Tiger Woman on Wall Stree (23 page)

  *  *  *  

Another American titan of investment who fell victim to the top-down investment approach in China was Warren Buffett. He famously said, “The 19th Century belonged to England, the 20th Century belonged to the U.S., and the 21st Century belongs to China. Invest accordingly.” With that, and a stock recommendation from Li Lu—a Chinese investor who fled to the United States after playing a prominent role in the protests at Tiananmen Square—Buffett began investing in China.

In September 2008, just after the collapse of Lehman Brothers pushed the financial crisis to new heights, Buffett’s Berkshire Hathaway bought nearly 10 percent of a Chinese battery and electric car-maker company named BYD for HK$1.8 billion ($232 million), or
HK$8 per share
. The company’s initials stood for “Build Your Dreams,” and it was indeed a dream, both then and for years after Buffett invested. Buffett’s reputation as the savviest investor of the twentieth century caused the stock to double on the day of the announcement to HK$16 per share, what was termed as a nice “Buffett pop.” The Buffett premium helped to push the stock to a peak of HK$88 in 2010 before it finally started to fizzle.

From a top-down perspective, this investment made a lot of sense. China was the world’s biggest auto market, and it would only get bigger as urbanization continued and the disposable income of the middle class rose. All these cars would exacerbate China’s terrible pollution problem, leaving the government desperate to cultivate a market for small electric cars. In 2009, Beijing reduced the sales tax on vehicles with engines under 1.6 liters to 5 percent from 10 percent, abolished road maintenance fees, and subsidized clean energy cars in a bid to prop up the economy and limit pollution. BYD was a major car maker with high historical margins and a promising battery division, well positioned to lead China’s electric vehicle (EV) market. Together, all these facts constituted a compelling investment thesis.

A bottom-up analysis revealed many flaws in BYD’s business model. The company’s past success was largely a result of free riding off the R&D of Japanese competitors. The Chinese government had also issued a subsidy for small vehicles in 2009, which created a bubble in car purchases. Many consumers moved their purchases ahead to take advantage of the subsidy, resulting in a drop in demand when subsidies were reduced a year later. Warren Buffett amassed his wealth through a highly selective buy-and-hold strategy, so when he bought into a Chinese company whose
valuation was largely driven by electric vehicles, a product yet to be commercially proved, I was surprised.

The stock’s value fell back to $8 by September 2011. Then BYD announced that corporate profit had fallen 94 percent to $13 million in 2012 due to weak
car demand
. BYD executives had reassured investors that the company could rely on its other business lines while it was waiting for China to build out charging stations for electric vehicles and for the still-tiny electric car market to grow, but that proved untrue.

In early 2013, the company was planning to raise $500 million in a rights issue to bolster its balance sheet and buy itself more time, putting further pressure on the stock. Four-and-a-half years after Buffett invested in BYD, the company’s EV division—the part of the business that drove the stock’s hype and investor hopes—was still a cash burner, and the stock had sunk to HK$22 from a 2009 peak of HK$85. Whoever bought into the Buffett bubble was likely sitting on a loss. That is, except for Buffett himself, who I believe had likely locked down and protected his profit from the trade by using sophisticated hedging strategies that are mostly beyond the reach of retail investors. The takeaway for investors? Don’t join the Buffett party unless you are an invited guest.

China’s economy has evolved quickly in the decades since Reform and Opening Up. The country has carried out its own Industrial Revolution in only a bit more than 30 years, a process that began in the United Kingdom in the mid-eighteenth century. The scale is also entirely different: The United Kingdom completed its Industrial Revolution with only around 16 million to 20 million people and the United States with maybe 50 million. China is doing it with more than 1 billion. This unprecedented speed and scale of change easily overwhelms investors in the public and private markets
who are not equipped with the latest and most accurate data, information, and analysis. Therefore, for most people, China is not yet a market for long-term investments, but rather short-term trades.

Investing and trading in the public market both involve purchasing a security, but they are drastically different games. Investing refers to buying and holding a stock for a long period of time despite short-term share price volatility. Trading, on the other hand, is buying and selling a security relatively frequently to profit from the fluctuations in the share price.

Those in it for the long haul must have a conviction, typically based on thorough fundamental research on a company. Long-term investors then develop their own views on the intrinsic value (the actual, not the market, value) of the underlying businesses. Only with that conviction can an investor comfortably ignore the market noise and short-term fluctuations in the market value reflected in the share price.

Trading is a different game. An investor trades on a stock when he or she thinks the market value of the stock will increase or decrease. Changes in a stock’s market value are typically driven by expectations of a company’s earnings—profits distributable to shareholders—quarter by quarter. If an investor thinks the company will deliver better than the Street consensus—the median estimate of prominent sell-side analysts—he or she will “buy into the quarter,” which stands for the quarterly earnings announcement in which all U.S.-listed companies are required to update investors of their operating performance. If the investor is pessimistic about a company’s ability to meet the Street consensus, he or she may take a short position into the quarter. Either way, a trade is typically made only when the investor disagrees with the Street view on a company’s near-term earnings.

For long-term investors, properly assessing the value of a business in China is a major challenge. Most value investors depend on what’s called a “mid-cycle analysis” to assess the normalized
earnings power before ascribing a value to a business. Normalized earnings or mid-cycle earnings are earnings adjusted for cyclical variations. To get that estimate, analysts look at the successive peaks and troughs in a company’s earnings and adjust them to a moving average.

But for both Chinese companies and China’s economy, mid-cycle references do not exist. Since the introduction of the market reform in 1979, the Chinese economy has only gone up, never down. Whenever the economy showed signs of slowing down, the government stepped in with fiscal stimulus and expansionary monetary policy. In other words, Beijing has so far defied the natural gravity of the economic cycle, with potential long-term structural damage.

In private, many economists argue that it is statistically improbable for any economy to have produced an economic growth trajectory as smooth as China has since 1979. In the 1980s, China’s economy was still overwhelmingly agricultural, so it should have been subjected to Mother Nature’s unpredictability in the form of bad harvests or bumper crops. In the 1990s, as manufacturing and industrial production grew as a percentage of GDP, business cycles driven by external demand and productivity fluctuations should have generated far more significant swings in economic growth than what reported Chinese economic data indicated.

Moreover, the structure of both the Chinese and global economy has evolved rapidly and unpredictably. China opened itself up to the global economy by engaging in international trade and accepting foreign direct investments, and therefore became more vulnerable to external economic shocks. Yet during this same period, Chinese official statistics show aggregate GDP advancing quickly and steadily, like a luxury car down an empty highway.

In the absence of a full cycle of growth, all projections of the intrinsic value of Chinese businesses are largely guesses. The tools commonly used on Wall Street to assess the intrinsic value
of companies, mid-cycle and discounted cash flow analysis, are therefore not exactly applicable. This is another level of complexity to consider when investing in Chinese companies or the Chinese economy.

B
Y NOW, ALMOST EVERYONE KNOWS THAT REPORTED DATA COM
ing out of China are not to be trusted, for various reasons. Investing solely based on official figures—whether put out by the government or companies—is a sure way to get burned.

For a long while, I had been strategizing how to develop a proprietary system to acquire and collect more accurate data and information for investment purposes. I had met with a few data suppliers, but none quite fit the bill. Many of them were founded and run by Westerners, mostly former American journalists. They were typically small outfits whose primary research methodology was journalistic investigation—sneaking into warehouses to count inventory, interviewing the suppliers and customers, and visiting department stores to assess foot traffic. Some of their work was good, but most was mediocre, superfluous, or even controversial—one shop in particular was rumored to have accepted payments from companies in exchange for favorable write-ups.

Then, in the summer of 2012, a West coast–based hedge fund manager and a friend from Middlebury College called me
unexpectedly to make an introduction. “You have to meet my friend Yifeng,” he said. “He runs this data processing outfit with a group of computer geeks in China that could complement your business very well.”

I said yes almost before he had finished his sentence.

I had realized from the very beginning that I wanted to partner with a native Chinese who operated on the ground in China, with a cutting-edge data acquisition and processing technology as well as a deep understanding of the country. I was convinced that data-driven fundamental equity research would be the future Ferrari of the research world. This is what I needed to take my business to the next level.

The kinds of journalistic investigations that most research outfits undertake in China are conducted by human beings and therefore subject to human bias and errors, whether conscious or not. But unbiased raw data generated directly from Chinese search engines, e-commerce sites, and social media—along with human fine-tuning to ensure relevance and accuracy—offer a far superior method of conducting equity research. Ultimately, that type of research can benefit not only stock investors but also corporations, governments, individuals, and research institutions that wish to gain insights into any particular company, sector, or subject.

  *  *  *  

One week later on a Friday afternoon, Yifeng flew in to New York from Shanghai for our first meeting. We met for coffee at the Tribeca Grill, Robert De Niro’s downtown establishment. A Shanghai native in his thirties, Yifeng was casually dressed, with the typical low-key persona of a tech guy. He told me he had earned a computer science degree from Shanghai Jiao Tong University, the leading science and technology university in China. After graduating, he worked as an auditor at a Big Four firm, where he audited
a few Chinese ADR companies, and then was recruited to open an Asia office for one of New York’s largest hedge funds. The hedge fund went under during the financial crisis, just as mine had. Like me, he had thought about launching his own fund but decided a data services start-up was a more interesting and lucrative business model because of the underserviced market.

Thirty minutes into the conversation, I started to realize that not only did we share a similar background but the potential synergy between our businesses was too great to ignore. His company, Goldpebble, specialized in the kind of real-time processing of massive amounts of raw data that I was looking for.

Yifeng explained his unique methodology to me. As Internet use grew in leaps and bounds in China, more real-time data had become available online. Most analysts were still reviewing huge amounts of information one piece at a time, such as news stories, postings on social media sites, or reviews of products on e-commerce sites. The process was labor intensive and slow, potentially subject to significant delays if the information had to be translated from Chinese to English. But with its robust and sophisticated IT infrastructure, Goldpebble was able to tap into a massive amount of real-time data simultaneously, analyze the data, and spit out impressive patterns and insights.

For example, his company was way ahead of the pack on Sina, a U.S.-listed Chinese Internet company with a $4.5 billion market cap as of
mid-2013
. Sina gained press and investors because of its popular microblog Weibo, a social media service that (as mentioned earlier) incorporated elements of Facebook and Twitter and revolutionized the way information spread in China. Like social media companies around the world, Sina was still unsure how best to monetize its more than
500 million registered users
. Even so, investors were typically kind to social media stocks, rewarding them a valuation based on their number of users rather than on the profits they currently earned from them.

Sina Weibo’s 500 million followers fetched the stock a high valuation. Goldpebble’s research, however, showed that many of these users were unlikely to ever generate value for the company, Yifeng told me. He and his colleagues developed an algorithm that tracked the number of users who logged in on a daily and weekly basis. They found out that among the 500 million registered accounts, nearly 80 percent were “zombie accounts,” or commercial accounts set up primarily for spamming or third-party marketing. Only about 14 percent of the total registered accounts created at least one new post each month, and only 4 percent were doing so on a daily basis. The program also gathered much intelligence on how people used the service—such as how long they stayed online and what kind of device and operating system they logged in from.