Predicting Stock Returns: Random Walk or Herding Behaviour?

Please cite the paper as:
Alessandro Zoino CFA, (2019), Predicting Stock Returns: Random Walk or Herding Behaviour?, World Economics Association (WEA) Conferences, No. 1 2019, Going Digital, 15th November to 20th December, 2019


The analysis is based on Efficient Market Hypothesis and predictability of stock returns. Both concepts will be theoretically showed and demonstrated following the most known asset pricing theories and studying the bid/ask spreads. The 3 Efficient Market Hypotheses will be empirically tested through the Technical Analysis and the Fundamental Analysis. The aim of the model is to highlight the Efficient Market Hypothesis’ limits by instructing several tests related to the irrationality of investors, such as the analysis of the noise trading and the respectively effects on volatility, the empirical test of hedging, arbitrage and speculation concepts, the demonstration of the divergence between behavioural finance and expected returns, to conclude with the comparison between random walk and herding behaviour.


Keywords: , , , ,

Recent comments



  • Bin Li says:

    The test is meaningful, but, even the theories tested herein are all “proved” or “falsified”, none of them is appropriate. How does stock market run? We need a comprehensive framework theory to clarify everything mentioned in the paper, which lead to an all-in-one economics-Algorithmic Economics. Economics have wandered for too long times, and should not go in old ways any more! Economists should not flatter the literature traditions any longer while reiterating “critics”, and should not suspect and reject new things while reiterating “innovation”!

    • Alessandro Zoino says:

      Thanks a lot for your feedback. Fully defining how the stock market run is quite challenging but the key focus currently is in identifying theories still valid and theories becoming obsolete due to new structural regimes. And, within the surviving ones, the new successful frameworks are the ones able to split and critic part of the no-longer-empirically-verified theories by reshaping the old frameworks and include the innovation impact. The macro and economical environments experienced a structural shift and this shift made many theories partially no-longer-empirically demonstrated. A common empirical example is the negative rates on the ECB side and the ZIRP concept – the current market balance is an environment never taken into account in past frameworks and successful models are the ones starting from removing the constraint of not accepting negative rates.

  • goingdigital2019 says:

    Dear Alessandro,

    After reading your paper, my main question is: to what degree the theoretical consideration of AI in finance is compatible with the real principle of ontological uncertainty in economic models?

    Thanks in advance,


    • Alessandro Zoino says:

      Hi Maria, thanks for your feedback. AI is disrupting the current environment at a global level and the impact will be even more relevant as soon as the availability and the access to AI outputs will be spread among all agents. As every innovation, the beginning is all about “test and see” to then start developing solid and empirically tested models. This current pioneering environment has uncertainty as one of the main variables.