Dushyant ThakkerResearch notes

Market ideas, tested on Indian data.

I take ideas that traders and investors use every day, test them properly on Indian market data, and write up what holds and what does not. Every note shows the numbers, the method and the misses.

Author
Dushyant Thakker, CFA, FRM
Background
About 20 years in various roles at global banks and asset managers, most recently JP Morgan Chase Asset Management
Now
Senior Research Analyst at Definedge and a systematic trader (see About)
Notes so far
5, since August 2026

Start here · studies on Indian market dataAll research papers

  1. No. 4

    Nifty broke its 200-week average. What happened the last five times?

    In four of the five earlier breaks since 1995 Nifty fell another 14% to 26% within a year; from a random week that happened about 3 times in 10. Big rebounds followed only the two crash breaks, 2008 and 2020. This time mid and small caps sit 15% to 18% above their own averages.

  2. No. 2

    The Overnight Drift

    Almost all of the gain came overnight. The overnight gap was positive in every complete year from 2015 to 2025 on both indices. The market-hours part was negative in ten of eleven years on Nifty and nine of eleven on Bank Nifty.

  3. No. 1

    The Phantom Close

    For about fourteen minutes a day the index chart stops updating while futures and options keep trading, then jumps to catch up at the close.

Latest · Research note No. 54 October 2026

131 funds froze in Turkey. What can an Indian investor learn from it?

The data fit funds that bought the shares that lifted their own prices. Their money-market funds paid 53% a year against 46% for peers, and accounts tripled after the Finance Minister's warning. Four checks for any fund you hold.

Dot chart: 12-month returns of Turkish money-market funds. Bank-owned and other funds cluster near 46%; the frozen funds sit further right with a median of 52.6% and a top of 60.2%.
12-month return of Turkish money-market funds to 28 August 2026. Source: TEFAS (Takasbank), my calculation.

All research notesNewest first

  1. No. 5 4 Oct 2026

    131 funds froze in Turkey. What can an Indian investor learn from it?

    How 131 Turkish funds grew 74 times in three years, why the warnings did not stop the money, and what it means for Indian investors in microcaps, ETFs and liquid funds.

    Investor risk
  2. No. 4 2 Oct 2026

    Nifty broke its 200-week average. What happened the last five times?

    Five earlier breaks since 1995, what a 12-month SIP from each one returned, and why this time the broader market looks different.

    Systematic investing
  3. No. 3 20 Aug 2026

    CAS Games and Algo Trading

    SEBI's first closing-auction order, and what one manufactured print does to pivots, ATR, Supertrend and Renko the next morning.

    Market structure
  4. No. 2 17 Aug 2026

    The Overnight Drift

    Split each day into the hours the market is open and the gap overnight. Over eleven years almost all of the gain came overnight.

    Market structure
  5. No. 1 14 Aug 2026

    The Phantom Close

    For fourteen minutes a day the index chart stops while futures and options keep trading, then jumps to catch up.

    Market structure

ArticlesPractitioner pieces, first posted on the Definedge forum

  1. Article 8 Oct 2026

    Ichimoku on Renko charts: three system ideas to test

    Ichimoku's periods become bricks on Renko, so its lines turn into brick counting. Three system ideas and the checks I would run before trusting a backtest.

    Full post on the Definedge forum

    Trading systems
  2. Article 6 Oct 2026

    IVP cheatsheet: when to buy and when to sell index options

    Are option premiums cheap or expensive against their own past year? Which structures suit low, normal and high IVP, and what to avoid in each.

    Full post on the Definedge forum

    Options

Talks and interviewsVideos, with a written summary

  1. Talk 9 Sep 2026

    Can options trading generate income? Notes from my Finding Edge episode

    My Finding Edge episode with Abhijit Phatak: index credit spreads from savings that stay invested, the rules I follow and two backtested strategies.

    Watch the video · Full write-up on the Definedge forum

    Options

How I workStandards, applied where they fit

  1. Point-in-time data. When a test depends on which stocks were in an index, it uses the list as it stood on the day, not today's list.
  2. A base rate for signal claims. When a note says a signal mattered, it compares the result with what random dates would have shown, so a lucky streak does not pass for an edge.
  3. Sensitivity, reported. I vary the assumptions that could change the answer and report what moved. Checks I did not run are listed as limits.
  4. Gross or net, always stated. Each note says whether its returns are before or after costs. When a note tests something you could trade, charges are included. A breakdown of an index's return is not a trade, and the note says so.
  5. The misses get published. Ideas that fail are written up too. They are usually the more useful notes.
  6. Sources you can open. Every paper cited links to a free copy where one exists.
  7. Versions and corrections. Each study shows its version and date and lists what changed. Corrections are welcome at [email protected].

The study summary at the top of each note says which of these checks apply to it.

New notes by email

Get the next research note by email

One email when a new note or talk is published. Nothing else, and you can unsubscribe any time.

Sent with Kit. Your address is used only for these emails.

About

I am a Senior Research Analyst at Definedge and a systematic trader. Before that I spent about 20 years in various roles at global banks and asset managers across Europe, the US and Asia, most recently at JP Morgan Chase Asset Management. I am a CFA charterholder and hold the FRM.

My aim is to bridge academic research and retail investing. I take ideas from published research and from traders' everyday practice, test them on Indian market data, and explain what I find in plain language. At Definedge I lead work on wealth technology and AI-powered research products designed to make market analysis and systematic investing more accessible.

The site has three kinds of work. Research notes each test one question against data and report the numbers, the method, a base rate and the limits; most come with a full paper as a PDF. Articles are shorter practitioner pieces, such as system ideas and checklists, that I first post on the Definedge forum; they suggest what to test and report no results. Talks are videos and interviews, each with a written summary.

I am associated with Definedge Securities Broking Pvt Ltd, a SEBI-registered research analyst (INH000010344) and stockbroker (INZ000301132). What I publish here is my own research and my own views. Where a piece uses Definedge data or discusses a Definedge product, it says so. I do not manage or solicit third-party funds.

The analysis runs on NoiselessOS, my AI-assisted research hub, using data from the external sources named in each note.

If something here is useful, or wrong, I would like to hear about it.