Initial Hypothesis and Thesis

Initial Hypothesis and Thesis:

        We first stumbled upon SFUN as we were searching for good, under-followed companies.  Initially we thought:

• Because of its leading market shares, we believe that customers prefer using SFUN as opposed to its other competitors in China, like WUBA or LEJU
• We believe that the Chinese economy and housing market will rebound sharply, or at the very least will not affect SFUN greatly as it acts more as a high margin tollbooth on transactions than a player beholden to the movements of the housing market
• SFUN currently has really high ROTCE and growth rates, but is valued very cheaply, so this could be potentially a great buy

However, as we furthered our research, we found that:

• SFUN is not the preferred search engine, both WUBA and Ganji are considered more popular
• The company’s earnings are set to flounder if a housing recession occurs in China. The company is already showing signs of slowing growth
• The company looks like an excellent stock in its financial statements, however we are worried about the integrity of management and the possibility of accounting problems

Our recommendation is to HOLD

Background:

SouFun Holdings Ltd (SFUN) is a Chinese Internet real estate portal like Zillow (Z) or Trulia (TRLA). The company allows Chinese citizens to find houses, condominiums, and apartments. The company generates revenue from four different sources: in essence most of its revenue comes from one source of advertising or another. SFUN’s first source of revenue is from marketing. Marketing refers to the banner advertisements and videos on the company’s website. E-commerce, the second source, is a bit of a misnomer. While the company does offer a small eBay-like business for home furnishings of which the company takes a cut of the transactions, a majority of the money in the e-commerce segment is made through selling memberships, which allow users to buy real estate, particularly condominium developments, at cheaper prices. This is no different than condominium development advertising in the same discount method like Groupon (GRPN). Marketing and e-Commerce are both currently at around 40% of revenue. The company’s third source of revenue is through listing services, the traditional advertisement method for companies like SFUN. This allows sellers, particularly real estate agents, to pay for premium real estate listing positions. This has been traditionally about 30% of revenue, but has taken a plunge recently perhaps due to the housing crisis. The fourth source of revenue, which is the smallest at about 3% of revenue, is other revenue, which includes using SFUN’s proprietary data and its real estate financing business. Currently, investors are anxious in SFUN’s performance because of the potential housing crisis in China. We think this may damage SFUN temporarily, but it is an excellent business that will be a lot larger than its current size a few years down the road. The company is currently trading at 13 P/E and has been growing 20% - 30% a year before the housing crisis and should continue to grow at the same pace after.  

Derived Revenue – Regression Results:

        To be able to forecast the future revenues of SFUN, we must first find out the derived demand of the housing market. Based on our research, we have determined that there are four main factors that drive housing demand: GDP per capita growth rate, population growth rate, changes in unemployment rate, and changes in mortgage/interest rates. We then used annual SFUN revenue growth rate as the Y variable, and conducted regressions using the other variables as the X variables. Our results were mixed. When we attempted to use all four X variables in one regression, all of the coefficients were insignificant. Therefore, we decided to have single variable regressions instead. On one hand, population growth rate and changes in unemployment rate seem to not really affect the revenue growth rate of SFUN at all. However, the regression of SFUN revenue growth vs. GDP per capita growth had high R-squared, and the coefficients are also significant (Exhibit 1). Curiously, the regression of SFUN revenue growth vs. changes in interest rates had an even higher R-squared, and the results seemed even better (Exhibit 2). However, this regression actually doesn’t make any sense, as we would expect there to be a negative correlation between the two variables. People are more like to buy houses when interest rates go down, not when interest rates go up. Because of this, we decided to go forward with the regression containing GDP per capita growth. Using this regression and forecasted Chinese GDP per capita growth of 2015 and 2016, we forecasted a 2015 revenue growth rate of 19.74%, with revenue of $841 MM, and a 2016 revenue growth rate of 12.40%, with revenue of $946 MM. These forecasts are similar to the ones forecasted by Bloomberg and Capital IQ. These results demonstrated that the housing market in China, and consequently the growth of SFUN, is strongly correlated to the overall economic performance of China.

Good Business Framework: