How Competitive is the Stock Market? Theory, Evidence from Portfolios, and Implications for the Rise of Passive Investing

Valentin Haddad, Paul Huebner, Erik Loualiche Video

Abstract

We develop a framework to theoretically and empirically analyze investor competition on financial markets. The classic view assumes that markets are very competitive: if a group of investors changes its behavior, other investors react such that nothing happens in equilibrium. Our framework quantifies the strength of the competitive response. We estimate a demand system of institutional investors in the US stock market accounting for two layers of equilibrium: how investors compete with each other in setting their strategies and how prices adjust to clear asset markets. We find that investors react to the behavior of others in the market: when an investor is surrounded by less aggressive traders she trades more aggressively. This reaction reduces the equilibrium consequences of changes in individual behavior by 50%. However, it also implies that the stock market is far from the competitive ideal. A consequence of this result is that the large increase in passive investing over the last 20 years has led to substantially more inelastic aggregate demand curves for individual stocks, by 15%.

Highlights

Each investor demands some amounts of an asset, decreasing in the price of the asset by the elasticity: $d_i={\underline{d}}_i-{\mathcal{E}}_i\times p$ In equilibrium, demand sums up to supply: ${\int }_i{D}_i(p)=S$ Elasticity is determined by individual effect and response to aggregate demand elasticity: ${\mathcal{E}}_i=\underline{{\mathcal{E}}_i}-\chi \times {\mathcal{E}}_{agg}$ In equilibrium, aggregate elasticity is equal to average investor elasticity (weighted by demand): ${\int }_i{\mathcal{E}}_i{D}_i/S={\mathcal{E}}_{agg}$

Can think about $\chi$ as representing competition. If $\chi =0$, there is no competition — each investor follows their own strategy totally independent of the rest of the market. In perfect competition, $\chi =\infty$, and any change is completely counteracted by investor reaction. Related to Statistical Arbitrage with Uncertain Fat Tails

Large increase in passive investors over time. In no competitive world, this lead to proportional reduction in elasticity. In perfectly competitive market, increase in passive investors has no effect on elasticity.

Authors estimate that $\chi =1.7$, relatively stable over time.

Follow similar methodology as A Demand System Approach to Asset Pricing, with logit specification for portfolio shares $w_{ik}$. Track portfolio shares relative to price as a measure of demand, modeled as investor-specific function of observables and investor-specific elasticity: $\log \frac{w_{ik}}{w_{i0}}-p_k={\underline{d}}_{0i}+{\underline{d}}_{1i}^{\prime }{X}_k-{\mathcal{E}}_{ik}{p}_k+{ϵ}_{ik}$

Investor-specific elasticity modeled as investor-specific function of observables and market elasticity: ${\mathcal{E}}_{ik}={\underline{\mathcal{E}}}_{0i}+{\underline{\mathcal{E}}}_{1i}^{\prime }{X}_k-\chi {\mathcal{E}}_{agg,k}$

Investors respond less to price movements for assets with more aggressive investors than for assets with less aggressive investors. If all other investors are more elastic by 1, lower my elasticity by 1.7.

Elasticities in general are low — this study finds 0.3. Price elasticity in the stock market is lower for large companies, approaching zero for the largest stocks:

online