China watchers have long debated the reliability of the country’s GDP numbers, and for many years the pessimists have argued that official figures are overstated. This is not a trivial debate anymore because China has become a major driver of global growth over the past decade, on par with the United States. Recent signs of slowing growth in China, blamed on trade wars and declining consumer confidence, have only heightened the debate.
The first thing to understand about China’s GDP is that the concept of GDP targeting in China is very different from what investors are familiar with, and this leads to confusion. One insightful China watcher, Michael Pettis, who is professor at Peking University’s Guanhua School of management, makes this argument in a recent article (What is GDP in China?). Pettis reminds us that the Chinese, with their deep tradition of economic planning, think of GDP as a pre-determined input figure not as a variable output. It is not a coincidence, therefore, that magically year after year the Chinese meet their GDP growth goal. Part of the reason for this may be some window-dressing for political reasons but much of it comes from active intervention. For instance, if the economy appears to be running below expectation, the authorities will respond quickly with increased spending and lending to set it back on target.
Unlike the U.S. Fed, which has only the blunt tool of monetary policy to achieve its pretentions of smoothing out the economic cycle (eg. Ben Bernanke’s “Great Moderation”), the Chinese authorities have an expanded toolkit of monetary policy, bank lending and fiscal spending which they have immediate access to.
Of course, achieving such fine tuning is easier said than done. The pessimists on China will argue that the Chinese authorities have exhausted the utility of these tools. This may be because of a combination of excess debt and declining returns on fixed capital investments in real estate and infrastructure, and also because saturated foreign markets have become a much more volatile driver of growth. This state of affairs increasingly has raised the issue of the “quality” of China’s GDP growth. If the GDP numbers are being achieved by increasing debt that won’t be repaid – either for domestic investments or for ports in Pakistan – and the end-result is more empty real estate and under-used bullet trains, than the effort is counterproductive. The Chinese have long been aware of the unbalanced nature and the limits of their debt-driven fixed-asset investment model, but it is not easy to change behavior. China’s vice president, Wang Qishan, reiterated the government commitment to its GDP growth target this week in Davos. Wang pledged, “There will be a lot of uncertainties in 2019, but something that is certain is that China’s economy, China’s growth, will continue and will be sustainable.” In other words, the authorities commit a-la-Draghi to do “whatever it takes” to meet the 6.5% annual growth target.
Nevertheless, foreign observers are always skeptical of China’s growth figures and seek alternative yardsticks to corroborate the official data. For example, electricity consumption is looked at in comparison to GDP growth. The recent numbers forelectricity consumption, shown below, at least have the merit of displaying year-to-year variability.
Along this line, Barclay’s bank looks at a series of alternative indicators to provide a comparison to official figures. Based on this exercise, Barclay estimates that China’s economy has been performing well below targets for the past five years.
A new paper by Yingya Hu and Jiaxiong Yao of John Hopkins University (“Illuminating China’s GDP Growth) uses a very innovative methodology and arrives at the same conclusion. Hu and Yao analyse satellite-reported nighttime light over time to measure changes in economic activity. As shown below, they estimate that China’s GDP may be some 20% overstated. The authors have done this for a wide variety of markets and find the data in China to be one of the most overstated. As shown below, India is also slightly overstated while both Brazil and South Africa are actually understated.
It is not clear what is causing this discrepancy in China. One theory is that a significant part of the real-estate stock remains dark, as properties are being bought for investment purposes and not occupied.
In any case, these theories of overstated GDP growth raise several worrisome questions. First, this may be evidence that the authorities may be pursuing unproductive policies as marginal returns from debt accumulation and fixed asset investments have declined. Second, the country’s very high credit/GDP ratio of 300% may be significantly understated, and could be closer to 360%.
Trade Wars
- The U.S. has turned sour on China (FT)
- Germany’s China problem (NYT)
- You can’t contain China (SCMP Rpbert Zoeliick)
- Jimmy Carter on China (Washington Post)
- On China relations, 40 years on (China File)
- Seven issues will drive the trade talks (Caixing)
- King dollar? (Kupy)
India Watch
- Amazon adapts to India (WSJ)
- India’s love of mobile video (WSJ)
- India’s potential in passive investing (S&P)
- India’s food-delivery startup, Swiggy, backed by Tencent (SCMP)
- Modi’s election troubles (WSJ)
China Watch:
- What next for China’s development model (Project Syndicate)
- An entrepreneur’s tale of adaptation (NYT)
- China boosts new airport spending (Caixing)
- An analysis of nightlight points to overstated Chinese GDP (JHU)
- China’s slowdown (CFE)
- China’s GDP (Carnegie Pettis)
- Zero growth in car sales expected for 2019 (Caixing)
- Looking back on 40 years (Ray Dalio)
- China steps up bullet train spending (scmp
- On sector investing in China (Globalx)
- The Future Might Not Belong to China (FT)
- Will China reject capitalism (SCMP
- The rise of China’s steel industry (WSJ)
China Technology Watch
- Can China become a scientific superpower? (The Economist)
- The quantum arms race (tech review)
- Baidu CEO says winter is coming (SCMP
- Zoomlion’s international ambitions (FT)
- Ping An goes digital (Mckinsey)
- China’s bet on AI chips (Tech Review)
- China’s 5G push (tech review)
Brazil Watch
- John Bolton’s Troika of Tyranny (The Hill)
- The rise of evangelicals in Latin America (AQ)
EM Investor Watch
- Indonesia’s economic populism (The Economist)
- EM’s Corporate debt bomb (FT)
- Holding’s digital transformation (Mckinsey)
- Lowy Institute Asia Power Index (Lowy
- In pursuit of prosperity (Mckinsey)
- What drives the Russian state? (Carnegie)
- Russia’s big infrastructure bet (WSJ)
Tech Watch
- Solving the productivity puzzle (Mckinsey)
- Autonomous trucks are coming fast (Mckinsey)
- Bloomberg energy finance, 2018 report (Climatescope)
- Fast-tracking zero-carbon growth (Ambition loop)
Investing
- 50 reasons not to invest for the long-term (Abnormal returns
- Not one Ivy League endowment outperformed past 10 years (Institutional Investor)
- Stock market charts you never see (Papers)
- The collision of demographics (Bain)
- The top 100 asset managers in the world (Thinking Ahead)
- An evolve or die moment for the world’s great investors (Fortune)
- Interview with William Eckhardt (Turtle Trader)
- Why momentum investing works (Anderson)
We can look at relative valuations to explain this relative performance. The chart below looks at Cyclicaly Adjusted Price-earnings (CAPE) ratios for both EM and the S&P 500. (The CAPE takes an average of ten-year inflation-adjusted earnings to smooth out cyclicality). This poor EM performance occurred because at the beginning of the period valuations in EM were relatively high and U.S. valuations were relatively low. 
The chart shows that a year-end 2008 EM was trading at a CAPE multiple 0f 14.7, in line with its 15-year average. Meanwhile, the U.S. market was valued at a CAPE mulitple of 15x, compared to its 15-year average of 26x. By the end of the 10-year period, EM CAPE multiples had declined and were well below historical averages while U.S. multiples were well above the historical average. This largely explains the relatively strong returns of the S&P 500 for the 2008-2018 decade. As a reminder, the previous decade 1998-2008 had been an entirely different story, with EM vastly outperforming the stagnant U.S. market. During the 1998-2008 period, EM CAPE multiples expanded and U.S. mulitples contracted. In fact, if we look at the past twenty years EM stock market performance is far ahead , providing returns of 444% vs. 204% for the S&P 500, as shown in the gaph below. 
So, what can current valuations tell us about probable future returns? In short, the prospects look good for EM. EM is now relatively cheap, trading at a CAPE valuation of 11.7 vs. a 15-year average of 16. Meanwhile, U.S. stocks trade at a CAPE of 28.4 times vs. an average of 24.7x. The U.S. dollar has also been strengthening for years against EM currencies, a trend that is likely to revert in the future. Though, in the words of baseball legend Yogi Berra, “it is difficult to make predictions, especially about the future”, we can use the historical context to make some guesses about probable future returns. We make three assumptions: 1. Valuations will move back to the historical CAPE average over the next five years; 2. Earnings return to historical trend; and 3. Normalized earnings grow by nominal GDP. To determine the historical earning trend we take a view of where we are in the business-earnings cycle. In the case of EM we consider that at this time earning are about 10% below trend and we assume 6.5% nonimal GDP Growth (vs. 4% for developed markets). Based on this simple framework and assumptions we project annualized returns for EM stocks for the next five years of 9.8% (9.3% for ten years). Adding dividends, we project total annualized returns of 12.1% (10.5% for ten years.) The U.S. ,by contrast, is likely to experience multiple contraction and is at peak earnings, so that returns can be expected to be low single-digits. Two outside opinions shown below arrive at similar results: first GMO (on the left) projects 4.4% real annualized returns for EM (7.9% for EM value) for its seven-year forecast period; Research Affiliates (
On a country-by-country basis, as one would expect, great differences appear. Countries find themselves at different points in the business-earnings cycle and their valuations may vary greatly depending on the mood and perceptions of investors. The chart below shows where country-specific valuations stand relative to the 15-year CAPE average for the primary EM markets. The third column shows the difference between the current CAPE and the historical average. For example, Turkey’s valuation, in accordance with CAPE, is 60% below normal. The markets in the chart are ranked in terms of probable long-term returns (5-10 years), with the last two columns to the right estimating annualized total returns (including dividends) for the next five and ten years. The table also shows where markets are currently in their business-earnings cycle and expected annualized earnings growth for the next five years. 
What does this table tell us? First, we can see that valuations are generally low. The majority of markets in EM trade at very deep discounts. India, Peru and Thailand, the most expensive markets, are valued only slightly above historical valuations and are not abnormally high in absolute terms. Second, most markets stand in the early-to-mid part of the earnings cycle. This provides the opportunity for concurrent earnings growth and multiple expansion for Brazil, China, Chile, Mexico, Malaysia, Colombia and Turkey. What does the table not tell us? First, this methodology serves best as a long-term allocation tool, not as a timing tool. Market timing is difficult because short-term stock movements are determined much more by liquidity considerations and the mood of investors than by valuation. So, for example, timing a stock market recovery in Turkey is not easy. The market may fall much further before it starts a recovery. Eventually, a new more constructive narrative will gain traction in Turkey and catch the attention of investors, starting a new cycle. Second, the assumptions of the model may be wrong. 
