How much of a stock's move is just the market?
When a share you own rises or falls on a given day, how much of that is about the company, and how much is the whole market or its sector moving?
This is educational content, not a recommendation. It describes a past analysis of a group of stocks as a whole. It names no stock and is not a forecast.
Every stock's price moves for three broad reasons. The whole market moves (interest rates, global news, foreign flows). The stock's sector moves (a bank-wide or IT-wide story). And the company itself does something (results, a contract, a management change). Splitting a day's move into those three parts shows how much of it was really about the company.
The finding
We split the daily moves of 118 liquid NSE stocks into those three parts. Measured as a share of the variance of each stock's daily move:
| Part of the move | Share of variance |
|---|---|
| Market | 31.7% |
| Sector, with the market taken out | 20.7% |
| The stock's own part | 54.2% |
In round numbers: about a third market, a fifth sector and half the stock's own. For a single liquid stock on a single day, close to half of the move had nothing to do with the company specifically.
Method. Universe: 118 liquid NSE stocks. Data: 5-minute price bars. Period: 248 trading sessions before 1 September 2025 (the data starts in September 2024). Move measured: each day's return from the open to the 15:20 bar. Market: the equal-weighted average move of all 118 stocks, scaled by each stock's beta estimated on the previous 60 sessions. Sector: the equal-weighted average move of the stock's sector (20 sectors, hand-built and checked against return correlations, each with at least three members), with that sector's own market beta taken out. Own part: what is left. Past analysis, not a forecast.
Why the three shares add up to more than 100%
31.7 + 20.7 + 54.2 is 106.6, not 100. That is not a mistake; it comes from how the parts were measured.
- Each part's variance was measured on its own. Variances add up to the total only when the parts are completely unrelated to each other. Here they are not.
- The beta comes from the past. Each stock's sensitivity to the market was estimated on the previous 60 sessions, never on the day itself, so it removes the market only approximately. Some market movement leaks into the other two parts.
- The overlap is not assigned to any part. When two parts move together, their shared movement (the covariance) belongs to neither bucket, so the three numbers overlap.
The right way to read the table is as proportions, "roughly a third, a fifth, a half", not as three numbers that should be added.
What this means for one stock
A good day for a stock is often a good day for its whole sector, and a good day for the sector is often a good day for the market. If you judge a company by how its share price moved over a week, a large part of what you see is the market and the sector, not the company. Comparing a stock with its sector index and with NIFTY 50 over the same days separates the two.
It also explains a common experience: owning several banks, or several IT companies, feels like owning several different investments, but on most days they move largely as one.
What changes for a whole portfolio
A portfolio behaves differently from a single stock. The companies' own moves partly cancel each other out: one stock's good results and another's weak quarter offset each other. The market part does not cancel, because it pushes every holding the same way. So the more holdings you have across different sectors, the larger the market's share of your portfolio's ups and downs, and the smaller the share that comes from any stock you chose.
Funds push this further. A NIFTY 50 index fund is almost entirely market. Many actively managed large-cap funds also move mostly with the market, because they hold many of the same large companies. The X-ray described below shows how far that is true of the funds you hold.
Measure your own portfolio
The Portfolio X-ray in the Vitta-Kosh Analytics tab applies the same idea to your own holdings, with two differences. It uses daily prices of the holdings you own today over the last 6 months, 1 year or 3 years. And it splits the variance with regressions (R-squared) rather than one variance per part, so the three shares sum to exactly 100%: market (NIFTY 50, plus the S&P 500 in rupees for US holdings), sectors (NSE sector indices with the market taken out, for direct Indian shares) and your own picks. It also shows each holding's share of the portfolio's swings next to its share of value, and pairs of holdings that moved together.
Limits of this analysis
- It covers liquid, widely traded stocks. Smaller companies were not in the sample.
- It measures moves within a trading day. Splits measured over weeks or months can differ.
- It covers about one year. A market dominated by a single big story, such as a crash, raises the market's share for a while.
- It describes the past. It says nothing about how any stock will move next.
The X-ray splits your portfolio's swings into market, sectors and your own picks.