Shen Ruoyu of China Universal Asset Management talks about Bayesian thinking in investment

Mondo Finance Updated on 2024-01-29

What is the biggest challenge in your life", in an interview, Tesla's founder Elon Mask pondered for nearly a minute in the face of this question, and gave an impressive answer: "Make sure to have a feedback mechanism that is constantly correcting and constantly evolving your thinking framework".

In fact, this kind of thinking is very similar to Bayes' theorem, one of the theoretical foundations of artificial intelligence technology, and it also has a very important reference for our investment practice in the current market environment.

Bayes' theorem in mathematics can be intuitively understood as "past experience" plus "new evidence" to obtain "modified judgment". It provides an objective method for inferring by combining newly observed evidence with existing experience.

Suppose there are random events A and B, and their conditional probability relationship can be expressed by the following mathematical formula:

p(a|b) = [p(b|a) *p(a)] / p(b)

Among them, event a is the target event to be investigated, and p(a) is the initial probability of event a, which is called a priori probability, which is often based on the probability data obtained from previous data analysis or statistics.

B is a new event that affects Event A. p(b) represents the probability of event b occurring.

p(b|a) denotes the probability of b when a occurs, and it is a conditional probability.

p(a|b) denotes the probability of a when b occurs (which is also a conditional probability), which is the posterior probability that we want to calculate, which refers to the probability that an event will occur after some observational information is obtained.

According to the Bayesian formula, the posterior probability can be modified and obtained on the basis of the prior probability, which is very similar to the process of improving the accuracy of the company and constantly approaching the true fundamentals of the company through research in our daily investment work.

There are a few key points in Bayesian thinking that are worth learning from when investing:

First, when the prior probability is strong enough, the fact that the prior probability can have a surprising influence on the posterior probability even when new information emerges, is often overlooked. The lesson is that we can't just focus on the latest available information, otherwise the performance of the investment side will follow the market. Maintaining a big-picture view is very important for investment, and it is important to fully consider the important premise of prior probability. For example, a priori assumptions in investment include "ROE determines the upper limit of the company's endogenous growth and also determines the return on investors' long-term holdings", "the competitive landscape affects the average earnings center of the industry", etc. The slow variables that affect enterprise value in these assumptions deserve long-term and repeated in-depth study. In recent years, due to non-fundamental factors such as transaction congestion and capital flow, the market has been keen to chase low-attention, apparently high-growth investment opportunities, but often ignores whether the key assumptions of the industry pattern and business model have changed, and whether there is any impact on the growth sustainability of related companies. In the end, most of the short-term investment behaviors have become part of the market volatility, or ignore long-term changes and blindly hold resulting in passive stop loss, and it is difficult to achieve real returns in the end. Realizing the importance of prior probability can help us to think about the key assumptions of investment logic repeatedly, rationally screen the impact of marginal information on enterprise value from the long-term dimension, avoid making decisions from a single perspective, and truly think about the value of the enterprise from the perspective of aggregate thinking.

Second, establish a correct and identifiable feedback mechanism. This will fully affect the stability of investment behavior under the framework of Bayesian thinking, the market generates a large amount of information every day, and how to establish a correct feedback mechanism for information to update decision-making is a compulsory course for investors. Taking the investment in the technology industry as an example, in recent years, the investment opportunities in the technology industry often come from fields that do not simply repeat historical experience, such as semiconductor localization, artificial intelligence, autonomous driving and other fields. The problem often encountered is that the market value of related companies reflects too fully on future business expectations, and once the industry progress is lower than expected, it will cause a large drawdown in stock prices. For this type of investment opportunity, it is necessary to first identify the key nodes of technological breakthroughs, and then fully verify the possibility of technological breakthroughs of relevant companies, and finally consider the market value space after technological breakthroughs. Under the premise of the probability of realization, the market value of the current company is fully compared to evaluate the risk-return ratio of the investment.

Third, fully respect the differences in cognition, and pursue the improvement of the winning rate. Bayesian thinking believes that probability represents a personal point of view, and that everyone can give their own determined probability of an event, which varies from person to person and there is no single standard. This view leaves room for cognitive differences between people. Everyone has different information, cognition, and judgment, and these differences lead to different people having different confidence in the occurrence of the same event, so this also explains why the periodic performance of current assets** and the intrinsic value of assets do not converge. Under the assumption of this investment philosophy, we cannot exhaust the truth of the world and pursue 100% accuracy, but we can get close to the real fundamentals through research to judge the reasonable range of various asset pricing, so as to improve the long-term winning rate.

Back to the current investment market, the world has undergone many changes in the past three years, especially the international situation, industrial policy and other macro-level events are not only difficult to advance in advance, but also have a long-term impact on the pricing mechanism of many listed companies. Navigating the complexities of market changes is the biggest challenge we face in our day-to-day investment work. In an uncertain environment, we can learn from Bayesian thinking to build investment strategies, actively learn Xi embrace changes, so that we can better cope with complex changes and maintain an open and calm mind.

Risk Warning:**There are risks and investment should be cautious. The views and judgments involved in the article only represent the views at the current point in time, and the opinions and judgments involved may be adjusted or changed subsequently due to the uncertainty and volatility of the market environment. This article is for communication purposes only and does not constitute any investment advice.

Related Pages