Abstracts
Paul Besson (Euronext) : When and at What Price are European Systematic Internalisers Providing Passive Liquidity?
Trades made by Systematic Internalisers (SIs) in Europe are reported via Approved Publication Arrangements (APAs). In our paper, based on public market data, we focus on SI Immediate trades based on MMT codes, which represent the largest share of SI trades on European Equities.
We first analyse the Price Improvements of these trades compared to consolidated Lit quotes. We evidence that SI trades take place at instances when consolidated spreads are larger compared to Primary trades. This leads to larger Realised Spreads on Sis, despite their Price Improvements. We then evidence that the Markouts following SI trades display very small adverse selection risk for SI market makers, compared to Lit venues. Lastly, using 5-minute buckets, we measure the average price at which passive liquidity is provided by SI market makers to aggressive traders. Average passive liquidity cost is computed using Volume-Weighted-Average-Passive-Price (Passive VWAP). We evidence that despite price improvements, aggressive participants on SIs consume liquidity at a greater cost than aggressive participants on Lit markets.
In addition, we provide two new methodological tools. We first propose a new way to flag midpoint trades sides based on the coinciding consolidated orderbook imbalances. We also propose a Passive VWAP computation over a time bucket to reflect the overall market timing of a passive liquidity provider
Disclaimer: This paper only represents the opinions of the authors. It is not intended to represent the position or opinions of Euronext, or the official position of any staff members. All errors or omissions are the authors’ sole responsibility.
Álvaro Cartea (Oxford): Spoofing with Learning Algorithms
This paper proposes a dynamic model of the limit order book to derive conditions to test if a trading algorithm will learn to spoof the order book. The testable conditions are simple and easy to implement because they depend only on the parameters of the model. We test the conditions with order book data from Nasdaq and show that market conditions are conducive for an algorithm to learn to spoof the order book.
(Based on joint work with Patrick Chang and Gabriel Garcia-Arenas)
CarteaCFM-Dec2023.pdfMihai Cucuringu (Oxford): Cross-Impact, Decomposition, and Co-Occurrence of Order Flow in Equity Limit Order Books
We investigate the impact of order flow imbalance (OFI) on price movements in equity markets in a multi-asset setting. We propose a systematic approach for combining OFIs at the top levels of the limit order book into an integrated OFI variable which better explains price impact, compared to the best-level OFI. We show that once the information from multiple levels is integrated into OFI, multi-asset models with cross-impact do not provide additional explanatory power for contemporaneous impact compared to a sparse model without cross-impact terms. On the other hand, we show that lagged cross-asset OFIs do improve the forecasting of future returns, and establish that this lagged cross-impact mainly manifests at short-term horizons and decays rapidly in time. Incorporating knowledge about order book event types leads to a decomposed OFI which attains significant improvement in a forward-looking predictive scenario.
Furthermore, we demonstrate that the time proximity of high-frequency trades contains a salient signal. We propose a method to classify every trade into five types, based on its proximity with other trades in the market, within a short period of time, ranging from 50 microseconds to 50 milliseconds. By means of a suitably defined normalized order imbalance associated to each type of trade, which we denote as conditional order imbalance (COI), we investigate the price impact of the decomposed trade flows. Our empirical findings indicate strong positive correlations between contemporaneous returns and COIs. In terms of predictability, we document that associations with future returns are positive for COIs of trades which are isolated from trades of stocks other than themselves, and negative otherwise. Furthermore, trading strategies developed using COIs achieve competitive returns and Sharpe Ratios, in an extensive experimental setup on a universe of over 450 stocks, for a period of three years.
Thierry Foucault (HEC): Algorithmic Pricing and Liquidity in Securities Markets
We let ``Algorithmic Market-Makers'' (AMs), using Q-learning algorithms, choose prices for a risky asset when their clients are privately informed about the asset payoff. We find that AMs learn to cope with adverse selection and to update their prices after observing trades, as predicted by economic theory. However, in contrast to theory, AMs charge a mark-up over the competitive price, which declines with the number of AMs. Interestingly, markups tend to decrease with AMs' exposure to adverse selection. Accordingly, the sensitivity of quotes to trades is stronger than that predicted by theory and AMs' quotes become less competitive over time as asymmetric information declines.
(Based on joint work with Jean-Edouard Colliard and Stefano Lovo. Paper)
Slides_CFM_Imperial.pdfOlivier Gueant (Sorbonne) : Market Making and Inventory Management Models: Where do We Stand?
Market making has been a classic subject in the market microstructure academic literature, generally divided into two main areas: informational aspects and inventory management. This talk aims to discuss recent advancements in inventory management models, especially those that include increasingly realistic features like tiering, externalization, complex price dynamics, and dynamic liquidity models. The talk will also address the challenges of the curse of dimensionality and present an effective approximation to overcome it.
slides Market Microstructure.pdfAlbert ("Pete") Kyle (Maryland): Flow Trading
We propose a new market design for trading financial assets. The design allows traders to directly trade any user-defined linear combination of assets. Orders for such portfolios are expressed as downward-sloping piecewise-linear demand curves with quantities as flows (shares/second). Batch auctions clear all asset markets jointly in discrete time. Market-clearing prices and quantities are shown to exist, despite the wide variety of preferences that can be expressed. Calculating prices and quantities is shown to be computationally feasible. Microfoundations are provided to show that traders can implement optimal strategies using portfolio orders. The proposal has several advantages over the status quo.
(Based on joint work with Eric Budish, Peter Cramton, Jeongmin Lee and David Malec. Paper)
KYLE-slides-CFM-Imperial-London-20231211.pdfVictor Le Coz (CFM and Ecole Polytechnique): When is Cross Impact Relevant?
Trading pressure from one asset can move the price of another, a phenomenon referred to as cross impact. Using tick-by-tick data spanning 5 years for 500 assets listed in the United States, we identify the features that make cross-impact relevant to explain the variance of price returns. We show that price formation occurs endogenously within highly liquid assets. Then, trades in these assets influence the prices of less liquid correlated products, with an impact velocity constrained by their minimum trading frequency. We investigate the implications of such multidimensional price formation mechanism on interest rate markets. We find that the 10-year bond future serves as the primary liquidity reservoir, influencing the prices of cash bonds and futures contracts within the interest rate curve. Such behaviour challenges the validity of the theory in Financial Economics that regards long-term rates as agents' anticipations of future short term rates.
(Based on joint work with Iacopo Matromatteo, Damiel Challet and Michael Benzaquen. Paper)
Cross_impact_presentation_Victor_Le_Coz.pdfFabrizio Lillo (Bologna): Online Learning of Market Liquidity
The estimation of market impact is crucial for measuring the information content of trades and for transaction cost analysis. Hasbrouck's (1991) seminal paper proposed a Structural-VAR (S-VAR) to jointly model mid-quote changes and trade signs. Recent literature has highlighted some pitfalls of this approach: S-VAR models can be misspecified when the impact function has a non-linear relationship with the trade sign, and they lack parsimony when they are designed to capture the long memory of the order flow. Finally, the instantaneous impact of a trade is constant, while market liquidity highly fluctuates in time.
This paper fixes these limitations by extending Hasbrouck's approach in several directions. We consider a nonlinear model where we use a parsimonious parametrization allowing to consider hundreds of past lags. Moreover we adopt an observation driven approach to model the time-varying impact parameter, which adapts to market information flow and can be easily estimated from market data. As a consequence of the non-linear specification of the dynamics, the trade information content is conditional both on the local level of liquidity, as modeled by the dynamic instantaneous impact coefficient, and on the state of the market.
By analyzing NASDAQ data, we find that impact follows a clear intra-day pattern and quickly reacts to pre-scheduled announcements, such as those released by the FOMC. We show that this fact has relevant consequences for transaction cost analysis by deriving an expression for the permanent impact from the model parameters and connecting it with the standard regression procedure. Monte Carlo simulations and empirical analyses support the reliability of our approach, which exploits the complete information of tick-by-tick prices and trade signs without the need for aggregation on a macroscopic time scale.
(Based on joint worl with Francesco Campigli and Giacomo Bormetti. Paper)
PresentationLillo.pdfCiamac Moallemi (Columbia): The Economics of Automated Market Making and Decentralized Exchanges
Automated market making (AMM) protocols such as Uniswap have recently emerged as an alternative to the most common market structure for electronic trading, the limit order book. Relative to limit order books, AMMs are both more computationally efficient and do not require the participation of active market making intermediaries such as high frequency traders. As such, AMMs have emerged as the dominant market mechanism for trust-less decentralized exchanges (DEXs) implemented through smart contracts on programmable blockchain platforms such as Ethereum. In cryptocurrency markets, the aggregate trading volume on the Uniswap DEX exceeds that of the much better known Coinbase centralized exchange.
We develop a model the underlying economics of AMMs from the perspective of their passive liquidity providers (LPs). Our central contribution is a ``Black-Scholes formula for AMMs''. Like the Black-Scholes formula, we consider the return to LPs once market risk has been hedged. We identify the main adverse selection cost incurred by LPs, which we call ``loss-versus-rebalancing'' (LVR, pronounced ``lever''). LVR captures costs incurred by AMM LPs due to stale prices that are picked off by better informed arbitrageurs. In a continuous time Black-Scholes setting, we are able to derive closed-form expressions for this adverse selection cost. Qualitatively, we highlight the main forces that drive AMM LP returns, including asset characteristics (volatility), AMM characteristics (curvature / marginal liquidity, fee structure), and blockchain characteristics (block rate). Quantitatively, we illustrate how our model's expressions for LP returns match actual LP returns for the Uniswap v2 WETH-USDC trading pair. Our model provides tradable insight into both the ex ante and ex post assessment of AMM LP investment decisions. LVR can also inform the design of the next generation of DEX market mechanisms --- in fact, in the short time since our work has been released, ``LVR mitigation'' has already emerged as the dominant challenge among practitioners in the AMM protocol designer community.
(Based on joint work with Jason Milionis, Tim Roughgarden, and Anthony Zhang. Paper 1 and Paper 2)
2023-12-cfm-imperial-lvr-2up.pdfRoel Oomen (Deutsche Bank): Pre-Hedging
This paper studies a dealer that pre-hedges anticipated potential trades and we analyse how this affects the client's overall execution outcomes. Pre-hedging can benefit both parties: improved risk management over an expanded horizon then enables the dealer to charge reduced spreads that more than offset any adverse impact the pre-hedging activity has on the execution price. However, when a dealer pre-hedges too aggressively, this can be detrimental to the client. This result is robust to a setting where competing dealers simultaneously pre-hedge. Any counter-productive pre-hedge activity can be mitigated by introducing timing uncertainty of the potential trade.
(Based on joint work with Johannes Muhle-Karbe. Paper)
Barbara Rindi (Bocconi): Optimal Tick Size
We use a model of a limit order book to determine the optimal tick size that maximizes welfare of market participants. When investors arrive sequentially and supply liquidity by undercutting or queuing behind existing orders, the optimal tick size is a positive function of the asset value and a negative function of trading activity. We use the introduction of MiFID II to empirically show that the new tick size regime based on price and trading activity benefited market participants. Our results suggest that both the European tick size regime and (partially) the 2022 SEC proposal dominate Reg. NMS Rule 612.
(Based on joint work with Giuliano Graziani and Bart Zhou Yueshen. Paper)
OTS_GRY_slides_Imperial_11122023.pdfMathieu Rosenbaum (Ecole Polytechnique): The Two Square-Root Laws of Market Impact and the Role of Sophisticated Market Participants
The goal of this work is to disentangle the roles of volume and participation rate in the price response of the market to a sequence of orders. To do so, we use an approach where price dynamics are derived from the order flow via no arbitrage assumptions. We also introduce in the model sophisticated market participants having superior abilities to analyse market dynamics. Our results lead to two square root laws of market impact, with respect to executed volume and with respect to participation rate.
(Based on joint work with Bruno Durin and Grégoire Szymanski.)
Pres_Rosenbaum_London_111223.pdfDimitri Vayanos (LSE): Long-Horizon Investing in a Non-CAPM World
We study dynamic portfolio choice in a calibrated equilibrium model where value and momentum anomalies arise because capital slowly moves from under- to over-performing market segments. Over short horizons, momentum’s Sharpe ratio exceeds value’s, the value-momentum correlation is negative, and the conditional value-momentum correlation positively predicts Sharpe ratios of value and momentum. In contrast, over long horizons, value’s Sharpe rati can exceed momentum’s, the value-momentum correlation turns positive, and the value spread becomes a better predictor of Sharpe ratios. Momentum’s optimal portfolio weight relative to value’s declines significantly as horizon increases. We provide novel empirical evidence supporting our model’s predictions.
(Based on joint work with Christopher Polk and Paul Woolley. Paper)
PVW_Imperial.pdfWei Xiong (Oxford): Dynamics of Market Making Algorithms in Dealer Markets: Learning and Tacit Collusion
The widespread use of market-making algorithms in electronic over-the-counter markets may give rise to unexpected effects resulting from the autonomous learning dynamics of these algorithms. In particular the possibility of “tacit collusion” among market makers has increasingly received regulatory scrutiny. We model the interaction of market makers in a dealer market as a stochastic differential game of intensity control with partial information and study the resulting dynamics of bid-ask spreads. Competition among dealers is modeled as a Nash equilibrium, while collusion is described in terms of Pareto optima. Using a decentralized multi-agent deep reinforcement learning algorithm to model how competing market makers learn to adjust their quotes, we show that the interaction of market making algorithms via market prices, without any sharing of information, may give rise to tacit collusion, with spread levels strictly above the competitive equilibrium level.
(Based on joint work with Rama Cont. Paper)