Yanita Oparlakova

A great insight piece in Research Live by our client davies+mckerr head of quant, David Power It's not all about charts and diagrams: “Quant research is tackling real-world consumer challenges, that need to be seen to be understood.” Research is about getting out there, gaining an understanding, asking the right questions and giving the answers that need to be heard. “Let’s win clients over with our thinking from the start and gain their trust to allow you to tell the story how it should be told.” https://lnkd.in/eXjPjXfM

1

“Savoring Stats” is a collection I created to introduce you to the world of statistical modeling in an engaging way. The goal is to convey the mathematical mindset intuitively, connect the dots between concepts, and share the fun I feel while studying it. It’s ideal for anyone seeking an overview of probability and statistics, refreshing core concepts, or ready to dive deeper into statistical modeling. This is the first article, starting with the cornerstone of everything: Probability Distributions.

13

1 Comment

https://lnkd.in/drYVAfn2 -- slightly disturbing, very insightful and shows the power of looking at a problem or subject from different angles using data! #data #analytics #ourworldindata

1

🎤 Speaker Confirmed! Catch David Ardia at our Machine Learning in Quant Finance Conference on 19th September. David's talk will discuss how to transform qualitative sentiment data into quantitative variables and apply them in finance. Don't miss out - book your free ticket today: https://lnkd.in/dRq_cf2N #CQF #CQFInstitute #quant #quantfinance #quantitativefinance #machinelearning #econometrics 

114

4 Comments

Data Valuations and data monetisation benefits to for the new PE ownership rules for the NFL The NFL's new private equity ownership rules have significant implications for data valuation and monetisation. Here's a breakdown: The Data Valuation Importance enables: - Fair Market Value: Data is a valuable asset, but its value can be difficult to quantify. Accurate valuations are crucial for determining fair prices in transactions involving team ownership stakes. - Investor Confidence: Investors need to understand the underlying value of a team, including its data assets. Clear data valuations can boost investor confidence and attract more capital. - Revenue Sharing: Data-driven revenue streams, such as advanced analytics and personalised fan experiences, may require revenue-sharing agreements between team owners and private equity investors. Accurate valuations are essential for fair revenue distribution. - Future Investment Decisions: Data valuations can inform future investment decisions. If data is deemed valuable, it could lead to increased investments in data infrastructure, analytics capabilities, and data-driven product development. Data Valuations Will Drive and Facilitate Data Monetisation for PE ownership by: Data-Driven Revenue Streams: Valuing data can unlock new revenue streams. Teams can monetise data through: - Licensing data to third-party analytics firms - Selling personalised fan experiences powered by data analytics - Developing proprietary data-driven products and services - Strategic Partnerships: Accurate data valuations can facilitate strategic partnerships with technology companies, media organisations, and other businesses that can leverage NFL data. - Increased Team Value: Data-driven insights can improve team performance, leading to increased on-field success and higher valuations, and - Enhanced Fan Engagement: Data-driven fan experiences, such as personalised content and interactive features, can increase fan engagement and loyalty, leading to higher revenue from merchandise, ticketing, and sponsorships.

2

Before anyone read Kimball books or build a data warehouse, they should read Ralph Kimball fundamental article about ER Modeling vs Dimensional Modeling: https://lnkd.in/eD3ab7Ve (thanks to Keith Belanger for reminding all of us today). In this article, Ralph Kimball explained very clearly, why a normalised database such as 3NF should not be used for reporting and analytics. He listed all the issues if we use such database for data warehousing. I know there are exceptions, such as the cases that Golov Nikolai has showed us in his paper with Lars Rönnbäck (Anchor Modeling) but in the majority of cases, Ralph Kimball is right: Dimensional Modeling is far more superior than a normalised database for #datawarehousing, #reporting and #analytics. Ralph wrote in that article: "The resulting master DM model of a data warehouse for a large enterprise will consist of somewhere between 10 and 25 very similar-looking star join schemas. Each star join will have four to 12 dimension tables. If the design has been done correctly, many of these dimension tables will be shared from fact table to fact table." That is amazing. To be able to come up with those figures he must have done many implementations! 10-25 stars, 4-12 dims per star. And that was back in 1997! That's 27 years ago. Amazing. Every #DataArchitect needs to read this article. It is such a fundamental article that everyone in the #Data world needs to read. Thank you Ralph Kimball for this amazing gift: Dimensional Model.

185

35 Comments

Interesting strategy assuming the market is endogenous. Using deep #RL to dynamically adjust portfolio allocation to max. returns while managing risk. An #Autoencoder to extract 3 latent features from the initial set of 12. Continuously learning from new data based on the input distribution. Would be interesting to see the performance when the strategy is exposed to exogenous factors. Missing a public repo for this but still a great read! https://bit.ly/4bwV1zt

1

"Savoring Stats" is my collection designed to guide you into the world of statistical modeling in a fun and engaging way. In this article, we dive into another estimation method: the Method of Moments (MoM). Enjoy!

9

It is the GCSE day in the UK, so let me ask a provocative question. Various ML people are bragging about "ChatGPT" solving math olypics problems or scoring high at some other equivalent math test or writing some "for dummies"- grade code on Python on even Rust. The natural question is, however, how well a prompted and post-training prepped (on the level of a school student) ChatGPT would score in stuff like English or English literature? 😉 #quants #quantitativeresearch #ai #ML #quantitativefinance

1

Here’s a list of my favourite tools for pricing, what I use them for, and their costs. Pyodbc - Accessing data from SQL databases - £0 Polars - Fast data processing and manipulation - £0 Statsmodels - Predictive modelling with GLMs - £0 LightGBM - Predictive modelling with GBMs - £0 MLFlow - Managing modelling workflows - £0 SHAP - Explaining machine learning models - £0 Plotly/Dash - Interactive data visualisation and dashboards - £0 Flask - Web Applications/model serving - £0 GreatTables - Great looking tables - £0 PyTorch - Matrix computation and optimisation frameworks - £0 Pytest - Testing code - £0 Github/Azure Devops - Version control and continuous integration - £0 WSL - power of Linux on a Windows operating system - £0 Mostly Python libraries and what I like about all these tools is that they integrate seamlessly, enabling me to build workflows and systems that can be run with code without any manual steps, meaning they are repeatable and can be automated. These are all popular tools in the wider data analytics field, yet incredibly underutilised in Pricing. However I do expect this to change in the long term. If any of the above sounds interesting, feel free to ask about them below. #pricing #datascience #programming

22

3 Comments

QUICK ANALYSIS: What next for the #Mag7 given the macro shifts we have seen ❓ 👉 We noted a few weeks back in our Linkedin post that the Mag7 12 month Forward PE was 2 standard deviation *expensive* to the overall macro information set (RED CIRCLE) 🚨 Now it has swung to 2 standard deviations **cheap** to macro (GREEN CIRCLE) One can model the PE as a function of economic growth, inflation, nominal & real rates, credit spreads, energy prices, the Dollar and various other measures of investor risk aversion. The chart shows: 1️⃣ The Actual Fwd PE (white line, left scale) 2️⃣ The GAP between the Actual PE and the Macro Warranted PE in standard deviation terms Indeed, you can see this for 12,000 stocks, indices and sectors globally. This has been one enormous swing. The Fwd PE has gone from pricing in strong growth (no landing at all) to a crash landing #mag7 #apple #tesla #meta #alphabet #amazon #netflix #nvidia

22

8 Comments

PSA - Quarto is great for writing more than just Python/R reports - It's awesome for many types of documents. I found it really simple to set up a two-column layout with different languages in each column! The Typst (https://lnkd.in/eeiaZCej) rendering is also much nicer than LaTeX, and even supported non-Latin characters out of the box! If you've never used Quarto, but you use Jupyter notebooks, there are easy ways to take your notebooks and make them render prettily in HTML/PDF in Quarto (https://lnkd.in/echgUyiC) . (For those in the niche intersection of my interests – I am excited because I used it to create a source sheet 📄 for my Shavuot shiur last week ⛰ 🌾 ) Posit PBC Typst #quarto

45

1 Comment

📚 𝐎𝐜𝐭𝐨𝐛𝐞𝐫 𝐁𝐨𝐨𝐤 𝐨𝐟 𝐭𝐡𝐞 𝐌𝐨𝐧𝐭𝐡: "𝐇𝐨𝐰 𝐭𝐨 𝐏𝐫𝐞𝐝𝐢𝐜𝐭 𝐄𝐯𝐞𝐫𝐲𝐭𝐡𝐢𝐧𝐠" 𝐛𝐲 𝐖𝐢𝐥𝐥𝐢𝐚𝐦 𝐏𝐨𝐮𝐧𝐝𝐬𝐭𝐨𝐧𝐞 📚 How well can you predict the future? 🔮 In 𝑯𝒐𝒘 𝒕𝒐 𝑷𝒓𝒆𝒅𝒊𝒄𝒕 𝑬𝒗𝒆𝒓𝒚𝒕𝒉𝒊𝒏𝒈, William Poundstone unpacks the science behind forecasting and introduces us to the fascinating world of superforecasters—people who consistently predict future events with remarkable accuracy. This isn’t just another theory-heavy book. It’s a practical guide on how to:  ➡️ Recognize and overcome biases that cloud your judgment.  ➡️ Apply probability thinking to make smarter decisions.  ➡️ Handle uncertainty with confidence. From business strategies to personal decisions, understanding how to make more accurate predictions can give you a competitive edge. Poundstone distills complex ideas into easy-to-digest concepts, making this book a powerful tool for anyone looking to sharpen their forecasting skills in an increasingly unpredictable world. Looking to think ahead and make better calls? This is the read for you. #BookOfTheMonth #HowToPredictEverything #Superforecasting #WilliamPoundstone #DecisionMaking #Strategy #ReadingForSuccess #BookXcess

11

Succinct article for Time Series Analysis Questions like 'WHY CAN'T WE JUST COMPARE ORIGINAL DATA FROM THE SAME PERIOD IN EACH YEAR?' which never even came to my mind but are important. I assumed that we need to account for account for seasonality by observing the data patterns in the model, but why can't we simply compare same period last year, is explained well https://lnkd.in/gxgkF-Zi

6

"Savoring Stats" is my collection designed to guide you into the world of statistical modeling in a fun and engaging way. In this second article, we dive into everything you need to know about estimators and Maximum Likelihood Estimation (MLE). Stay tuned for the next article, where we’ll explore the Method of Moments and compare it to MLE. Enjoy!

15

What is the difference between a Quant Researcher and Quant Trader? This was a question my client asked me in a meeting yesterday as they are planning on what types of people they need to hire for next year. In short I said it depends as it's very different from Firm to Firm and there is certainly a lot of crossover but this was my attempt at answering. A Quant Researcher typically has PhD in Applied Mathematics or Physics. Their main tasks are to analyse large quantities of historical data to understand market behaviour and find subtle anomalies in order to develop models and strategies. Often they will code with Python and R. A Quant Trader may have more of a Quant Finance background and is focused on implementing and executing trading strategies developed by Quant Researchers or themselves. This a risk taking role and there job is to monitor the market and deliver PnL. They will have strong conviction and be able to act quickly under pressure. Often they will code in C++ / Python / C#. It's worth noting some of my clients who are fully systematic in their approach only have Quant Researchers as a job title in their team other Firms that come from more discretionary background more often have Quant Traders while others have both collaborating with each other. I'm keen to understand people's views of this. Do you agree with this summary? #QuantTrader #QuantResearcher #NordSearch

37

3 Comments

In the latest paper published in Risk.net, Stuart Baumann PhD shows how to compare trading strategies using importance sampling and how the method may be preferrable to the standard A-B test. https://lnkd.in/d6KqDWRz

26

This is definitely an interesting read. No one knows for certain what the impact of AI will be on the job market, though some predictions do seem reasonable. However, what is important is to keep upskilling and leveraging AI to enhance your skills and become more efficient. Here is one interesting extract from the article: "By lightening the load for data scientists, AI could well raise expectations for them to be more creative by giving them the opportunity 'to just try out different ideas.' You have to step up in bringing value, otherwise you could be replaced by someone else who could potentially add more value at a lower cost.” 

1