Narratives, the stories we tell about markets, have a strong influence on market dynamics. And a study of Chinese stocks shows that these narratives may even create their own momentum effect.
As part of my job at Liberum, I have recently launched a weekly publication that investigates market narratives using a large number of economic and investment terms. As Robert Shiller has pointed out in his Presidential Address to the American Economic Association in 2017, which later turned into a book these narratives may be much more important than we thought in the past. But it’s only through recent technologies like AI and natural language analysis that we can systematically exploit these narratives.
A team of researchers from China has now looked into the concept of “concept stocks” to see how market narratives drive the return of different types of shares. What they mean with concept stocks is a classification of stocks not along the usual industry classification schemes, but along the nature of their business. For example, some stocks in the technology sector may be associated with AI, others with cybersecurity, etc. They are all technology stocks, but part of different niches in the tech universe. Sometimes, industrial stocks may be exposed to concepts in other sectors. For example, a company that builds robots is linked to the concept of AI.
They could do that because Chinese trading apps assign stocks to specific concepts to help retail investors grasp what a company does. These concepts are created and adjusted very quickly. For example, the concept of “mask” was created on 4 April 2020, just a couple of months after the outbreak of the COVID-19 pandemic.
Using these concepts from Chinese trading apps the researchers could group stocks into different concepts or narrative baskets and then look for trends in these narratives. What they found is that there is a momentum effect that is uncorrelated to the traditional price or earnings momentum effect and based on market narratives. Stocks that are part of a specific narrative gain when the narrative becomes more important. When the narrative loses importance, their share prices suffer. This allows investors to create narrative baskets of investing in stocks linked to narratives that gain popularity and selling stocks that are linked to narratives losing popularity.
The chart below shows the result of a strategy investing in stocks associated with narratives of rising popularity vs. stocks associated with narratives of falling popularity. The outperformance is a stunning 15% per year. While one can argue a lot about using the Chinese stock market as an example since the dynamics in that market are significantly different than in developed equity markets, the size of the outperformance is so large that it seems unlikely to be spurious. And indeed it seems likely to me that there is outperformance to be had in developed markets by investing in this kind of narrative momentum.
Performance of narrative momentum strategy in China
Source: Du et al. (2022)
Thanks for the article, that looks quite interesting! May I know the name of the paper?
Are there any ETFs or funds in the US or European markets that are using this kind of narrative grouping? Are thematic ETFs achieving something similar?