A Short Introduction to Investor Networks
What’s an Investor Network, how to build one, and what can we do with it?
Yu Zhu (祝宇)
Ph.D. Candidate, Information SystemsUniversity of Utah, USA
Ph.D., FinanceZhejiang University, China
Quicktake: BPE, WordPiece, and SentencePiece
One table to compare popular tokenization methods: BPE, WordPiece, and SentencePiece.
Implementation details: From the original Transformer to GPT
This article comapres the implementation details between the original Transformer and GPT. These tricks are critical to performance but not always explained in the paper.
Transformers (deep learning models) are better at predicting the tail distribution
In a setting of PEAD (post-earnings-announcement-drift) prediction using earnings call transcripts, I found Transformers (deep learning models) have a larger performance lead on extreme data points (data at the tails of the distribution).
(R) Super Efficient Event Study Code
A highly efficient (>10x speedup) and concise (core part is less than 40 lines) event study R program that replicates WRDS’s SAS version.
All you need to know about data/time in R and SAS
How date and time are stored/processed in R and SAS, and how to use them wisely in WRDS or in your daily programming.
Git tips
As a Git non-prouser, I found these tips very helpful.
Some facts on Earnings Announcement and Analyst Revision
Some facts on earnings announcement and analyst revision that you may not know before.
Expected Return Model
How to choose the right expected return model?
Synthetic Control
Quick recipe for a synthetic control study
Difference-in-Differences
Quick recipe for a DD study
Regression: How to use it for causal inference?
Under what conditions does a significant coefficient means causality?
How to bypass WRDS DUO verification
Bypass WRDS’s DUO verification
Fastest way to download from WRDS
Downloading with Rsync could be 10X to 100X faster than downloading from Web or PostgreSQL
Pandas may be the bottleneck of your neural network training
Indexing a row from Pandas is order of magnitudes slower than indexing from a native
dict or numpy.arrayAccounting delay of WRDS Factors
10K/Q are only available to the public several days after the fiscal period end, how does WRDS Factors deal with that?
How to Parse 10K and 10Q
How to extract items from 10K/Q files
Nailing down the earnings event time to the millisecond
Find the millisecond-level earnings event time by combining Captial IQ and RavenPack
The "Accepted Date" from WRDS SEC suite doesn't match EDGAR
The issue and my conversation with WRDS
How to query an interval? Understand `foverlap` and `roll` in data.table
How to query the surrounding days given an event?