Slow Diffusion of Information in NSE Stocks
Cross-sectional tests of delayed price response and subsequent return continuation in Indian equities.
Research question
Do returns in NSE-listed stocks display evidence of slow information diffusion, and can delayed price response help explain subsequent return continuation?
Why this matters
If information reaches prices unevenly across stocks, the stocks it reaches slowly should continue to move after the information arrives. That is an economically motivated reason to expect continuation, and it is distinguishable from continuation caused by illiquidity or by size effects — but only if those controls are built into the test from the start.
Hypotheses
- 01H1: Stocks with slower measured price response to market-wide information exhibit stronger subsequent return continuation.
- 02H2: The effect survives controls for liquidity, market capitalization, and turnover.
- 03H3: The effect is weaker after realistic transaction costs at monthly rebalancing frequency.
- 04H4: The effect is not confined to a single subperiod of the sample.
Methodology
Economic motivation
Slow diffusion predicts that prices of less-followed or less-liquid stocks incorporate common information with a lag, producing predictable continuation in the following period.
Signal formation
A delay measure is estimated per stock from the response of its returns to lagged market returns, then used to sort stocks cross-sectionally at each formation date.
Rebalancing
Portfolios are formed and rebalanced monthly. This is a monthly-rebalanced specification, not a daily-rebalanced strategy, and turnover is reported on that basis.
Liquidity controls
Stocks below stated traded-value and price thresholds are excluded before sorting, and liquidity enters the panel regressions as a control rather than only as a screen.
Market-cap controls
Sorts are performed within size groups, and size is included as a regression control so that the delay effect is not a restatement of a size effect.
Portfolio tests and inference
Long-short spread portfolios across delay groups are tested with Newey–West standard errors; panel regressions cluster by date and stock.
Transaction costs
Costs covering brokerage, statutory charges, and spread are applied to monthly turnover, and results are shown gross and net with a sensitivity range.
Data
Universe
[UNIVERSE DEFINITION TO CONFIRM] NSE-listed common stocks meeting stated liquidity and price filters at each formation date.
Sample period
[SAMPLE PERIOD TO CONFIRM]
Frequency
Daily prices and volumes, monthly formation points.
Data source
[DATA SOURCE TO CONFIRM]
Cleaning and validation
- Corporate-action and split adjustment checks
- Suspension and trading-halt handling
- Price and volume outlier screening
- Symbol-change mapping over time
Delisting and survivorship
Delisted stocks are retained until their final trading month so that portfolios reflect the investable universe at formation, and delisting returns are handled under a stated rule.
Key results
No results published yet
Long-short spread (net)
—
Awaiting final output
t-statistic (clustered)
—
Awaiting final output
Monthly turnover
—
Awaiting final output
Maximum drawdown
—
Awaiting final output
Planned exhibits
- Delay-group portfolio returns with confidence intervals
- Panel regression tables with clustered errors
- Subperiod stability comparison
- Turnover and gross-to-net cost waterfall
- Liquidity and size control diagnostics
Every published chart will carry a descriptive title, axis labels, units, legend where needed, its sample period, a gross or net label, a short written interpretation, and accessible colors with tooltips.
Interpretation
What the tests can show
Whether continuation is stronger among slower-responding stocks after controls, and whether the spread survives cost assumptions.
What I would infer
[INTERPRETATION TO COMPLETE AFTER RESULTS] Any diffusion interpretation requires the controls to hold; without them, illiquidity is the simpler explanation.
What remains uncertain
Whether the delay measure captures information diffusion or measurement noise in thinly traded stocks.
Robustness checks
- Alternative delay estimation windows
- Alternative holding periods
- Stricter liquidity filters
- Size-group subsamples
- Subperiod tests
- Cost sensitivity
- Bootstrap inference
- Multiple-testing correction across sort variants
Limitations
- Indian equity history includes structural and regulatory changes within the sample.
- Delay measures are noisy for less liquid stocks, the very stocks the hypothesis concerns.
- Short-selling constraints and borrow availability limit implementability of the short leg.
- Cost estimates are assumptions rather than realized fills.
- Survivorship and delisting treatment can materially affect measured spreads.
- Capacity is limited in the smaller-cap segment.
Conclusion
[CONCLUSION TO COMPLETE] Will be written after results and robustness checks are final, with explicit separation between the statistical finding and its economic interpretation.
Reproducibility
Repository structure
[REPOSITORY LAYOUT TO CONFIRM] Ingestion, universe construction, signal estimation, portfolio tests, and regression modules.
Data requirements
An authorized NSE price and volume history in the documented schema; the repository does not ship licensed data.
Output generation
All tables and figures regenerate from a single command against cached intermediates.
Downloads and links
Related research
Citation
This is working research, not a peer-reviewed publication. If you refer to it, please cite it as work in progress and note the status shown above.
BibTeX
@misc{bang_nse_information_diffusion,
author = {Bang, Pratik},
title = {Slow Diffusion of Information in NSE Stocks},
year = {2026},
note = {Status: In Progress. Working research, subject to revision.},
url = {[PROJECT URL TO ADD]}
}Plain text
Bang, P. (2026). Slow Diffusion of Information in NSE Stocks. Working research (In Progress). [PROJECT URL TO ADD]