What happens when you apply big data to investment?

Hu Pingping

What will be the result of applying big Internet data to investment?

GF Fund's Guangdong China card Baidu Hundred Strategy 100 Index fund was approved, to be issued on October 20, this will be the first to use large data index funds. Previously, by Sina Finance, the South Fund and the Shenzhen Information company to launch the Big Data Strategy index i100 and i300 had been listed on September 12 in Shenzhen stock.

After implanting the internet's large data gene, these indices quantify and express the previously elusive market sentiment, and how much can they add to the investment? What are the secrets behind the big data?

Excess income comes from three major data "chemical reactions"

Friends at the party, a stock market talent casually recommended a few stocks, usually also buy some stock of you, afterwards remembered unavoidably first open Baidu search, will these several stocks research.

But at this point, do you think that this action is now like the butterfly effect, can affect an index, a fund of the ups and downs?

"When investors face market mass information, their concern is a scarce resource." "Jifeng 100 research and development staff," the first financial daily newspaper "financial quotient" (Micro-signal: CAISHANG02) Reporter said, 100 of the Internet Big Data application is actually based on the financial "limited attention" theory.

In particular, only the limited attention of investors, there will be a corresponding network search behavior, this large data more directly and accurately reflect the investors to a specific stock of concern.

However, it is hard to get excess profits by relying on large data on the internet alone. According to Jifeng, before the Hundred 100, the team tried to use large data to do a similar index products, but this index performance is not good, historical simulation returns and the Shanghai and Shenzhen 300 Index of the income is basically flat, meaning little.

"To really get excess income, you need to add two other types of quantitative data that are commonly used, and we're surprised by the chemistry of the three." "Jifeng said, in addition to using the user search and behavioral data index provided by Baidu, the model integrates the financial factors, momentum factors and momentum reversal strategies of 8 financial indicators into the BFS model, and filters out 100 basic fundamentals and good growth space, Can reflect the next one months in the market industry rotation hot stock.

"In short, it combines the first half of the investor's decision behavior with the data in the second half." "GF Fund Quantity Department general manager, hundred Hair 100 proposed fund manager Lu Zhiming to reporter said."

What is the result of the "chemical reaction" of these three data?

From historical data, the 100 index has performed well since its release. Cumulative earnings, hundred 100 since 2009 has achieved 545% (as of June 30, 2014), far higher than the 300 index 19% of the same period of earnings, higher than the Shanghai index of 12%, also far more than the overall market performance benchmark in the total 56% of the cumulative income.

In addition, as of September 24, the Hundred 100 Index has risen by 47.42% this year.

Also with big data "Golden Key" came from the Big Data Strategy index i100 and i300, the highlight of the index is the South Fund and Sina Financial cooperation, the use of Sina finance in large data advantages, analysis of the current market focus on investors and focus, and then combined with financial factors and market driving factors, The selected component stocks are optimized by a quantitative model.

From the historical data, the two index also significantly exceeded the market, January 29, 2010 to September 12, 2014, i300 rose more than 180%,i100 rose more than 290%, the same period, the Shanghai and Shenzhen 300 index fell nearly 24%.

In addition, i100 has risen 35.56% this year.

Large retail market data more effective?

"Large Data index is a new thing, from home and abroad, there is no successful precedent." "Good buy fund" one analyst told reporters that the big data strategy to continue to maintain the previous historical performance of long-term effective, still need future performance to prove.

It is reported that there is no typical case for the actual application of search data in the capital market, but the research on social and economic behavior prediction based on Internet search data has gradually become a new academic hotspot. In the study of capital market application, there are academic studies abroad that the company's name or related keywords search volume, and the company's stock volume is positively correlated.

"Now the domestic public offering fund in some respects is no more backward than abroad." Lu Zhiming said that domestic data could be more effective than overseas, while applying the big Internet data to investment.

In fact, whether Baidu or Sina financial Internet financial data, in a sense is a measure of investor sentiment, one of the quantitative methods can be used to reflect the investment will or expectations of investors. The institutions that develop such indices believe that investor sentiment plays an important role in the pricing of a-share securities.

Lu Zhiming explained that, because a a-share is still a retail-oriented market, individual investors more, and individual investors in making investment decisions will be more use of search engines such tools to find information, so the Internet investors behavior data more reflect the mood of most investors, This is also Baidu 100 since the release can obtain the reason of excess income.

"Big data is really a big gold mine, and now we are using only a small part of the future, including behavioral data, good and bad screening and so we all hope to be able to apply to the investment." "Jifeng said.

What are the risks of a large data index?

Hundred 100 will make a transfer every month. According to the introduction, hundred 100 from June 20 since the 3 period of operation, basically each period of 50% of the stock will be replaced, a high proportion of the warehouse plus the current high income, hundred 100 will become a chase up the index fund?

Jifeng to this explanation, to June 20 This issue hundred 100 of the constituent stocks for example, gem stock accounted for 5%, SME board stock accounted for 20%~30%, Shanghai and Shenzhen board stock accounted for more than 60%.

"From this composition, hundred 100 is not a small-ticket style index." "Jifeng said, how to avoid chasing up and down, in the model has been taken into account, through the combination of three major data, some of the market has been hyped, the fundamental risk of stocks have been" filtered. "

"We hope that through these three data and models, to find some in the future there will be a certain amount of room for the next one months, market or industry hot stocks, rather than chasing the rise and fall of the short-term stocks." "Lu Zhiming said.

In addition, some analysts also point out that in the early days of Big Data index funds, investors will be more likely to gain excess income. However, the increase in the number of products that will mimic this strategy in the future could also lead to a fall in excess earnings, similar to the risk of quantitative funds being homogeneous in the market.

However, Lu Zhiming said that once the fund was established, once the performance continued to go well, because domestic investors are relatively short-term, I am afraid will encounter a lot of redemption, this is the most worried.

"The target of the 100 index will be able to hold a large amount of redemption in one day, which is the key consideration in our operation." "Lu Zhiming said.