How big is a bad day for a share?
Before you buy, it helps to know what a normal bad day looks like for that share - not a crash, just the kind of day that comes about once every two weeks. If your stop or your loss limit is smaller than that, you will be stopped out by noise.
The idea: a 1-in-10 bad day
Our Risk Range answers: on about 1 trading day in 10, how much could this fall? It looks at how much the share has moved each day over the last 60 trading days (its volatility) and multiplies that by about 1.25 for shares and FX, and about 1.09 for crypto. Those multipliers were measured on past data so that the bad day is reached about 10% of the time.
A worked example
A share's daily moves over the last 60 days have a typical size (standard deviation) of 2%. The 1-in-10 bad day is about 2% x 1.25 ≈ -2.5%.
On a ₹1,00,000 or $1,000 position, that is a fall of about ₹2,500 or $25 in one day. If you planned a stop 1% below your entry, a completely ordinary bad day would hit it.
Typical sizes
| Kind of market | Typical daily move | 1-in-10 bad day (approx.) |
|---|---|---|
| Large index (NIFTY, S&P 500) | 0.6-1% | about -0.8% to -1.3% |
| Large company share | 1-2% | about -1.3% to -2.5% |
| Small or volatile share | 3-4% | about -4% to -5% |
| Bitcoin | 2-3% | about -2% to -3.5% |
Ranges for illustration. Each market's real number changes with its recent volatility - look up the one you trade.
How it was tested
A risk number is only useful if it is right about as often as it says. We checked the 1-in-10 bad day and bad week against what happened next on 1,805 markets (shares in 30+ countries, FX, commodities and crypto), about 4.8 million market-days in all. The real fall was worse than the bad-day line on:
| Market type | Bad day hit | Bad week hit | Target |
|---|---|---|---|
| Shares (1,143 markets) | 8.7% | 9.6% | 10% |
| FX (63) | 9.5% | 8.9% | 10% |
| Crypto (469) | 10.7% | 11.0% | 10% |
| Commodities (38) | 9.0% | 10.6% | 10% |
Our own tests on past prices, not customer results; checked a second way on 2.29 million fresh market-days. Past behaviour does not promise future behaviour.
What it does not tell you
- It is not a worst case. On 1 day in 10 the fall is bigger, and crashes can be many times bigger.
- It says nothing about whether the share will go up or down.
- News days (results, budget, rate decisions) are often bigger than normal.
How to use it
Put your stop further than a normal bad day, or accept that you will be stopped often. Size the position so that a bad day costs you no more than you are willing to lose. Our search box shows the 1-in-10 bad day for a share when you type its name.
Bad day vs bad week
Moves add up over several days, but not in a straight line. A rough rule: a 1-in-10 bad week is about 2.2 times a 1-in-10 bad day (the square root of 5 trading days). If a share's bad day is -2.5%, its bad week is roughly -5.5%. If you hold for a week, size for the bad week, not the bad day.
Why recent volatility, not a fixed number
Markets have calm and stormy spells. A share that moved 1% a day last year might move 3% a day this month after bad results. Using the last 60 trading days keeps the bad day close to the market you are actually trading, and it changes as the market changes.
General information about the past and about your own numbers. Not financial advice and not a recommendation. It does not consider your goals or situation.
Frequently asked questions
- What is a normal daily move for a share?
- It depends on the share. Large companies often move 1-2% on a typical day, small ones 3-4% or more. Look at the last 60 days of moves rather than a general rule.
- Is the 1-in-10 bad day the maximum fall?
- No. About 1 day in 10 is worse, and crashes can be much bigger. It shows a normal bad day, not the worst case.
- How accurate is the Risk Range?
- On 1,805 markets and about 4.8 million market-days of past data, the real fall went past the bad-day line on 8.7-10.7% of days depending on the market type, close to the 10% target. These are our own tests on past prices.
- Does it work for crypto?
- Yes, with its own multiplier. In our tests crypto went past the line on about 10.7% of days.
Look up a bad day for your share
Type a company name in the search box. The answer shows the price, how it makes money, and our 1-in-10 bad day for it.