Quant Research for seeking alpha: Join our journal reading group if you're interested!

Hi all! I received an overwhelming interest for quant research in my previous post: Looking to partner with data/quant analyst to research for alpha. With the same spirit in mind, I'm looking to form a small journal reading group that meets on weekly basis or more frequentely based on the interest of various collaborators. Colloborative research always has a greater advatage in terms of better reach of good papers/ideas and hence mutually beneficial to everyone involved. The idea is to bring together and encourage motivated people to read their own areas of interest, bounce and share ideas. The overarching goal is towards emprical finance, ie. data tested, implementable trading strategies rather than purely academic research.

Full Disclosure: There are currently 3 people who are actively reading research papers. We are amaetuer researchers ourselves. The reading primarly involves reading/discussing/understanding ideas from papers and discusing them. If you have an unrealistic timeline of finding great strategies within 1-2 months then this may not be a good fit.

To give a basic idea of what the paper reading looks like, I've picked out some papers from my recent reading that are sorted in order of readability/difficulty along with contextual comments for a new researcher:

  1. Seasonality, Trend-following, and Mean reversion in 🅱️itcoin - Padysak et al. A simple paper to understand how to systematically make a statistical hypothesis and verify it. Results indicate a very promising sharpe-kinda metric of 2.06.
  2. Barclays US Equity Derivatives Strategy Impact of Retail Options Trading - A well written paper to gauge how a big market participant approaches the market and performs a rigorous hypothesis testing.
  3. Momentum in the Indian Equity Markets: Positive Convexity and Positive Alpha - Srivastava et al. - Written by a quant fund manager in India, not well-written but helps in gauging how a indian quant is approaching the indian market.
  4. Performance of Quality Factor in Indian Equity Market - Jacob et al. - Understanding are some of salient indicative fundmental metrics through factor analysis, in particular the "quality factor".
  5. High Frequency and Dynamic Pairs Trading Based on Statistical Arbitrage Using a Two-Stage Correlation and Cointegration Approach - Miao et al. - An old but introductory read of how to implement a statistical arbitrage strategy along. Results indicate a very promising sharpe of 9.7, not sure how reliable the tests are but most probably the alpha has decayed out.
  6. Optimal Execution of Portfolio Transactions (Currently reading) - Almgren et al. - Covers things like temporary and permanent market impact for optimally liquidating a large position. Interesting to see how if one can estimate/reverse engineer some of these liquidations.

If you're interested in the reading group, please pick out one of the above papers that you'd like to discuss, DM and let's schedule a video google meets to see if its' a good fit. If you're not interested in the above, you can also checkout some ready-made strategies on Quantconnect or Quantpedia or another paper of your liking. Give me a heads-up so that I can briefly go over it as well before we schedule a call. FYI: chatpdf.com helps in efficient reading

Personally, all my research is primarly directly towards the following board statement: "Given a motivated research timeline of 6-8months, with an objective function of say a very small average daily return of say, 1lakh, are there any strategies that a retailer can exploit that are unfeasible for other market participants with higher informational advantage". In short: End-to-end exectuable trading strategy with exaggerated returns over short periods.

To that end, I'm interested along the lines of: illiquid equities-option pricing, measurement of market impact, measurement of information asmmetry, news/event volaility, factor analysis, 0DTE, overnight positioning.

PS: We are all amateur researchers ourselves. Research in a new topic always starts out with a nascent understanding. Therefore, there is no need to worried/be intimidated with fancy financial terms. If you have an interest for quantifiable research, feel free to reach out!
 
@reformedchristian453 I have not gone through individual references mentioned in the footnotes of the capitalmind article, unlikely that I will. Do look at footnote no. 1, maybe you would like to read that book.

Also look at Nifty Alpha 50 Index and the way it is structured.
 
@reformedchristian453 Any time limit to check these papers and get back to you for a call? I'm interested in going through these papers but it would take me some time to sink into them and get back to you as I don't really have free time apart from weekends. Please do let me know!
 
@reformedchristian453 w.r.t these strategies,

i. How do you take care of broker fees / transaction fees ? For intraday strategies, broker fees eats up entirety of my profits.

ii. What timeframe do you look at ? Based on the papers you linked, my guess is that you are looking at 1D prices.

iii. How do you systematically run backtest (accounting for broker fees), live dry run and then graduate the algorithm to test with a small capital ? Do you use any framework / have you built one yourself ?

iv. How do you get high quality data (L2/L3 orderbooks). Are you subscribed to any brokers who are providing this data ? Among the free options, we don't get much fine grained data. Personally for me, It does not make sense to subscribe to paid brokers with a full time job.
 
@merekas I have a python codebase with data for equities from 2018 in 1-second timeframe, options in 1-minute. I had open sourced some of the data previously: here and in the previous article, I had mentioned what data I already have and seems reasaonibly sufficient for out task at hand.

A key limiting contraint of a retailer is good data and infrastructure, which involves lack of the best available market data like historical/live trade, quotes, orderbook etc.

The stat arb paper that I linked deals with 15min -OHLC.
 

Similar threads

Back
Top