being achieved fast. The built-up in speculative long positions in the US dollar also augurs for further US dollar strength, and thus, a weaker CNY (Focus Chart 10).

Focus Chart 10: USD non-commercial net-long surging, portending further dollar strength, and weaker CNY.

Bonds to Underperform Equities; Volatility Set to Rise

The Chinese economy is stuck between reflation and stagflation; L-shaped since 2012: Focus Chart 11 outlines different phases in the economic cycle that China has navigated through since the recovery initiated in 2009. However, since 2012, the Chinese economy appears to be stuck between reflation and stagflation, as highlighted in the red rectangle in Focus Chart 11. The path that the Chinese economy has traversed since 2012 resembles the “L-Shaped” growth phase that has been discussed by the “Authoritative Figure”.

These two economic phases hold different implications for asset allocation. When in reflation, the economy tends to see falling interest rate, rising stock and commodity prices. But when in stagflation, falling FX reserve, tighter money and falling property prices are more commonly observed. Recall in the past four years, China has indeed experienced all of these phenomenons, supporting our model conclusion empirically. Consequently, Chinese equities have been stuck in a wide trading range, with bouts of volatility periodically. 

Focus Chart 11: The Chinese economy L-shaped since 2012, and presents a dilemma to asset allocation. 

Bonds set to underperform equities; look for convexity trades with quasi-option return: Our bond yield vs. earnings yield model (EYBY model hereafter), which has helped us pinpoint the peak of China’s stock market bubble in June 2015 and negotiate the rough waters after the bubble burst, started showing relative value of equities relative to bonds in June this year around the Brexit period (Focus Chart 12). 

That said, EYBY’s relative value saw a significantly higher peak in 2008 that eventually corresponded to the market bottom. The local peaks in 2012, however, did not correspond to the eventual lows of the market. As equities’ relative value improved, the rotation should progress gingerly, in tandem with market volatility.

The pace of how fast equities’ valuation can expand relative to the rise in bonds’ yield determines how far the stock indices can rise, as funds rotate from bonds to equities. As liquidity conditions tighten and bond yield rises, it is likely that the market will be increasingly unwilling to ascribe a higher valuation multiple for each unit of earnings. Or equity valuation may not be able to expand fast enough to offset the rise in bond yield. As such, the outlook for Chinese equities will look much less sanguine than many pundits are forecasting based on the false assumption of expanding liquidity conditions.

Focus Chart 12: Chinese equities showing relative value to bonds; rotation has begun. 

The Shanghai Composite 2017 likely trading range = 2800 – 3800, with widely dispersing outcomes. In our report “Outlook 2016: the Chinese Curse” on December 5, 2015, we applied our EYBY model to estimate the trading range for the Shanghai Composite. This time last year, the relative value between equities and bonds had not yet completed its trend towards historical highs, and was clearly set to do so. It made our job easier then. Our target range for the next twelve months set in last December was 2500-3300. As of writing, the Shanghai Composite has largely confined itself within the trading range of 2600-3300, after a dramatic 1000-point plunge and circuit breaker meltdown that many failed to anticipate earlier this year. And it has been exactly 12 months.

In June, around the time of “Brexit”, our EYBY model has moved past its high point in recent years, and started to fall. We believe its decline should continue, before settling in a range, as the history in 2010 or in 2014 suggests (Focus Chart 12, highlighted in red rectangles). As the relative value trend is entering into a range-bound phase without apparent direction, the exact upper and lower bounds of this year’s range are more difficult to discern. As such, we have applied a sensitivity analysis based on different levels of bond yield and various levels of relative valuation between bonds and equities to arrive at a likely trading range (Table 1).

The historical average of relative valuation between bonds and equities was around 0.4. We believe the current level should continue to fall towards its long-term historical average, before fluctuating within the range of +1/-1. Such moves should see the Shanghai Composite trading at 2600-4000 – a very wide trading range. But as bond yield is likely to rise above 3%, relative valuation should linger at the range below its long-term average, or from 0.4 to -1, implying a trading range of 2800 – 3800. 

Table 1: Shanghai Composite 2017 trading range = 2800-3800 (scenarios with higher confidence highlighted in blue). 

Note that when relative valuation is at its long-term average of 0.4, the implied trading range from our analysis in Table 1 is indeed below the index’s current level of around 3300. Within the likely trading range for 2017 we highlighted in blue, only 1/3 is above 3300, with a wide dispersion of potential outcomes. That said, our estimated trading range of 2800-3800 for 2017 is higher than the range of 2500-3300 estimated exactly twelve months ago. Further, assuming rising HKD relative to CNY, and the connect program goes smoothly, southbound funds should offset outflows from Hong Kong due to a strong Dollar (Focus Chart 13). The A shares are moving closer to the oversold territory, but the post-bubble correction has not gone deep enough compared with historical precedents (Focus Chart 14).

Focus Chart 13: Hong Kong has shown allocation value since February 2015. 

Focus Chart 14: Shanghai’s post-bubble correction appears too shallow compared with historical levels. 

Hao Hong, CFA

2016-12-4