Weak-Form Market Efficiency in the Dhaka Stock Exchange: Evidence from a Multi-Test Approach Around the 2024 Political Transition

In the world of finance, there is a long-standing idea called the Efficient Market Hypothesis. It suggests that stock prices are like a fair coin toss: they change randomly because they instantly absorb every piece of news and information available to everyone. If this idea is true, then looking at how a stock moved yesterday cannot help you guess where it will go tomorrow. The past holds no secret clues for the future. This concept is the bedrock of modern investing, but it is often tested in places where the rules are less clear. In many developing nations, markets can be thin, meaning not many people trade every day, and information might not spread as quickly as it does in wealthy countries. When prices do not move randomly, it opens the door for investors to potentially spot patterns and make money based on history alone. Understanding whether a market is truly random or if it follows hidden patterns is crucial for anyone trying to understand how value is determined in these economies.

Researchers from the Bangladesh University of Professionals decided to put this idea to the test on the Dhaka Stock Exchange, the main stock market in Bangladesh. They were particularly interested in a specific moment in history: the political transition that occurred in August 2024, when the government resigned and an interim administration took over. Markets often react violently to such changes, and the researchers wanted to know if this shift in leadership also changed how the market behaved. Did the chaos of the transition make the market more chaotic, or did the new order bring more predictability? To find out, they gathered daily price data for thirty of the largest and most active companies on the exchange, along with the main index that tracks them, covering a period from early 2016 through mid-2025. They did not rely on just one way of checking; instead, they used a battery of five different statistical methods, each designed to catch a different kind of pattern, from simple connections between days to more complex trends over weeks.

The results from the long-term data, covering nearly a decade, were clear and consistent: the market was not behaving randomly. In fact, the tests showed that for the vast majority of these stocks, the past did influence the future. When the researchers looked at the daily price changes, they found that prices tended to move in clusters rather than in a straight, unpredictable line. If a stock went up one day, it was more likely to keep moving in a similar direction for a few days afterward, or it would show a pattern that could be detected by their tools. This happened for about eighty-four to ninety-seven percent of the stocks they studied, depending on which specific test was used. Even the main index, which is supposed to represent the whole market, showed these predictable patterns. This suggests that the market is not yet efficient in the strictest sense; there are still frictions, such as slow information flow or trading rules, that prevent prices from adjusting instantly and randomly.

The researchers then zoomed in on the period after the political transition in August 2024 to see if the new government had fixed these issues. At first glance, the numbers looked promising. When they ran the same tests on the data from just the year following the transition, the number of stocks showing predictable patterns dropped dramatically. The tests that had previously rejected the idea of randomness now mostly accepted it. It seemed as though the market had suddenly become more efficient and more random. However, the researchers were careful not to jump to this conclusion. They dug deeper and found that the apparent improvement was likely an illusion caused by the short amount of time they had to study. The period after the transition was only about one year long, which is a very small sample compared to the nine years of data they had before. When they checked the actual numbers behind the scenes, they found that the underlying patterns had not actually changed. The mathematical coefficients that measure how much yesterday's price affects today's price did not get smaller; they stayed roughly the same. The tests simply failed to detect the patterns because there was not enough data to be sure.

Furthermore, the researchers looked for a specific "break" in the data, a moment where the rules of the market seemed to fundamentally shift because of the political change. They found no evidence of such a shift. The way the stocks behaved before the transition was statistically indistinguishable from how they behaved after it. The decline in the number of detected patterns was not because the market had improved, but because the statistical tools were less powerful when applied to a shorter timeline. The study concluded that the inefficiencies in the Dhaka Stock Exchange are rooted in deep-seated structural issues, such as thin trading volumes and regulatory constraints like price limits that prevent stocks from moving freely, rather than in who is running the government. A change in leadership alone did not fix the market's ability to process information.

One specific factor that complicated the picture was the use of "floor prices" by regulators. For a time, authorities set a minimum price below which certain stocks could not trade. This meant that even if a stock should have dropped in value, the price stayed stuck, creating long strings of days with no change. This artificial stability created patterns that looked like predictability but were actually just the result of a rule. Even after the political transition, some of these restrictions remained in place for a few major companies, further muddying the water. The researchers noted that while the market showed some statistical signs of predictability, these patterns were small and likely too weak to be exploited for easy profit once trading costs and fees were taken into account. The market was not random, but it was also not a place where a simple strategy could guarantee a fortune.

Ultimately, the study paints a picture of a market that is still maturing. The Dhaka Stock Exchange does not yet follow the perfect random walk seen in some of the world's most developed financial centers. The patterns the researchers found are typical of emerging markets where trading is less frequent and information moves more slowly. The political transition of 2024, while a historic event for the country, did not act as a magic switch that instantly modernized the stock market. The inefficiencies remain, driven by the mechanics of how the market operates rather than the politics of the day. For the market to become truly efficient, the researchers suggest that the focus must shift to improving liquidity, removing artificial price barriers, and ensuring that information flows freely, rather than waiting for a change in government to solve the problem. The data shows that the market is listening to history, but it is still learning how to listen to the present.