Endogenous participation in otherwise segmented markets

The activity of arbitrage is an essential part of a healthy financial market. Arbitrageurs presence increases market elasticity allowing to trade larger quantities with lower price impact. Overall this is a desirable feature of a market as it allows for improved risk sharing and higher welfare (Gromb and Vayanos 2002). A number of papers build on the observation that arbitrageurs always participate in otherwise segmented markets. However, the participation of arbitrageurs is likely to depend on market conditions, and in particular, volatility.

There are contrasting views on the impact of volatility on arbitrageurs' participation. Grossman and Miller (1988) predict that volatility is associated with increased market participation. In their model, arbitrageurs face a fixed participation cost and since volatility increases the price discount, it helps recover the fixed cost. Theories based on financial frictions such as Gromb and Vayanos (2002) and Brunnermeier and Pedersen (2008) predict the opposite. This is because volatility tightens the financial friction, reducing resources available to arbitrageurs. Thus, according to one view arbitrageurs act as shock absorbers, while in the other as shock amplifiers. What drives arbitrageurs' participation, how does their participation relate to market conditions such as volatility, and what effects does endogenous participation have on optimal policy?

In this paper, I study empirically the drivers of arbitrageurs' participation using novel data on the prime example of arbitrageurs in the real world: hedge funds. Specifically, I test whether arbitrageurs are more likely to participate around episodes of market volatility, using a granular transaction level dataset that tracks hedge funds behaviour in real time. To address concerns about the endogeneity of volatility, I use scheduled macroeconomic news releases and government bond auctions as instruments for heightened fundamental and demand volatility, respectively. Consistent with shock-absorber view, I find that arbitrageurs are net buyer of government bonds at risky times, while they are net sellers on the average day. Moreover, this effect is largely driven by their purchases of long maturity bonds.

Much of the debate surrounding the role of hedge funds is their role around large shocks, such as financial crises. While scheduled events provide a relatively clean volatility test, it is unlikely that it is equally informative of large shocks. To address this concern, I use the LDI crisis in September 2022 in the United Kingdom as an environment in which shocks to information and central bank gilt purchases can be used to further test how arbitrageurs react to extreme bouts of volatility. I find that also on this occasion, hedge funds were net buyers of UK government bonds and helped to absorb the negative demand shock from pension funds. However, I do find that they purchased mostly short-dated bonds, accumulating a large net long position which then gradually reversed over the following six months.

Overall, my findings point to a positive role of hedge funds in the market: they absorb shocks and provide liquidity when other market participants are less willing to do so. My results contrast with the previous literature that identified hedge funds engaging in destabilizing speculation. It is plausible that this discrepancy is the result of different shocks in different sample, with most studies focusing on hedge funds around the 2007-08 financial crisis where the shock originated in the financial sector. My study instead focuses on a more recent sample (2018-2024) with shocks originating mostly from the real side of the economy.