arXiv subjects

Tianshu Lyu

Publications and source records attributed to Tianshu Lyu.

3 recordsLinked to original sources

arXiv2025

Causal Inference in Financial Event Studies

Financial event studies, ubiquitous in finance research, typically use linear factor models with known factors to estimate abnormal returns and identify causal effects of information events. This paper demonstrates that when factor models are misspecified -- an almost certain reality -- traditional event study estimators produce inconsistent estimates of treatment effects. The bias is particularly severe during volatile periods, over long horizons, and when event timing correlates with market conditions. We derive precise conditions for identification and expressions for asymptotic bias. As an alternative, we propose synthetic control methods that construct replicating portfolios from control securities without imposing specific factor structures. Revisiting four empirical applications, we show that some established findings may reflect model misspecification rather than true treatment effects. While traditional methods remain reliable for short-horizon studies with random event timing, our results suggest caution when interpreting long-horizon or volatile-period event studies and highlight the importance of quasi-experimental designs when available.

econ.EM↗

arXiv2021

On Length Divergence Bias in Textual Matching Models

Despite the remarkable success deep models have achieved in Textual Matching (TM) tasks, it still remains unclear whether they truly understand language or measure the semantic similarity of texts by exploiting statistical bias in datasets. In this work, we provide a new perspective to study this issue -- via the length divergence bias. We find the length divergence heuristic widely exists in prevalent TM datasets, providing direct cues for prediction. To determine whether TM models have adopted such heuristic, we introduce an adversarial evaluation scheme which invalidates the heuristic. In this adversarial setting, all TM models perform worse, indicating they have indeed adopted this heuristic. Through a well-designed probing experiment, we empirically validate that the bias of TM models can be attributed in part to extracting the text length information during training. To alleviate the length divergence bias, we propose an adversarial training method. The results demonstrate we successfully improve the robustness and generalization ability of models at the same time.

cs.CL↗

arXiv2019

Compositional Network Embedding

Network embedding has proved extremely useful in a variety of network analysis tasks such as node classification, link prediction, and network visualization. Almost all the existing network embedding methods learn to map the node IDs to their corresponding node embeddings. This design principle, however, hinders the existing methods from being applied in real cases. Node ID is not generalizable and, thus, the existing methods have to pay great effort in cold-start problem. The heterogeneous network usually requires extra work to encode node types, as node type is not able to be identified by node ID. Node ID carries rare information, resulting in the criticism that the existing methods are not robust to noise. To address this issue, we introduce Compositional Network Embedding, a general inductive network representation learning framework that generates node embeddings by combining node features based on the principle of compositionally. Instead of directly optimizing an embedding lookup based on arbitrary node IDs, we learn a composition function that infers node embeddings by combining the corresponding node attribute embeddings through a graph-based loss. For evaluation, we conduct the experiments on link prediction under four different settings. The results verified the effectiveness and generalization ability of compositional network embeddings, especially on unseen nodes.

cs.LG↗