Community Research and Development Information Service - CORDIS

Final Report Summary - MSAEO (Market Selection and Aggregate Economic Outcomes)

The research project “Market Selection and Aggregate Economic Outcomes” aims at investigating possible failures of the Market Selection Hypothesis (MSH) and whether they are helpful in describing aggregate economic data. The hypothesis states that well functioning competitive markets reward those economic actors who act rationally and have the most accurate information on future events. As a result, these actors gain all the economic resources in the long run and determine aggregate economic outcomes, such as prices and output. The validity of the hypothesis underlies most of mainstream economic models of both pure exchange and production economies. The alternative view is instead that markets, even when they are competitive and allow for the exchange of all possible hedging contracts, do not perfectly select among risk attitudes, information sets, propensity to save and invest or, more generally, behaviours. The main objective of the present project is to investigate the validity of the MSH and the impacts of its possible failures. Intermediate steps are both the development of the general economic model that accounts for the interaction of heterogeneous economic actors and the development of the mathematical tools needed for its investigation.

During the first phase of the project, the fellow concentrated his effort on benchmark market selection model of financial markets that improves upon the existing literature by considering more general investors behaviours and possibly incomplete spot markets. In particular, the fellow worked on a general framework where existing approaches that validate the MSH for complete financial markets can be confronted with the negative results that emerged from preliminary investigations. The framework is based on the comparison of agents' intertemporal substitution rates and portfolio expected log-returns. The latter are shown to depend both on agents' information accuracy and on their preferences. The common framework allowed the fellow to shed light on the fact that models that support the Market Selection Hypothesis rely on a trade-off between intertemporal substitution rates and relative risk aversion that arises from the time separability of preferences. When preferences are non time-separable, e.g. when utility has a recursive formulation, market selection failures occurs even if some investors have perfect information on assets' dividend process and optimally allocate their consumption across time and states. Other critical assumptions are shown to be 1) agents have rational price expectations even when they 'agree to disagree' on the exogenous dividend process, 2) agents' beliefs are exogenous and do not interact with eachothers through prices. The relaxation of either assumption has been investigated in two separate lines of research and has shown that, everything else being equal, coexistence of heterogeneous trading strategies is a generic outcome of the market selection process.

Results summarised above are presented in the following papers:
- G. Bottazzi and P. Dindo (2014) "Evolution and Market Behavior with Endogenous Investment Rules" published in the Journal of Economic Dynamics and Control, Volume 48, Pages 121–146.
- P. Dindo (2015) “Selection in Speculative Markets”, LEM Working Paper 2015-32; (main paper)
- G. Bottazzi, P. Dindo, and D. Giachini (2015) “Long-run Heterogeneity in an Exchange Economy with Fixed-Mix Traders”, LEM Working Paper 2015-29, investigates point 1) above.
- P. Dindo and F. Massari (2016). “The Wisdom of the Crowd Revisited”, investigates point 2) above.

During the second phase of the project, a complementary effort was devoted to providing a market-selection-based explanation of empirical regularities in financial markets, linking them to the real performances of their underlying assets. The first line of research has concentrated on the explanation of positive returns autocorrelation at short horizons and negative returns autocorrelation at long horizons. Whereas the behavioural finance literature links the emergence of these empirical regularities to a representative investor's cognitive biases in the processing of new information, the fellow has proposed a model where the coexistence of agents with different portfolios is the key factor. Building upon the project output Bottazzi et al (2015), the fellow has showed that a ”market sentiment” emerges as opposed to behavioural finance “investor sentiment”: despite single investors hold constantly rebalanced portfolios, their relative wealth dynamics is such that the aggregate investor reacts to news as behavioural models suggest.

The second line of research has looked at the phenomenon of fire sales and at their impact on the overall market performance. Consider the case of a trader being hit by an exogenous liquidity shock. If her portfolio mainly consisted of cash or other low-risk securities such as government bonds, it would just take to borrow what necessary using (part of) her holdings as collateral. Conversely, if her portfolio was primarily made of high-risk securities, she would have no choice but selling them to be able to honour her short-term liabilities, since risky assets are not generally accepted as collateral. Under this scenario, should a shock trigger a fire sales phenomenon on the side of the holders of risky stocks, a trade-off between the two outlined strategies clearly emerges. On the one hand, taking more risk brings a positive reward since riskier securities carry higher expected returns. On the other hand, a portfolio which is exceedingly skewed towards risk also turns out more vulnerable to external liquidity shocks. Building upon Bottazi and Dindo (2014), the fellow was able to analytically investigate this trade-off and provide sufficient conditions for the long-run survival of each strategy in terms of relative wealth. This is the channel by which, even in bullish markets, a riskier portfolio may systematically fail to outperform a relatively safer one and yet cause market instability.

Results summarised above are presented in the following papers:
- G. Bottazzi, P. Dindo, and D. Giachini (2016) “A Model of Market Sentiment” proposes a market selection based explanation of momentum and reversal to the mean.
- P. Dindo and J. Staccioli (2016) “Wealth-driven asymptotic survival in a financial market with demand shocks”, investigates the performance of portfolio strategies under liquidity shocks.

Technically all these paper analyse a Random Dynamical System of some sort. Another output of the project has been the development of tools to analyse the long-run behavior of a stochastic process by exploiting so called asymptotic drift conditions. Such results relies on applications of the martingale convergence theorem and are presented in the following paper
- G. Bottazzi and P. Dindo (2015) “Drift criteria for persistence of discrete stochastic processes on the line”, LEM Working Paper 2015-26;

During the opening address at the ECB Central Banking Conference held in Frankfurt on November 18, 2010, Jean-Claude Trichet, President of the ECB, declared that “In the face of the crisis, we felt abandoned by conventional tools”. Most conventional tools do rely on the validity of the Market Selection Hypothesis. Having shown that the hypothesis fails generically, at least in exchange economies, provides a better understanding of the forces that are shaping asset prices dynamics in our financial markets. Further work is needed to test the proposed model, comparing its predictions with known financial returns stylized facts, and to complement the model by considering the linkages between the financial and the real sectors. However, when successful, we might have a deeper understanding in the functioning of markets in producing aggregate economic outcomes.

Related information

Reported by

Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top