Using publishing behaviour to predict market volatility

Why not Sentiment Analysis?

Sentiment Analysis on news as a form of alternative data is becoming mainstream. Data providers have already been offering text-mining products on their platforms and in-house research teams in financial institutions have been working on it for some time. NLP models could also get very complex in trying to interpret the words, phrases and numbers in a report.

What would a number like 130 mean in this context? There are more negative than positive words, does this mean the paragraph has a positive sentiment? Positive for whom? Even after taking into account the context of the passage, a human reader…


A reflection on our current crises

I read the highly acclaimed Extreme Economies by Richard Davies over the weekend and I could not help but draw parallels between the behaviour of the economies he studied and our present-day economies’ response to crises. Although the apparent nature of the economies and crises he studied seem to be very different from today, the agents in our systems share the same human condition in reaction to new shocks: resilience.

My focus is on the first 2 of the 9 economies studied in the book, these are examples of economies which have survived and arised from extraordinary circumstances by adaptation…


A systematic Sector Rotation strategy in US equities

Objective

Investors of hedge funds may be generally concerned by lower liquidity, higher use of leverage, and higher fees than in traditional asset management. With the use of sector ETFs, I aim to mimic hedge funds’ risk/return profiles in sector allocation, but enjoy the benefits of using highly liquid assets with low costs. A sector rotation strategy allows investors to capture returns from different stages of the market cycle while diversifying their holdings by putting more (less) weight into sectors that are rising (falling). Sector ETFs provide a cheap vehicle to invest in hundreds of stocks across different sectors.

Strategy

The strategy…


Insights about Singapore‘s Online VC landscape

Using Social Capital as a resource

Venture Capitalists use a variety of accumulated resources (screening, monitoring, advisors, etc) to improve their strategy & investment processes. Although they may not be as apparent as financial statements or research reports, Social Media presents a goldmine of alternate data VCs could leverage on in the form of Social Capital.

Relevance of Social Network Analysis

SNA is the process of investigating social structures through the use of networks and graph theory. It emerged as a key technique in modern sociology and our goal is to apply this on web data to identify social capital in different actors in the market: investors, start-ups, news agencies, ecosystem…


The future is Nowcasting

Nowcasting Singapore’s Real GDP Growth Rate

According to MTI’s advance estimates, Singapore was projected to contract by 10.6% (QoQ, annualised & seasonally-adjusted) in Q1 2020. Their methodology largely involved computing data in January and February 2020 to serve as an early indication of GDP growth in the quarter. It is also subjected to revision when more comprehensive data becomes available. The GDP is a lagging indicator because it is only officially released a month after the quarter, making it difficult to use in assessing the economy now. Singapore Government Agencies periodically release macroeconomic data, and research analysts have been constantly trying to interpret them instead to…


Quantifying Forward Curve Dynamics with a single metric

The WTI futures market fell below $0 for the first time in history as the May contract (CLK20) went down 306% and closed at -$37.63/bbl on Monday, 20th April. Trading volume and positions had been moving into the June contract (CLM20) which became the front month as the May contract expired the day after on Tuesday, 21st April. The June contract also fell as it became the representative of the short-term market, resulting in a steeper contango.

I was curious as to how contango was commonly calculated as I realised several articles were using spreads from the different back month…


Quantitative Finance for Market Indices

It’s been a whole year since I last posted and I thought to write about a project I’ve done recently during the stay-home period. It was difficult finding practical resources online. There were many math-heavy and theoretical academic papers, but I wanted to support others out there who were looking out for code and a layman’s understanding since I’ve already spent the past 4 days researching and working on this project.

I will leave most of the code out from this post but you can access everything here:

👉🏻Link to Code / Notebook

👉🏻Link to Repo

Index-Tracking Investing

is a passive investment…


with a single function

Scraped news article


Gathered methods to analyse data, diagnose models and visualize results

This analysis was a project which I decided to undertake for the Regression Analysis module in school. I have learnt and gathered several methods you can use in R to take your depth of analysis further. As usual, I always learn the most discovering on my own.

Data

Response variable: Chance.of.Admit

Predictors: GRE Score, TOEFL Score, University Rating, SOP, LOR, CGPA, Research

Link to csv

Libraries

library(dplyr);
library(ggplot2);
library(GGally);
library(vioplot);
library(corpcor);
library(ppcor);
library(mctest);
library(ggfortify);
library(lmtest);
library(MASS);
library(car);
library(DAAG);
library(jtools);
library(relaimpo);

Descriptive stats

summary(df)

Distribution plots

par(mfrow=c(4, 2))
colnames = names(df)
for(name in colnames) {
vioplot(df[name], horizontal=TRUE, col='gold', lineCol='gold', lty=0, colMed='floralwhite', yaxt='n',rectCol='dodgerblue4')
title(main=name)
}


Helping the blind hear what they cannot see.

Final exams are finally over!!! I was fascinated by the idea of object detection in Computer Vision and wanted to start a project on it. I realized that we could probably help the blind “see” better using Image-to-Text and Text-to-Voice, without any complex hardware.

Having a complete understanding of every single piece in the whole pipeline is extremely difficult but I have attempted to understand as much as I need to know for this task.

Object Detection

is a field of Computer Vision that detects instances of semantic objects in images/videos (by creating bounding boxes around them in our case). …

Jason Yip

Financial Analytics Major, Math & Economics Minor from NUS | www.linkedin.com/in/jasonyip184 | www.github.com/jasonyip184

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store