Risk management and position sizing
The answer
Position size, not stock selection, decides whether an account survives. Decide in advance the largest amount you are willing to lose on one position, then let the distance to your stop — the price at which you will sell if you are wrong — tell you how many shares that amount permits.
Why this costs you money
Almost everybody chooses the size of a position by feel. A strong idea gets a big position. A weak idea gets a small one. The number itself is picked because it is round, or because it is what is left in the account.
Here is what that produces.
An account holds Rs 5,00,000. The reader is confident, so Rs 1,50,000 goes into one stock. The stock is volatile, and a sensible stop sits 15% below. That is Rs 22,500 at risk, which is 4.5% of the whole account in a single position. Nobody calculated that number. Nobody would have chosen it if they had.
Now the ordinary thing happens. Six of those go wrong in a row. Six losses in a row is not bad luck; with any realistic accuracy it happens to everyone within a few hundred trades. The account is down about 24%. To get back to where it started it now needs a 32% gain.
The account did not die from a bad stock. Every one of those 6 decisions could have been reasonable. It died from the size.
That is the whole subject. Being right is a variable you do not control. Position size is a variable you control completely. Almost all retail effort goes into the first one.
How it works
Three numbers and one division.
- A — your account value. All of it, including cash.
- r — the fraction of the account you will risk on one position. Start at 1%.
- d — the stop distance in currency per share. Entry price minus stop price.
The number of shares you may buy is:
Shares = (A × r) ÷ d
Work it once and the idea becomes obvious.
An account of Rs 5,00,000 at 1% risk means Rs 5,000 may be lost. Entry at 400, stop at 370, so d = 30. Shares = 5,000 ÷ 30 = 166. The position is worth Rs 66,400, or about 13% of the account.
Now change only the stop. Same account, same entry, but the stop goes at 390, so d = 10. Shares = 5,000 ÷ 10 = 500. The position is worth Rs 2,00,000, or 40% of the account.
Same risk. Same 1%. Position size 3 times larger.
The stop distance sets the position size. That is the mechanism, and it is the opposite of how people work. Most people choose the size first and then look for a stop that fits it. Choosing the stop first turns a feeling into arithmetic.
The cap that must sit on top
The example above should worry you, and it should. A 1% risk rule with a very tight stop permits a position worth 40% of the account. If that stock gaps overnight — opens far below the stop, which happens on results, on news, on a regulatory order — the stop does not protect you at all, and 40% of the account is exposed to whatever the gap turns out to be.
So the risk rule needs a second, independent limit: a maximum position value, regardless of the stop. For most retail accounts, 15% to 20% of the account in any one company is a sensible ceiling.
You take whichever number is smaller. A tight stop earns you a bigger position only up to the cap, and never past it.
The stop that is a round number, and the stop that is a decision
These are 2 different things and they get the same name.
A stop at 370 because 370 is a round number is arbitrary. So is a stop 10% below entry, or a stop at the amount of money you are prepared to lose. None of these have anything to do with the stock. Placing them is a way of feeling protected, not of being protected.
A stop that is a decision sits at a price where the reason you own it stops being true. That price comes from the chart or from the business, not from your wallet. In practice it is one of:
- Below the swing low that defines the current uptrend. Under it, the trend structure you bought is gone.
- Below the level that was resistance and became support after a breakout. Back under it, the breakout failed.
- Below a moving average that has framed the whole advance.
- A multiple of ATR, the Average True Range, which is the stock's average daily movement over a recent period. A stop at 2 or 3 ATR below entry sits outside normal noise.
The test is one question. If the price reaches this level, is my reason for owning it still true? If the answer is yes, the stop is in the wrong place, and it will take you out of a position that is still working.
The ATR version does something the others do not. Set the stop as a multiple of the stock's own daily movement and a volatile stock automatically produces a wide d, which automatically produces a small position. The formula sizes volatile stocks down without you deciding anything.
One more thing about round numbers. Prices like 500, 1,000 and 2,500 attract orders, and many of those orders are other people's stops. Placing your stop just below a round number puts it exactly where a cluster of selling will occur. You are not choosing a level. You are joining a queue.
The mathematics of recovery
This table is the reason the entire discipline exists. It is not an opinion.
If you lose a fraction L of your capital, the gain you need to get back is L ÷ (1 − L).
| Loss | Gain needed to return to the starting value |
|---|---|
| 5% | 5.3% |
| 10% | 11.1% |
| 15% | 17.6% |
| 20% | 25.0% |
| 25% | 33.3% |
| 30% | 42.9% |
| 40% | 66.7% |
| 50% | 100% |
| 60% | 150% |
| 70% | 233% |
| 80% | 400% |
| 90% | 900% |
Read the left column and the right column together and notice where the relationship changes.
Between 0% and about 20%, the 2 columns are close. A 10% loss needs an 11.1% gain. That is an ordinary year. Recovery is normal and unremarkable.
After 30% the columns separate, and after 50% they separate violently. A 50% loss requires a 100% gain — you must double what is left, simply to return to zero progress. An 80% loss requires a 400% gain, which for most people is not a recovery plan. It is a new career.
This is why the goal is not to make money. The goal is to stay in the shallow part of this table. Everything about position sizing follows from that single asymmetry.
What a losing streak does at different risk levels
Six losses in a row, compounded, at 3 different risk settings.
| Risk per position | Account after 6 straight losses | Gain needed to recover |
|---|---|---|
| 1% | down 5.9% | 6.2% |
| 5% | down 26.5% | 36.0% |
| 10% | down 46.9% | 88.2% |
The difference between the first row and the third row is not skill, not research and not stock selection. The same 6 decisions produced all 3. Only the size changed.
Risk-to-reward and why accuracy is not the point
Measure everything in R, where 1R is the amount you risked on that position. Risk Rs 5,000 and make Rs 10,000, and that was a 2R win.
Expectancy per trade = (win rate × average win in R) − (loss rate × average loss in R).
- Win rate 40%, average win 2R, average loss 1R: (0.4 × 2) − (0.6 × 1) = +0.2R per trade. Profitable.
- Win rate 40%, average win 1R, average loss 1R: (0.4 × 1) − (0.6 × 1) = −0.2R per trade. A losing system.
Identical accuracy. Opposite outcomes. What changed was the size of the wins relative to the losses, which depends on where you place the stop and where you take the profit — not on how often you are right. That is why the professional description is "good at losing small" rather than "usually right".
What it tells you, and what it does not
The formula tells you the largest number of shares that keeps your loss inside a limit if the stop executes at your price.
That last condition does the heavy lifting, and it fails in 3 ordinary ways.
Gaps. A stop order becomes a live order when the price is touched. If a stock closes at 400 and opens the next day at 340, your 370 stop sells at roughly 340. You planned to lose Rs 5,000 and you lost Rs 10,000. Results, regulatory news and index changes all produce gaps.
Stop-limit orders that never fill. A stop-limit order will not sell below your limit price. In a fast fall it protects you from a bad price by leaving you with the entire position.
Circuits and halts. In India, a stock at its lower price band may have no buyers at all. The stop cannot execute because there is no trade. The price on the screen is a historical fact, not an offer.
So the honest description is this. A stop is an intention, not a floor. The only thing that genuinely limits your loss is the size of the position.
Where liquidity breaks the formula
A position sized correctly by risk can still be impossible to exit. This is the part most explanations skip, and it is where real damage happens.
The risk formula only knows about the stop distance. It knows nothing about how many shares the market can absorb.
Suppose an account is large and the stop is close. The formula happily permits 40,000 shares. If the company trades 60,000 shares on an average day, your position is two-thirds of a normal day's entire trading. When the stop triggers, your sell order is not a participant in the market. It is the market. You push the price down with your own selling, over several days, and every day you are the reason it fell.
So there are 2 independent constraints, and you take the smaller answer.
- The risk constraint. Shares ≤ (A × r) ÷ d.
- The liquidity constraint. Shares ≤ 5% of average daily volume, if you want to leave inside a day without being noticed.
Almost nobody applies the first. Almost nobody at all applies both.
Then apply the professional haircut: assume liquidity halves on the day you want out. The volume you measured was recorded on a calm day by people who were relaxed. The day you need to sell is not that day.
The decision rule
Before you buy, in this order:
- Fix the stop from the chart, not from your wallet. It goes where the reason
for owning it stops being true.
- Risk 1% of the account.
Shares = (account × 0.01) ÷ (entry − stop).- Cap the position value at 15% to 20% of the account, whatever the formula
says.
- Cap the share count at 5% of average daily volume.
- Take the smallest of the 3 numbers.
If the answer is far fewer shares than you wanted, that is the answer, not a problem to work around. The most common way people break this rule is by moving the stop closer, which does not reduce risk. It only reduces the distance to the exit.
Try this now
Five minutes, and this is the number most readers have never seen.
- Open your largest position. Write down the number of shares you hold and the current price.
- Write down the price at which you would sell if you were wrong. If you have never decided one, write the price you would have to see before you admitted the idea had failed. Everybody has that number. Most people have never said it out loud.
- Multiply:
shares × (current price − that price). This is your rupees or dollars at risk on this one position. - Divide that by your total account value, including cash and every other holding. Multiply by 100.
What you should see. A number far above 1%.
Most readers who do this honestly find 4% to 15%. A meaningful number find more than 25%. If the figure is above 2%, you do not have a position. You have a bet on being right, and the bet is large enough that being wrong 3 times in a row would change your year.
Now do the second half, which takes another 2 minutes and is the real lesson. Repeat steps 1 to 3 for every holding and add the results together. That total, as a percentage of the account, is your open portfolio risk. It is what you lose if everything you own goes to your stop at the same time.
And things do go to their stops at the same time. Ten positions at 1% each are not 10 separate risks if all 10 are Indian small-caps, or all 10 are American technology companies. Correlated positions behave as one position on the day it matters. If your holdings share a sector, a country and a market capitalisation band, add their risks together and treat the total as a single number.
Three real cases
1. Barings Bank, 26 February 1995 — size, not direction Barings was 233 years old. A single trader in Singapore, Nick Leeson, built enormous unhedged positions in Nikkei futures and options, hidden in an account that was not properly reconciled. Losses reached roughly £827 million and the bank was sold for £1. No view Leeson took was exotic. The Nikkei went the wrong way, and the position was large enough that the wrong way was fatal.
2. Long-Term Capital Management, September 1998 — right eventually is worth nothing The fund's partners included Myron Scholes and Robert Merton, who shared the 1997 Nobel prize in economics. Their positions were mostly small mispricings that were reasonable to expect to converge. Leverage was around 25 to 1 before the crisis, with far larger exposure through derivatives. When Russia defaulted in August 1998 the mispricings widened instead of narrowing. Many did eventually converge, after the fund was gone. A consortium of banks recapitalised it in an arrangement organised by the Federal Reserve Bank of New York on 23 September
- Being correct about a price is useless if the size does not let you wait.
3. Archegos Capital Management, March 2021 — concentration plus leverage Bill Hwang's family office held very large, concentrated positions in a handful of shares, built through total return swaps at several banks, so no single bank saw the whole picture. When a few of those shares fell, margin calls forced an unwind that began around 26 March 2021. Credit Suisse alone reported a loss of about $5.5 billion and published a report on the failure in July 2021. Hwang was later convicted on fraud and related charges. Every position was in a real listed company. The failure was in how much of each one there was.
The question that resolves it
A novice asks: how much can I make on this?
An expert asks: how much do I lose if I am wrong, and can I take that 6 times in a row without changing how I behave?
The second question is answerable in 30 seconds and it settles the position size without any opinion about the company.
What would make this wrong
If a drawdown were symmetric — if a 50% loss needed only a 50% gain to recover — position sizing would be a minor comfort rather than the central skill. The table above is arithmetic, so it cannot be wrong the way an empirical claim can.
Here is what genuinely limits the advice.
The 1% rule assumes many positions. It is built for somebody taking 50 or more positions a year, where a run of losses is certain and surviving the run is the objective. A long-term investor holding 8 companies for 10 years runs a different system. Their risk is permanent loss of capital in a business, not a stop being touched, and a stop-based rule does not translate directly.
Small accounts cannot use it literally. One percent of Rs 50,000 is Rs 500. With a realistic stop that permits a position so small that brokerage, taxes and the spread are a large share of the outcome. The honest answer for a small account is fewer positions, longer holding periods, and accepting that the account cannot support frequent trading. The common answer is to keep the trade frequency and raise the risk.
One percent is a convention, not a law. What matters is that the number is chosen before the position, is the same every time, and is small enough that 6 consecutive losses do not change your behaviour. Some professionals use 0.5%, some use 2%. Nobody serious picks it after seeing the chart.
Correlation can make the rule a fiction. Ten positions at 1% each look like a 10% worst case. If they all sit in the same sector, they are closer to a single 10% position. The rule protects you from being wrong about a company. It does not protect you from being wrong about a theme.
In India
Two Indian features change position sizing in ways that imported advice never mentions.
Lot sizes make precise sizing impossible in derivatives. Futures and options trade in lots. You cannot buy 0.6 of a lot. SEBI's October 2024 measures raised the minimum contract value for index derivatives, and lot sizes were revised upward accordingly from late November 2024.
The consequence is concrete. The sizing formula produces an answer like "0.6 lots". The market offers 1 lot or nothing. If 1 lot risks 4% of your account, the correct action is not to round up. It is to conclude that you cannot trade that instrument at that account size. Most people round up and quietly change the risk limit instead, which is how a 1% rule becomes a 4% rule without anybody deciding.
Cash equities are different. You can buy 1 share, so sizing in Indian cash equity is nearly as precise as anywhere. The lot problem is a derivatives problem, and it is largest for the smallest accounts.
Circuit limits can stop a stop from executing. Individual shares have daily price bands of 2%, 5%, 10% or 20% depending on the surveillance category. When a stock is at its lower band with no buyers, no trade occurs. Your stop sits unfilled while the screen shows a price you cannot get. In a small company this can repeat for several consecutive sessions, each one lower.
The whole market can halt too. Index-level circuit breakers trigger at 10%, 15% and 20% moves in the Nifty 50 or the Sensex, with halt durations depending on the level and the time of day. On 13 March 2020 the Indian market hit the 10% lower limit and trading was halted. The evidence. SEBI has published the most direct evidence anywhere of what happens to individual traders without risk control, because the sample is a whole population rather than one broker's clients.
A study released in January 2023 examined individual traders in the equity futures and options segment and reported that roughly 89% of them made losses. A larger follow-up released in September 2024 covered the 3 financial years to 2023-24 and reported a loss rate of roughly 93% across more than 1 crore individual traders, with aggregate losses in the region of Rs 1.8 lakh crore.
Two honest notes. A substantial share of the reported loss is transaction costs rather than market direction, which is itself a risk-management lesson: costs are a guaranteed negative a strategy must overcome before it does anything else. And the studies do not perfectly separate speculation from hedging. Neither point changes the conclusion. A loss rate near 9 in 10, measured across crores of accounts, is not a sampling artefact.
In the United States
Fractional shares make the formula exactly executable. Most large American retail brokers now allow you to buy a fraction of a share, or simply to buy by dollar amount. Robinhood introduced fractional trading in December 2019, Fidelity in early 2020, and Charles Schwab launched "Stock Slices" in 2020. When the formula says 166.4 shares, an American investor buys 166.4 shares. The gap between the calculated position and the executed position, which is a permanent irritation in India, is zero.
The regulator imposes a position-sizing rule of its own. Under FINRA's margin rules, an account that makes 4 or more day trades within 5 business days in a margin account is designated a pattern day trader and must maintain at least $25,000 in equity. This is the only place in either country where a rule directly restricts how often a small account may trade. Whatever you think of it, it exists because of the arithmetic in the recovery table.
Halts are designed to reopen. Individual American shares are covered by the Limit Up-Limit Down mechanism, which pauses a stock for 5 minutes when it moves outside a band, then reopens it, usually with a wider band. Market-wide circuit breakers halt all trading for 15 minutes at a 7% fall (Level 1) and a 13% fall (Level 2), and close the market for the day at 20% (Level 3). Level 1 triggered on 4 separate days in March 2020: the 9th, 12th, 16th and 18th.
Leverage is available and standardised. Regulation T permits 50% initial margin on shares. Portfolio margin, available to larger accounts, permits considerably more. The ability to borrow is the fastest way to convert a correctly sized position into an incorrectly sized one.
Tax interacts with exits. The wash sale rule disallows a loss for tax purposes if you buy the same or a substantially identical security within 30 days before or after the sale. An American investor honouring a stop and then wanting back in has a tax consequence an Indian investor does not.
Where they differ, and what that tells you
Three differences, and each one changes what you should actually do.
1. Whether the answer can be executed. American fractional shares mean the sizing formula's output is always available. Indian derivatives lot sizes mean it usually is not. What that tells you is that the Indian version of this method needs a step the American version does not: if 1 lot risks more than your limit, do not take the position. That step is the whole difference between having a rule and having a preference. The imported version of position sizing was written in a market where the answer is always purchasable, so it never had to say this out loud.
2. Whether the stop can execute at all. The American design pauses trading and then deliberately restarts it, with wider bands, because the objective is to keep a market functioning. The Indian design can hold a stock at a limit price with no trades for an entire session, and repeat that the next day. What that tells you is that an Indian stop is a weaker promise than an American stop. In India, position size is not the best protection. It is often the only one that works. An Indian investor should therefore be more conservative on size than the American textbooks suggest, not less.
3. The quality of the evidence about people like you. American behavioural evidence comes mostly from academic studies of one broker's client data in the 1990s. Excellent work, small samples, 30 years old. India has something no other market has: a regulator publishing population-scale studies of retail derivatives outcomes, updated within the last few years, covering crores of accounts, free to download. What that tells you is uncomfortable. The clearest evidence in the world about what happens to individual traders without risk control was produced in India, about Indian traders, and Indian readers are the least likely to have read it.
Carry this
Shares = (account × 1%) ÷ (entry − stop). Everything else is commentary.- The stop goes where the reason for owning it stops being true, not at a round number and not at an amount of money.
- A 50% loss needs a 100% gain. Stay in the shallow part of the table.
- Cap position value at 15% to 20%, and cap share count at 5% of daily volume. Take the smallest of the 3 answers.
- A stop is an intention. Position size is the only real protection.