Intraday Forecasting Ability and Liquidity Provision of Institutional Traders

summary

The gist

Intraday Forecasting Ability and Liquidity Provision of Institutional Traders This research investigates the relationship between forecasting ability, liquidity provision intensity…

In short

The research analyzed institutional traders' intraday profitability using high-frequency limit order book data. It found that a trader's ability to forecast future price changes from their own trading flow is strongly linked to higher profits. Furthermore, providing liquidity through passive executions significantly boosts returns, with two passive executions yielding substantially more profit than zero.

Key concepts

Forecasting Ability ($f_i$)
This measures a trader's skill in predicting future price movements based on their current execution flow. It is quantified by linking the trader's behavior to subsequent price changes over different time horizons, showing how well they anticipate market shifts.
Liquidity Provision Intensity
This refers to how much a trader acts as a market maker by placing passive orders (limit orders). The study found that increasing this intensity—from zero passive executions to one or two—directly correlates with higher realized profits for the institution.
FIFO-Matched P&L Framework
This is the mathematical structure used to link an individual trader's profit and loss observation to specific trading actions. It ensures that each profit calculation is matched precisely with a corresponding measure of forecasting skill and execution characteristics.

Terminology used across episodes

This episode discusses

The paper

Intraday Forecasting Ability and Liquidity Provision of Institutional Traders · Read on arXiv

Xinlei Hao, Jiajun Guo, Jussi Keppo, Xianhua Peng

HSBC Business School, Peking University · Department of Mathematics, The Hong Kong University of Science and Technology · Business School, National University of Singapore

Transcript

Introduction to the show: ident: Quantitative Finance. The day in quantitative finance: models of markets, risk and prices.

Ruth: We're looking at the paper: "Intraday Forecasting Ability and Liquidity Provision of Institutional Traders".

Warren: Here's the gist of the paper. Comprehensive Research Synthesis: Intraday Forecasting Ability and Liquidity Provision of Institutional Traders This research investigates the relationship between forecasting ability, liquidity provision intensity, and proprietary intraday profitability among institutional traders using a high-frequency dataset derived from stock exchange limit order book events.

Ruth: The study meticulously constructs a framework linking these factors through a FIFO-matched Profit & Loss (P&L) structure, allowing for the examination of how realized P&L covaries with trading skill and liquidity provision characteristics.

Warren: Data and Methodology Foundation The analysis is grounded in a high-frequency dataset encompassing limit order book events for five stocks spanning November 22 to November 26, 2010.

Ruth: This data is rich, including the time of event, type of event (limit or market order), volume, price, identity of both the liquidity taker and provider, and crucial trading capacity information distinguishing proprietary trading from client orders.

Title and authors: Ruth: So we’re diving into the specific title and authors of this paper, "Intraday Forecasting Ability and Liquidity Provision of Institutional Traders." It’s important to understand who did this research because it sets the stage for what they found.

Warren: The authors are Xinlei Hao, Jiajun Guo, and Jussi Keppo, with Xianhua Peng as well. They’re bringing together expertise from markets and incentives, quantitative trading, and econometrics.

Iris: That mix is key because you need someone to handle the econometric identification of what variation is actually driving the results in this complex P andL structure.

Ruth: Exactly. And when we look at the title again, it highlights two different channels for profit: forecasting ability and liquidity provision intensity. It’s not just one or the other; it’s about how they interact.

Warren: They are essentially asking if being good at predicting prices helps you make better money, and if being a good liquidity provider—supplying those trades—also adds to that profit.

Iris: And that interplay is what makes the empirical accounting demanding, because you have to measure both skill and market making roles simultaneously at a high frequency.

Ruth: It’s not just about finding skill; it’s about figuring out how that skill manifests financially when you consider the constraints of the order book and your execution style.

Warren: And that leads into the core question they answer: is it information rent from prediction, or is it revenue from supplying immediacy?

Iris: The methodology hinges on matching buy and sell executions into FIFO P andL observations so they can link those outcomes to the specific holding interval and the roles of those two executions.

Ruth: So, to recap, we’re looking at how forecasting ability and liquidity provision intensity relate to proprietary intraday profitability in this paper. It sets up a way to test if traders with certain skills or certain market-making behaviors earn more than others.

Warren: And it opens up the idea that HFT profits might come from both anticipating future prices and supplying necessary market liquidity at the right time.

The paper's summary: Ruth: Now, let’s summarize what this paper actually says about its findings regarding forecasting ability and profitability. The core finding is that there is a real correlation between a trader's ability to predict future price changes from their own execution flow and their actual P andL.

Warren: That correlation isn't just theoretical; they show that hypothetical trading strategies based on the execution flows of these skillful traders can achieve annualized Sharpe ratios ranging from sixteen to sixty-three across various trading frequencies <ref:2610.31141#pg1,annualized Sharpe ratios ranging from 16 to 63 across various trading frequencies>.

Iris: And when you look at the direct link between skill and profit, a one-standard-deviation increase in that forecasting ability proxy is associated with currency units higher P andL per matched position, specifically between one point zero two and one point one nine currency units higher P andL per matched position <ref:2610.31141#pg1,1.19 currency units higher P&L per matched position>.

Ruth: That’s a pretty concrete number showing the direct payoff of having better foresight when you’re trading intraday in this way, holding all other factors constant in the model construction.

Warren: But it doesn't stop there; they also look at liquidity provision intensity—how many passive executions are involved in a matched position: zero, one, or two.

Iris: And the paper shows that having just one passive execution yields P andL that is one point two eight to one point three one currency units higher than positions with no passive executions, and having two passive executions yields an even bigger jump of three point seven nine to three point eight two currency units higher than those with one passive execution.

Ruth: So, the summary paints a picture where both being skilled at predicting price moves and being active in providing liquidity are significant drivers for institutional traders' returns.

Warren: It confirms that liquidity provision isn't just some side effect; it’s a distinct source of financial benefit when you provide that immediacy.

The paper's improvements: Ruth: The authors suggest a few ways to take this research forward, focusing on how we can make the identification of these effects even stronger and more actionable for trading strategies.

Warren: They propose developing a persistent trading ability detection system by calculating a group measure, GM1, for traders who show that positive correlation between their execution flow and future price changes in early periods.

Iris: That GM1 idea is interesting because it’s meant to provide group-level evidence that supports the persistence of the price-forecasting ability they captured in the initial classification.

Ruth: They also suggest creating a high Sharpe ratio flow-following trading strategy that specifically targets those traders identified as having significant positive forecasting correlations across different trading frequencies.

Warren: And they propose building a P andL prediction model that uses the horizon-matched forecasting proxy to estimate how many currency units higher P andL per matched position you can expect, based on that one standard deviation increase finding.

Iris: Finally, they suggest designing a liquidity provisioning optimization module that analyzes how realized proprietary intraday P andL covaries with both the trader’s price-forecasting ability and the passiveness of its two executions.

Ruth: That last suggestion seems to be the practical application; it’s about creating an execution strategy that prioritizes positions with two passive executions because those positions have P andL that is three point seven nine to three point eight two currency units higher.

Warren: It’s moving from just descriptive findings to prescriptive strategies based on understanding the interaction between skill and liquidity provision intensity within a single trade structure.

Conclusion: Ruth: So, wrapping up the discussion on this paper, it really shows that institutional traders are better off if they combine good forecasting ability with active liquidity provision. The findings from "Intraday Forecasting Ability and Liquidity Provision of Institutional Traders" confirm that both skill and market-making roles contribute significantly to proprietary intraday profitability.

Warren: The Sharpe ratios they found, ranging from sixteen to sixty-three show the potential upside when you leverage these forecasting abilities through flow-following strategies <ref:2610.31141#pg1,ranging from 16 to 63>. It’s a measure of how much return you can expect based on that skill alone.

Iris: From an econometric view, the paper successfully establishes that the variation in forecasting ability is a valid driver for P andL, and it provides concrete numbers—like the one point zero two to one point one nine currency units higher P andL per matched position—that give us a tangible link between skill and profit <ref:2610.31141#pg1,1.19 currency units higher P&L per matched position>.

Ruth: It’s about understanding that profitability isn't just one thing; it’s a combination of anticipating what’s coming and being present to facilitate trades with liquidity.

Warren: And the paper suggests that if you want to build a real strategy, you need to incorporate both the skill aspect and the liquidity aspect into your optimization module.

Iris: The work provides rigorous evidence for how these two mechanisms interact, but it also flags a limitation: they are working with a specific set of data from November 22nd to November 26th, two thousand ten. So we don't know if those findings hold up over much longer time spans.

Ruth: Right. So that’s the picture on this paper, "Intraday Forecasting Ability and Liquidity Provision of Institutional Traders." It gives us a solid empirical basis to think about how institutional trading profits are actually generated in real-time.

Warren: We're ready for the next piece of work when we are.

More episodes

← Home